Mathmuse's page

Thank you for your experiences. So if I summerize it, it'd be something along these lines:

• Free Archetypes are in general horizontal power, so in many cases they player have more options, but won't be outright better at what they do anyway
• be beware of the archetypes and options that do give more vertical power, like more HP due to multiclass resiliency feats or feats like acrobat dedication, which grants legendary acrobatics in the end
• having no restrictions on archetypes can encourage powergaming behavior ...

With tactical players, horizontal power, i.e.,having more tactical options but their numbers are not better, is as effective at winning combats as vertical power, i.e. better numbers. For example, if the players with piercing weapons find themselves battling skeletons that resist piercing and slashing damage, having a cantrip that deals acid damage is a winning alternative. For an example from my campaign, Roshan's rogue build specializes in debuffing a single strong opponent, because if she is faced with multiple weak opponents, she can switch to casting cone-shaped spells from her free archetype. Also, the spellcasters in my party stay at range and do not flank, but bard Jinx Fuun will send her bird animal companion from her druid free archetype to flank for the melee rogue and champion.

Nevertheless, horizontal power causes the players to invent clever tactics. That entertains me as a GM. Vertical power is less interesting to watch.

Did you let the Starlit Span Magus free archetype Psychic and take Imaginary weapon?

No, Zandre took the Dragon Disciple archetype. Her backstory is that her Ekujae elf village in the Mwangi Expanse was attacked by a dragon and Zandre decided to become a dragon hunter.

As a dragon disciple, you study and learn from the example of dragons, but that doesn't necessarily mean that you admire the dragons you emulate. While most dragon disciples do indeed revere their chosen dragon, countless tales are told of dragon disciples born of great personal tragedy at the hands of boundless draconic fury. These disciples have transformed themselves into what they despise in order to gain the power to destroy their draconic tormentor and at long last earn their revenge.

However, despite the possible anti-dragon theme the archetype is about becoming more like a dragon, which does not mix will with an archer magus.

Furthermore, though I offered more flexibility on the Strength of Thousands Player's Guide recommendation of Druid Multiclass or Wizard Multiclass archetypes, I insisted that their free archetypes had to do something with arcane or primal magic. Dragon Disciple barely qualified because of the arcane spells granted by Dragon Arcana feat 4. It was one of the most martial options, and I suspect that that is why Zandre's player chose it. Psychic Multiclass archetype offers occult spellcasting with no arcane or primal magic, so it was not an option.

And if Zandre had taken Psychic Multiclass for extra focus points and imaginary weapon, I would not worry about her combat prowess. Instead, Zandre usually uses plain cantrips, such as ignition, with her spellstrikes.

And the Diamond Distributor Bankrupty has disrupted some standard distribution channels, so simplifying the stocking Paizo publications in the stores will aid the resumption of business.

I have loved free archetype in the games that I GMed, and didn't feel that it moved the power needle much -- largely because most of the folks in my homegames aren't that rules savvy, so they chose stuff that resonated with their characters.

My own players are good at rules. The four that embraced the free archetypes (see comment #9 above) fortunately use their skill with character design for roleplaying rather than powergaming. As for the other three, one does not put time into character development, and another is a newbie enjoying experiencing Pathfinder as is without major plans.

The third is the smartest in our group, a genius biophysicist, and his Starlit Span magus Zandre is built for combat power. But he does not browse through lists of feats to find the most powerful options. I used to play weekly boardgames with him, and he sticks with one strategy that he planned from the beginning rather than switching strategies in Eurogames where the winning tactic is to constantly adapt to circumstances. (My wife wins these 4-player boardgames half the time, because she adapts constantly and cleverly.)

And the player characters are leveling up to 8th level this week for the module Hurricane's Howl, which has more emphasis on combat than the previous two modules. Thus, I pointed out to Zandre's player that Zandre has three Dragon Disciple archetypes already, so he could expand out to a 2nd archetype in his free archetype slots, and I said it did not have to be magical like the 1st archetype. Then I directed him to Ascalaphus's comment #15, which describes how to get more combat power from archetypes.

He decided that Zandre will take Rogue Multiclass Dedication for his 8th-level free archetype feat, one of Ascalaphus's examples. Zandre often shot her spellstrike arrows from hiding for the reduced AC from off-guard and a rogue's surprise attack would fit her favorite tactics.

Yes, free archetype can be exploited for additional combat power. But sometimes that works with the campaign. The magus Zandre took the heavy hitter role in the party, the champion Wilfred is the defender, the bard Jinx is the healer, the bard Stargazer is the buffer, the rogue Roshan is the debuffer, and the wizard Idris and kineticist Cara are wild cards. I want the heavy hitter at the top of her game so that the others can have fun in their chosen roles, so I gave him a hint to step up his game.

These Paizo forums are helpful in unexpected ways.

The Strength of Thousands Player's Guide recommends the free archetype but limited to druid multiclass and wizard multiclass. The PCs are students at the Magaambya Academy, which teaches arcane and primal magic. Those archetypes would gives PCs with non-arcane non-primal classes the ability to cast arcane or primal spells, to explain why they are at the Magaambya. I decided to open the options wider, and just said that their archetype had to do something with arcane or primal magic. The result was:

Cara'sseth Ti'kali, a catfolk fire kineticist with wizard free archetype
Idris, an anadi divination wizard with Magaambyan attendant free archetype
Jinx Fuun, a tengu enigma bard with druid free archetype
Roshan Azar, a fleshwarp eldritch trickster (elemental sorcerer) rogue with Gelid Shard free archetype
Stargazer, a ghoran enigma bard with druid free archetype
Wilfred Eugenus Rosehill-Aglag, a dromaar redeemer champion with magus free archetype
Zandre, an elf starlit-span magus with dragon disciple free archetype

Cara, Idris, and Jinx have strongly roleplayed their archetypes. Kineticist Cara wanted to become a wizard in her youth, but faulty training led to her tapping into fire impulses rather than arcane spells. Now she is at the Magaambya to learn proper wizard spellcasting, but kinetic fire impulses are still her main tool. Bard Jinx is more about druid culture rather than druid spells, though as the party healer she likes the extra Heal spells from the archetype's spell slots. She used druidic Animal Order to gain an animal companion. Wizard Idris picked a non-multiclass archetype directly related to the Magaambya Academy. He acts like a typical academic student, spending long hours in the library studying the Magaambyan tradition of combining arcane and primal magic. Idris's player has him prepare spells related to his classwork and then creatively use them for combat.

Rogue Roshan is weird. Her player was the main advocate of the Strength of Thousands campaign, because she wanted a rogue character with two spellcasting archetypes. She needed a free archetype to manage that. Her Gelid Shard archetype gives her spontaneous arcane cold spells and her Sorcerer Multiclass archetype she devoted to primal fire spells. In combat, she mostly uses her Athletics skill for grappling and tripping, and then punches the off-guard opponent for sneak attack damage (her ranged-spell teammates are seldom in position for flanking). Her archetype spells are for less frequent occasions when she needs a ranged attack or an area of effect.

Stargazer, Wilfred, and Zandre barely remember that they have a free archetype.

My wife, player of Jinx, was the other strong advocate for Strength of Thousands. In my previous PF2-converted Ironfang Invasion adventure path, her character Sam was a Charisma-based scoundrel rogue with Sorcerer Multiclass archetype. This campaign did not have free archetypes. My wife said that she constantly struggled with the multiclassing, because the needs of the rogue class and the needs of the sorcerer archetype both wanted her one-per-two-levels class feat slots. She wanted the free archetype so that she could multiclass without the conflict. Two other PCs in the Ironfang Invasion campaign took archetypes at 16th level, but that was after they had gotten everything they wanted out of their class feats so they did not feel the struggle.

Thus, in my experience the free archetype is an option to open up more roleplaying without costing the character their class feats. And one in seven players went weird with it.

The biggest problem with Staves is their spell ranks trail behind their item level, and it only gets worse as you level up further. Blasts and incapacitation effects won't be useful when they are stuck at 3-4 ranks below what you can cast. So you want spells which are useful at any level without heightening. This will also let you maximize your uses per day thanks to how charges work.

That is the main reason why a staff for the 6th-level primal sorcerer with bard dedication seemed the most interesting. The occult spells of the Bard Multiclass already lagged behind the spell ranks of a primary tradition. Spellcasting Archetypes give a 1st-rank spell at 4th level, 2nd-rank at 6th level, 3rd-rank at 8th level, 4th-rank at 12th level, 5th-rank at 14th level, 6th-rank at 16th level, 7th-rank at 18th level, and 8th-rank at 20th level. That is 3 levels behind at best. Staff ranks look good in comparison.

Levels at Which Spell Ranks Become Available
Rank Archtype Staff Wand Class
1 4 3 3 1
2 6 6 5 3
3 8 8 7 5
4 12 10 9 7
5 14 12 11 9
6 16 14 13 11
7 18 16 15 13
8 20 18 17 15
9 __ 20 19 17

Or sorted another way
1st level: spellcaster class has 1st rank, some items have cantrips
2nd level: archetype has cantrips
3rd level: class gains 2nd rank, wands and staves have 1st rank
4th level: archtype gains 1st rank, more 1st-rank staves
5th level: class gains 3rd rank, wands gain 2nd rank
6th level: archetypes and staves gain 2nd rank
7th level: class gains 4th rank, wands gain 3rd rank, more 2nd-rank staves
8th level: archetypes and staves gain 3rd rank
9th level: class gains 5th rank, wands gain 4th rank
10th level: staves gain 4th rank
11th level: class gains 6th rank, wands gain 5th rank, more 4th-rank staves
12th level: staves gain 5th rank, archetype gains 4th rank
13th level: class gains 7th rank, wands gain 6th rank
14th level: staves gain 6th rank, archetype gains 5th rank
15th level: class gains 8th rank, wands gain 7th rank. more 6th-rank staves
16th level: staves gain 7th rank, archetype gains 6th rank
17th level: class gains 9th rank, wands gain 8th rank, more 7th-rank staves
18th level: staves gain 8th rank, archetype gains 7th rank
19th level: class gains 10th rank, wands gain 9th rank
20th level: staves gain 9th rank, archetype gains 8th rank
The current inventory of staves lacks 5th-, 9th-, 13th-, and 19th-level staves.

My wife, who plays Jinx Fuun in my Strength of Thousands campaign (Hilary Moon Murphy might recall it from Common Sense Versus The Plot), decided that Jinx would organize a round-robin gift exchange between the party members for Winter Solstice. The gift price was capped at 80 gp unless the gift-giver has an excuse how they got the item cheaper than its list price, such as crafting it themselves or Bargain Hunter. I added temporary NPC party member Virgil Tibbs to the gift exchange to join in the fun.

And I noticed that some of the staves are musical instruments, which are pretty cool. I do like the look of the Pipes of Compulsion or the Seer's Flute.

And I also found the Coda musical instruments that act like staves in the hands of a bard: "Instruments with the coda trait work mostly like staves and have the staff trait. There are two differences: Coda instruments are in the form of musical instruments, and they can be prepared only by bards. Because they're not physically staves, you can't attack with a coda instrument, nor can you etch it with weapon runes." Virgil Tibbs gave a Trickster's Mandolin to the bard Jinx Fuun as a gift. It was a little above the price cap at 90 gp, but Virgil had the excuse that it was auctioned off at a discount by the police station after being confiscated from a trickster criminal.

Entertainer's Lute would have been more appropriate for singer, actor, and playwright Jinx Fuun, but Trickster's Mandolin seemed a little more practical. Prestidigitation is often on a cantrip repertoire for non-combat utility when pulling out a musical instrument would be no trouble, and would free up repertoire space for a combat cantrip. Illusory Disguise and Item Facade would help in stagecraft for a play.

My impression about staves in Pathfinder 1st Edition is that they were handy but terribly high priced. Thus, the party would keep a staff if found, but never buy one. I have not analyzed the price of staves in Pathfinder 2nd Edition, so let me do a quick analysis. Suppose a spellcaster wants to cast 2nd-level Translate. A Wand of 2nd-level Translate, item 5, would cost 160 gp. A Librarian Staff, item 6, can also cast 2nd-level Translate and costs 225 gp. And it has the option of other spells, such as Pocket Library, a favorite among my PCs. Okay, the price is comparable to wands. In contrast, a PF1 Staff of Understanding costs 14,400 PF1 gold pieces and a PF1 bard's Wand of Tongues costs 6,000 PF1 gold pieces and can be used multiple times per day.

An interesting exploit for the 6th-level primal sorcerer with bard dedication is that the sorcerer can cast occult spells only up to 2nd rank via Basic Bard Spellcasting, even when using a staff of occult spells. And a 2nd-rank spell from a staff costs 2 charges. On the other hand, the sorcerer preparing the staff will give the staff 3 charges, because the Preparing a Staff rules don't mention spell tradition and the sorcerer can cast 3rd-rank primal spells. At 7th level the sorcerer would get 4 charges for free each day. A staff or coda would noticeably enhance the Bard Multiclass archetype.

I'm gonna be honest, I just straight-up tell my players what RK skills would be appropriate. I'm pretty generous with Recalling anyway, because the conservative reading of the rules is kinda bad.

Likewise for me, but my alteration of Recall Knowledge has a long history behind it. First, I also do not bother with secret rolls. My players are good at roleplaying mistakes, such as recalling erroneous knowledge.

Second, during the public playtest of Pathfinder 2nd Edition, my wife wanted the player character's background and backstory to matter more. The one place we were able to implement it was in Recall Knowledge. If a wizard from a wizard school succeeded Recall Knowledge, I would emphasize information that would come out of textbook. If a barbarian from a tribe succeeded at Recall Knowledge, I would emphasize information that was told as stories around the campfire, such as, "You recall a hunter in the tribe telling stories about fighting this kind of creature. He told of its deadly sting and the venom's effects on his teammate. It does ..." I have not had as much fun this this in my current Strength of Thousands campaign, because all the characters are students at the Magaambya School of Magic, so they study similar knowledge.

Third, I also give three times the information recommended in the Pathfinder 2nd Edition Core Rulebook, because I have more fun when the players know what they are fighting.

As for which skill my player non-secretly roll for Recall Knowledge, I tell them the skill from the creature entry. Sometimes I give them a list, because some entries give a list; for example, identifying a Hell Hound is DC 18 in Arcana, Nature, or Religion. Nevertheless, they will sometimes negotiate, such as, "Can I roll Society on this shady count rather than Religion, because he has been involved in the society of this town?" And if I allow it, I will tell a tale about his political machinations and the loyalty of his staff rather than on his supernatural vampiric abilities.

I did not change my system for the Remaster, which emphasizes a specific question. That is better than the pre-Remaster recommendation of giving the most common knowledge. However, I like tailoring the answer to what the PC would care about based on their roleplayed priorities.

In conclusion, house rules on Recall Knowledge do work out when they make the action more flavorful.

This thread has been enlightening. Thank you for the discussion.

I came here from a weird direction. I am running a Strength of Thousands campaign with seven players. Due to spreading the experience points among more than four players and not able to plausibly increase the size or difficulty of every encounter to make up for that, I have been adding extra events to Strength of Thousands, such as a sequel to the 2008 module River into Darkness. My current additional event is that the party will assist in the preparations for the Convocation of Rival Academies featured in Lost Omens: Rival Academies. That book has a chapter about having PCs attend the Convocation, but Strength of Thousands starts four years before the Convocation and even after the first two modules, it is still three years in the future. Thus, they will help with preparing the Magaambya's biggest display at the Convocation, the Tree of Stories. The PCs will help transplant a kapok tree from the Magaambya campus in the Mwangi Expanse to the Academy of the Reclamation in Mendev, 2,500 miles away.

That is beyond the abilities of the spellcasting teachers currently named at the Magaambya, who can manage 1,000-mile 7th-rank Teleport spells but not the longer-range 8th-level ones. But I once ran Rise of the Runelords, and continued that campaign to 20th level with The Witchwar Legacy. All my past campaigns are canon to my present campaign. Furthermore, the Sihedron Spires, a New Thassilon academy, will participate in the Convocation, so they have connections to my Rise of the Runelords characters. The new PCs could team up with the old PCs--I have two current players who participated in the previous campaign. This would add the flavor of yet another academy to the adventure. I have done a team-up before between Rise of the Runelords and Jade Regent: What Happens to the Former Heroes of Adventure Paths, comment #24.

However, I did not run the 2018 adventure path Return of the Runelords. I read the entry on Belimarius and Sorshen in Lost Omens: Legends, but that does not tell give me enough information about their personalities for their encounter with the 20th-level Rise of the Runelords PCs in my incarnation of Golarion. I mostly need to tell the tale of whether the ancient runelords fought or cooperated. And I remembered seeing this thread about Sorshen, so I am here to learn.

In my own campaigns, I had a druid in Ironfang Invasion worried whether his fireball would set the forest on fire. I told the player that looking at the text, the fireball spell was specifically designed to hurt creatures rather than objects, so the forest would not catch on fire.

A wizard in my Strength of Thousands campaign had the same fears, but about damaging books in a library. This was a Magaambya Academy library, so the library ought to have strong fireproofing spells.

The Strength of Thousands campaign has a rogue with Magical Trickster feat that allows sneak attack with spell attacks against AC. She noticed that higher level spells seldom target AC. We have plans for a custom spellshape that will let area of effect spells concentrate on a particular target's AC in order to allow sneak attack. Thus, in my campaigns the damage from spells is malleable, to a small degree without a feat. Thus, the spellcaster could choose whether objects are protected or vulnerable in an area of effect.

Claxon wrote:

Finoan wrote:

Something I wish the rules made more clear is the idea of a non-combat Encounter mode. The point in the narrative where you are tracking events in the very short amounts of time measured by actions, but there isn't an initiative order.

I can't recall the wording, but I think the rule book makes it clear that any time where time matters, and tracking individuals actions are important is exactly what Encounter mode is, it doesn't even matter if there are no NPCs present.

I would agree with that. The rules for non-combat Encounter mode do exist... if you read the rules very, very carefully.

I just wish the rules made the idea more clear. So that more GMs would use it when it is appropriate.

My game session yesterday was entirely non-combat.

After some conversations with other students and with teachers, the 7th-level party headed to the North Docks of Nantambu. Those docks are not canon to Golarion lore; instead, I had added them for a runesmith playtest encounter. I had explained that the North Docks were the slums of Nantambu, because they flood occasionally. Therefore, when winter rainy season arrived, I had a flood. I altered the map to fill the roads and gabs between buildings with water. The party was forewarned that the rain would bring a flood and had volunteered to help.

Relevant to this thread's issue of the gap between Exploration mode and Encounter mode, I asked my players if their rescue of people trapped atop buildings should be in Exploration mode or Encounter mode. Their ship was already docked at the submerged docks, unable to go into the neighborhood itself because the flood was too shallow. They selected Encounter mode. We rolled initiative for player characters and other rescuers. No opponents rolled initiative because no-one was an opponent. Townsfolk in trouble acted on command after a Diplomacy or Intimidation check.

The kineticist Cara with wizard archetype cast Enlarge on the champion so that he could wade through the water with his Athletics and pick up people. The actual wizard Idris cast 4th-rank Water Walk on the entire party, but I said the flood waters were difficult terrain. The bard Jinx Fuun, whose backstory had her born on an ocean-going ship, brought a folding rowboat from Magaambya supplies and used her Sailing Lore to row about the neighborhood while maintaining a Triple Time song to speed everyone up. Idris had a second 4th-rank spell, Sliding Blocks, that made platforms for jumps and could move some slowly. The players pointed out that the Triple-Time speed meant that they could Leap 15 feet between rooftops. Some PCs simply leaped over to diplomatically or intimidatingly persuade panicked people to leap onto blocks or the rowboat.

The two advantages of doing this in Encounter mode rather than Exploration mode is that it clarified the timing of who could rescue particular people the soonest, and made stunts that sped up the PCs more dramatic. For example, the champion had Quick Jump and discovered that Long Jumps were faster than Striding over difficult terrain.

I occasionally throw a creativity challenge like this at my players. They enjoy the chance to stretch their imagination.

As for the transition between Exploration mode and Encounter mode, I had a houserule that simplified some issues. Say the player had listened at a door in a dungeon during Exploration mode and knew that enemies awaited in the next room. They would plan for one PC, typically the rogue who had checked the door for traps, to open the door and then the martial characters would rush in. But left to the initiative dice, the rogue could roll lower than the martials, so the martials would spend their first turn saying, "I wait until the door is open." That was boring and eliminated the advantage of rolling high initiative. Instead, I would allow an inciting event in Exploration mode. The rogue would open the door and then we would roll initiative. Or in a different scenario, with the party finding the bandit camp in the forest and still being unnoticed, the inciting event would be the wizard casting Fireball on the camp while the bandits were all grouped together. The inciting event was not a full turn; instead, it was merely enough actions to perform the event.

My take on it is that the magic warrior archetype can represent a historical time when the 10MW fought the King of Biting Ants and they needed to hide their identities to prevent the King from going after their families etc., but that after he was defeated this wasn't so important anymore. It's a canonical point that armies don't make it to Nantambu because the Tempest-Sun have very sufficient firepower.

So most Magaambyans use the "new" mask practice of using the mask to express themselves. The "old" tradition of using the mask to disguise still exists, and might be useful to agents operating further afield in hostile territory.

Another factor about the Magic Warrior tradition changing over time is the vast amount of time. The Old-Mage Jatembe and the Magic Warriors founded the Magaambya Academy in -2832 AR, The Strength of Thousands adventure path starts in 4721 AR. That is 7553 years. In that amount of time the Magic Warrior tradition could have been lost, restored, altered, and reverted over and over again. The only anchor points across the millennia are the stories and the masks.

So for me, the point of the archetype isn't to turn into an animal, or to make you a martial. It's about being able to be a resistance warrior against evil overlord wizards.

Ah, evil overlords. Peter Anspach, one of the authors of The Evil Overlord List, had been my boss 20 years ago. He was a good boss. He no longer updated the list, but he still supported the website.

Resistance warrior against powerful enemies of the Mwangi Expanse is a good theme.

Mathmuse wrote:

I am fairly sure that the developer's intent for the Magic Warrior was to illustrate the Mwangi Expanse rather than add another class archetype to magus.

I'm not really seeing why these two things are mutually exclusive. Would a flavorful Magus class archetype not be an awesome way of illustrating the Mwangi Expanse?

Class archetypes are not as fun as general archetypes.

The PF1 class archetypes came out of the game design standards of Paizo re-inventing Dungeons & Dragons 3rd Edition as Pathfinder 1st Edition.

Multiclassing was beloved by players, but it could lead to abusive combinations in Dungeons & Dragons. Furthermore, high-level 3rd EditionD&D often had so-called dead levels that offered nothing interesting except +1 to a few numbers, which made a switch to multiclass tempting at those levels. Thus, Wizards of the Coast put penalties on multiclassing in D&D 3rd Edition. Multiclassed characters would earn fewer experience points. (In retrospect, that was bad design.) Yet to allow the fun of multiclassing when it was not disruptive, Wizards of the Coast developers invented ways around the penalties such as Prefered Classes for various races. For example, elves could multiclass to wizard without penalty. Prestige classes were another workaround. They were deliberately designed for multiclassing and had no penalty, but they had strict prerequisites that restricted their use. Sometimes a prestige class was designed to fit with only one base class.

Paizo removed the multiclass penalties from Pathfinder 1st Edition and tried to design each class to simply encourage players to stick to one class to enjoy the high-level benefits. This left prestige classes in an awkward place. Players still wanted to play variants on classes like the prestige classes offered. So in the PF1 Advanced Player's Guide Paizo developers offered class archetypes. They let the players play a variant without having to multiclass.

The tight math in Pathfinder 2nd Edition was incompatible with classic multiclassing. Instead, multiclassing was by general archetype, based on a PF1 optional system from Pathfinder Unchained called Variant Multiclassing. Since these PF2 archetypes were intended for multiclassing, the developers saw little reason to force an archetype to stick with a single class as a class archetype.

I said "little reason" rather than no reason, because some exceptions appeared. Flexible Spellcaster and Wellspring Mage from Secret of Magic (September 2021) each required a specific style of spellcasting from 1st level and altered that spellcasting at 1st level before the player formally took the archetype. Elementalist from Rage of Elements (August 2023) had similar requirements. Spellshot from Guns & Gears (October 2021) was the first true PF2 class archetype. It required the Gunslinger's Way of the Spellshot as a prerequisite, so it was only for Gunslingers. That opened the gates for later class archetypes in War of Immortals (October 2024) and Divine Mysteries (January 2025).

Magic Warrior could be a class archetype in the same sense as Flexible Spellcaster, Wellspring Mage, and Elementalist. Give it a prerequisite that the base class must offer martial weapon proficiency and spell slots at 1st level. Then the character would already have magic and be a warrior as a Magic Warrior and the archetype would simply offer a few tricks with the Magaambyan mask. The only classes that fit that are bard and magus, because warpriest cleric does not gain martial weapon proficiency until 3rd level. A rogue like Noor Khan could never become a Magic Warrior, so I would have to give her a Wizard Multiclass archetype instead, less flavorful for a spy. I refer to see alchemist Magic Warriors, barbarian Magic Warriors, champion Magic Warriors, druid Magic Warriors, exemplar Magic Warriors, fighter Magic Warriors, etc., implying that Magic Warriors can come from all walks of life.

In summary, class archetypes perform a narrow modification of particular class features. Magic Warrior is not modifying particular features; instead, a magus class archetype Magic Warrior would be piggybacking on an existing gish class as a shortcut to obtain both magic and martial prowess. It would narrow the flavor of Magic Warriors.

Random thought, but if the intent here for some people is to make the Magic Warrior play like a Magus, would that not be a great opportunity for a Magus class archetype?

The intent behind Magic Warrior is regional favor about the Mwangi Expanse. According to the Archives of Nethys, which is great about tracking sources, the PF1 Magic Warrior magus archetype was first published in Pathfinder Campaign Setting: Inner Sea Intrigue in May 2016 and later republished in the hardcover Pathfinder Roleplaying Game: Adventurer's Guide, released a year later. The archetype was ported to PF2 in the Pathfinder Lost Omens: World Guide, released in July 2022.

Overview

Heroes aren’t born, they’re forged: forced by extraordinary circumstances to rise above and shaped by the world they inhabit.

This book is a guide to that world. In its pages, you’ll learn about the heart of the Pathfinder setting, known as the Inner Sea region, from the towns and cities your heroes call home to the wild frontiers and monstrous lairs where they’ll risk everything in search of fortune and glory. The information presented here provides players with a sense of the setting as a whole and is perfect for both fleshing out existing character concepts and inspiring you when you’re not sure what to play. Is your wizard a necromancer from haunted Ustalav, raised to battle the undead armies of the Whispering Tyrant, or a young student at the Arcanamirium in Absalom, the City at the Center of the World? Perhaps you’re a daring priest from the burning deserts of Qadira, blasting manticores and genies with the blazing light of the sun goddess Sarenrae, or a troll-hunting warrior from the icy Lands of the Linnorm Kings. Regardless, here you’ll find the rich details you need to give your character the perfect backstory, as well as new backgrounds, gods, feats, spells, and other rules to help you customize your ideal hero.

Paizo offers new character options in their Golarion sourcebooks, perhaps for more immersion than just reading about the regions and perhaps to lure non-GM players to buy the books. The Magic Warriors were more representative of the entire Mwangi Expanse than other possible archetypes, because they traveled the entire region. In contrast, the Pathfinder Lost Omens: Character Guide, released two years earlier in October 2019, offered the Magaambyan Attendent and Halcyon Speaker archetypes, which were tightly focused on the Magaambya Academy.

I am fairly sure that the developer's intent for the Magic Warrior was to illustrate the Mwangi Expanse rather than add another class archetype to magus. The PF1 developers David N. Ross and Ross Byers had a choice of making a Magic Warrior its own class (too important for a regional tradition), a prestige class (Pathfinder Campaign Setting: Inner Sea Intrigue did that with Taldor's Lion Blade), or a class archetype. They chose class archetype and the magus class was easiest to adapt to the concept.

As for my intent, I added classroom courses to my Strength of Thousands campaign, because my players wanted to seriously roleplay as students, I invented a course for every skill. For the first semester, Spring 4721 AR, I had the inventory manager Xhokan teach inventory security for the Thievery class, but really I needed a rogue to teach future Thievery classes in later semesters. So I made a new teacher, Noor Khan, based on real-life British resistance agent Noor Inayat Khan. Most teachers at the Magaambya Academy have a half-page description but no stat block. I found that the courses let me add teacher-supervised field trips to the campaign, which required a stat block for the teacher. Winter semester was time to give Noor Khan her stat block and the Magic Warrior archetype seemed a flavorful way to add magic to an Emerald Boughs (i.e., Magaambyan spy) rogue.

I also like to study game design, and an exercise such as repairing a broken archetype is practical experience in game design. Teridax's thread Multiclass dedication feats could be improved in a number of ways brings up notions that can be explored in a practical exercise. Ryangwy's objections to my latest ideas reveal the difficulty of balancing a gish class ("gish" refers to some early D&D foes who had both martial and spellcasting abilities, and became the nickname for martial-spellcaster hybrids: What A D&D "Gish" Character Build Is & How To Make One).

I don't like how you give so many proficiencies via feats, and in ways that don't match how they're given normally via archetype. It's very jarring, looking at a PF1e relic kind of feeling.

Usually I sit on a draft for a week to get better perspective and tone it down. But I am in a hurry this time, because Noor Khan is already active in my games, with her five homebrew Magic Warrior feats. I have a plot worked out in which the 7th-level player characters will team up with Noor Khan and 8th-level visiting scholar Kassi Aziril against trouble that followed them to Nantambu. It will happen in next week's game session or the week after.

Mathmuse wrote:

Mask Armor Feat 4

Your mask becomes armor. You are trained in light armor, medium armor, and unarmored defense. When you cast Don the Mask, the mask can transform to clothe you in explorer's clothing, leather armor, quilted armor, or hide armor for as long as you wear your mask. Removing the mask is still an Interact action, but casting Don the Mask again can make the armor disappear while masked. You may attach armor runes to the mask that become active when its armor appears.

Extremely awkward, nonscaling proficiency in armour when there's no reason why wearing a mask makes you comfortable wearing medium armour in the first place. Make it a bespoke armour that you can choose between the stats of leather armour or hide armour for but that uses your unarmoured proficiency, and maybe give it a free rainment rune.

That would be more magically elegant. I could scale it off the Mystic Armor spell. On the other hand, the mask does give a way to use actual armor runes instead of mimicking them with a scaling spell.

Nevertheless, Champion archetype gives light and medium armor proficiency, so gaining armor proficiencies from an archetype has happened before. And the PF1 magus Magic Warrior had light and medium armor proficiency, so I copied that.

Mathmuse wrote:

Mask Transformation [one-action] Focus 2

Uncommon Manipulate Polymorph
Based on Magic Warrior Transformation, World Guide pg. 95
Requirements You are wearing your Magic Warrior mask.
Duration 1 minute
You transform into the animal from your mask. You transform into a battle form copying your mask's creature, following the polymorph rules. You gain the features and attributes in the creature's Animal Companion entry, except all the attributes are increased by 1, you keep your original hit points, and you do not gain the creature's Minion trait, Support Benefit, nor Advanced Maneuver. ...

Is there a reason why you went against the standard polymorph template (especially when animal form is right there) for this instead? It looks pointlessly messy when you could just make it untamed form.

Yes, I have a reason. The reason is similar to the difference between casting a Summon Animal spell and adopting an Animal Companion. The Summon Animal, Animal Form, and Untamed Form spells are a one-shot appearance of a different animal each time. Animal Companion and a Magic Warrior mask are about keeping to the same animal or animal form.

Furthermore, while Summon Animal gives lots of detail about the animals by referencing Monster Core entries, Summon Animal uses the same template for each animal except for the difference in their speeds and unarmed attacks. Animal Companions offer more individualism in the animals, making the choice of mask more significant.

Mathmuse wrote:

Magic Warrior Expertise Feat 6 ...

The last similar one got purged in fire, and honestly I really don't see why you need this if your primary combat is through a polymorph spell (you know... a regular one). Let the martials be martials and the casters be casters.

Is a Magic Warrior magic (caster) or warrior (martial)? The Ten Magic Warriors were apprentice wizards, except that Azure Leopard was known as a defender, Ibex was known as a healer, Shifting Frog was known as a ranger-like elf, Verdant Spider was known as a baker, and White Bull was known as an architect. And the archetype outright states, "You mix magic and martial prowess, following in the tradition of the Ten Magic Warriors of Old-Mage Jatembe."

A martial, or a gish who is half martial, needs their weapon proficiency to increase to keep up with the increasing AC of opponents. If I wanted the Magic Warrior to be entirely a shifter, relying totally on transformed unarmed attacks for melee combat, then I could have their morphed or polymorphed form increase in attack bonus instead, but as far as I know the original Magic Warriors were not shifters. I presume that some Pathfinder 1st Edition developer had the cute idea that a character who always wore an animal mask should embrace aspects of that mask, and a PF2 developer added a full mask transformation, but I have to handwave those ideas as more recent Magic Warriors discovered that they could do more magic with their masks than the originals could. I don't want their combat to always be through a polymorph spell. Maybe I should drop Mask Transformation.

Okay, a weakness in my design is that I am applying band-aids to an archetype developed in separate visions rather than forged as a unified concept. What unified vision should I have for Magic Warrior that does not make the character rely on polymorphing?

Ryangwy wrote:

Oh! New idea! What if Magic Warrior was a 12th level archetype which doesn't lockout Magaambyaan Attendant/Halycon Speaker, and that can be taken as a mythic destiny like Mortal Herald? That lets you put a lot more power into the mask if it's that high level, better matches the story of the Ten, and allows you to gate truly legendary stuff behind mythic point use.

This is a great idea. The Magic Warriors are legendary figures and there has been very few of them. Mythic would definitely fit.

Magic Warrior archetype is not directly about the legendary Ten Magic Warriors. It is about newer heroes of the Mwangi Expanse inspired to follow in the footsteps of the original Magic Warriors. Requiring these inspired heroes to have earned their own prowess before donning a mask at 12th level to re-earn their reputation in a new identity feels like the wrong story.

I own a copy of Lost Omens Legends. Just because someone is a legend does not make them mythic. Most of the legendary people in the book are contemporary rulers of nations, such as Queen Abrogail Thrune II, ruler of Cheliax, and Nankou, Linnorm King of Icemark. A few figures in the book, such as Baba Yaga, are probably mythic, but most aren't. The Licktoad Goblins from We Be Goblins are in Legends. Some renown Legends figures have non-mythic stat blocks in adventure paths. For example, Kevoth-Kul, Black Sovereign of Numeria, is a male human barbarian 15 in Palace of Fallen Stars whose only amazing feature is that Numerian fluids have rendered him ageless. General Azaersi of Oprak is a female hobgoblin swashbuckler 20 in Vault of the Onyx Citadel.

Here is the backstory and build of Magaambyan teacher Noor Khan.

Vudrani scout Noor Khan and several other Vudrani volunteers sailed to Vidrian to assist in the Vidric Revolution. Arriving during the Free Captains’ retaliatory assault in 4717 AR, their ship participated in the naval battles. Alas, this gained them little recognition among the people of the new nation, many who looked upon them as outsiders. Most volunteers returned home. Noor, instead, traveled north to the Magaambya Academy to study medicine.

She returned to Vidrian as a physician and alchemist. Wearing her damibwaa mask from the Magaambya made her more accepted by the Vidric people, so that became her routine. She also found a growing magical connection between herself and her mask, granting her the abiltiy to cast arcane cantrips. Upon further study at the Magambya, she found tales of the the magic that developed in devotees of the Ten Magic Warriors. She joined the Emerald Boughs branch of the Magaambya, which focused on studying culture. Her research took her to other nations, where she adopted a dual identity: unmasked as an apolitical medical researcher and masked as Swimming Dog who rescued the downtrodden. She crafted a second mask of human visage for remaining anonymous.

She is currently on hiatus from field work and teaching at the Magaambya due to pregnancy.

Noor Inayat Khan Rogue 10
Human female sociologist
Ancestry Human, Heritage Skilled
Background Spotter
Racket Eldritch Trickster (Magic Warrior)
Perception +19
Languages Elvish, Garundi, Kelish, Mwangi, Osiriani, Taldoran, Vudrani
Skills Athletics +13, Crafting +16, Deception(m) +20, Diplomacy(e) +18, Emerald Boughs Lore +16, Intimidation +16, Magic Warrior Lore +16, Medicine(m) +19, Nature +15, Performance +16, Scouting Lore +16, Society(m) +20, Stealth(m) +20, Survival +15, Thievery(e) +18
Str +1, Dex +4, Con +1, Int +4, Wis +3, Cha +4
Items +1 ghost-touch striking shortsword, +1 rainment resilient leather armor, alchemist's toolkit, healer's toolkit (expanded), thieves' toolkit (infiltrator), Magic Warrior damibwa mask, Magic Warrior human mask, two moderate antidotes, two lesser elixirs of life, scroll of fireball, scroll of aerial form, 30 gp
Rogue Features Sneak attack 2d6, surprise attack, deny advantage, weapon tricks, debilitating strike
Class Feats
Eldritch Trickster Racket: Magic Warrior Dedication
Natural Ambition: Nimble Dodge
2nd: Magical Trickster
4th: Magic Warrior Arcana
6th: Basic Magic Warrior Spellcasting
8th: Mask Transformation
10th: Ancestry Mask
Ancestry Feats Courteous Comeback, Natural Ambition (Nimble Dodge), Sense Allies
General Feats Breath Control, Feather Step
Skill Feats Alchemical Crafting, Backup Diguise, Battle Medicine, Continual Recovery, Glad-Hand, Medical Researcher, Multilingual, Slippery Secrets, Swift Sneak, Terrain Stalker (underbrush), Underwater Marauder
AC 28; Fort +15 resilience, Ref +20 evasive, Will +17
HP 98
Speed 25 feet
Melee [one-action] +1 ghost-touch striking shortsword +19 [+15/+11] (agile, finesse, versatile S), Damage 2d6+3 piercing
Arcane Prepared Spells DC 26, attack +16
3rd dive and breach
2nd helpful wood spirits
1st admonishing ray, ant haul
Cantrip (5th) needle darts, spout
Magic Warrior Focus Spells (2 focus points) don the mask, mask transformation

"Swimming Dog" Animal Mask
Based on Wolf, Player Core pg. 210
Size Small
Melee [one-action] jaws +19 [+14/+9] (finesse), Damage 1d8+4 piercing
Str +4, Dex +5, Con +3, Int -2, Wis +2, Cha +2 (increases added already)
AC 28; Fort +17 resilience, Ref +21 evasive, Will +17
Temporary Hit Points 6
Skill Survival +14
Senses low-light vision, scent (imprecise, 30 feet)
Speed 30 feet, Swim 10 feet

"Goro Takei" Ancestry Mask
Ancestry human; Heritage versatile(Fleet); Language Tien

Magic Warrior Dedication Feat 2
Archetype, Dedication
Based on Magic Warrior Dedication, Lost Omens World Guide pg. 95, and Vigilante Dedication, Player Core 2 pg. 218
Archetype Magic Warrior
Prerequisites Trained in Arcana or Nature, trained in unarmed attacks.
Access You are from the Mwangi Expanse or attended the Magaambya Academy
You have donned a magical animal mask and learned the way of the magic warriors. You become trained in Magic Warrior Lore.

You make a Magic Warrior mask representing an animal of your choice from the Animal Companion list or similar to one of those animals, such as an Ibex instead of a Goat. You become trained in your choice of Acrobatics, Athletics or an additional skill listed on the selected animal companion. If you are already trained in all those skills, you may train another skill instead.

You have only one Magic Warrior mask, unless a feat grants you another one. You also name the mask based on the animal selected, such as Azure Leopard or White Bull. You cannot change the creature or name of the mask. If your mask is lost or destroyed, you can craft a replacement with the same appearance in one day of downtime. The mask is designed so that you can eat, drink, and sleep while masked.

You gain the focus spell Don the Mask. You Refocus by wearing your mask. If the mask is absorbed in a polymorph, that still counts as wearing the mask. Putting on the mask with an Interact action instead of Don the Mask does not changed your identity.

Special You cannot select another dedication feat until you have gained two other feats from the magic warrior archetype.

Don the Mask [one-action] Focus 1
Uncommon, Mental
Requirement You are holding, carrying, or wearing your Magic Warrior mask.
You mask your identity and become a mysterious figure. You Interact to put on your Magic Warrior mask. If you are already wearing your mask, you may Step instead. You assume an identity under the name of your mask. Your normal identity becomes obscured, except to people that you want to recognize you, as long as you do not remove your mask. That normal identity is not forgotten; rather, people remember it as another person. Effects that detect you based on your identity work only if you are currently in the identity the effect is trying to detect; otherwise, the effect fails as if the target didn't exist. No-one can discern the connection between your normal and Magic Warrior identities by observation, divination, or deduction from clues. Even people present as you put on your mask forget that your identity had changed right before their eyes. Likewise, if you remove this mask, people still do not connect your masked identity to your normal identity.

While in your Magic Warrior identity, you gain a +2 status bonus to saving throws against effects that try to detect you.

The Historic Ten Magic Warriors were Azure Leopard wearing cat mask, Black Heron wearing bird mask, Carmine Jaws wearing hyena mask, Elephant wearing elephant mask, Golden Snake wearing snake mask, Ibex wearing goat mask, Shifting Frog wearing a custom frog mask, Verdant Spider wearing custom spider mask, Whistling Kite wearing bird mask, and White Bull wearing custom bull mask.

I have no idea whether the original Ten Magic Warriors could use Mask Aspect or Mask Transformation, but here are the stats for the custom masks for Shifting Frog and Verdant Spider. White Bull's custom bull mask copies the rhinoceros animal companion.

Shifting Frog's Mask
Based on Capybara from Grand Bazaar pg. 54 and Animal Form from Player Core pg. 315
Size Small
Melee [one-action] jaws Damage 1d6 bludgeoning
Melee [one-action] tongue (reach 10 feet), Damage 1d4 bludgeoning
Str +2, Dex +2, Con +2, Int -4, Wis +2, Cha +0
Temporary Hit Points 6
Skill Survival
Senses low-light vision
Speed 25 feet, swim 25 feet

Verdant Spider's Mask
Based on Cave Gecko from Pathfinder #154: Siege of the Dinosaurs pg. 73 and Hunting Spider from Monster Core pg. 320
Size Small
Melee fangs Damage 1d4 piercing plus 1d4 poison
Str +2, Dex +3, Con +1, Int -4, Wis +2, Cha +0
Temporary Hit Points 6
Skill Stealth
Senses imprecise tremorsense 30 feet
Speed 25 feet, climb 25 feet

Magic Warrior Arcana Feat 4
Archetype
Archetype Magic Warrior
Prerequisites Magic Warrior Dedication, Trained in Arcana
You cast spells like the wizard Old-Mage Jatembe. You gain a spellbook with two common arcane cantrips and two common 1st-rank arcane spells of your choice. You gain the Cast a Spell activity. You’re trained in the spell attack modifier and spell DC statistics. Your key spellcasting attribute for Magic Warrior archetype spells is Intelligence. While wearing your Magic Warrior mask, you can prepare one cantrip and one 1st-rank spell each day from your spellbook. Also while wearing your Magic Warrior mask, you can use arcane scrolls and wands. You can add further spells to your spellbook through the Learn a Spell Activity

Mask Armor Feat 4
Your mask becomes armor. You are trained in light armor, medium armor, and unarmored defense. When you cast Don the Mask, the mask can transform to clothe you in explorer's clothing, leather armor, quilted armor, or hide armor for as long as you wear your mask. Removing the mask is still an Interact action, but casting Don the Mask again can make the armor disappear while masked. You may attach armor runes to the mask that become active when its armor appears.

Mask Aspect Feat 4
Archetype
Based on Magic Warrior Aspect, World Guide pg. 95
Archetype Magic Warrior
Prerequisites Magic Warrior Dedication
You can alter your form to gain an aspect of the animal your mask represents. You gain the mask aspect focus spell.

Mask Aspect [one-action] Focus 2
Uncommon Manipulate Morph
Based on Magic Warrior Aspect, World Guide pg. 95
Requirements You are wearing your Magic Warrior mask.
Duration 1 minute
You call upon the aspect of the animal from your mask, morphing to gain some physical features reminiscent of that animal. You gain the senses and the unarmed attacks of your mask's creature. You can Dismiss this spell.
Heightened (3rd) You also gain all the the Speeds of the mask's creature.
Heightened (5th) The duration increases to 10 minutes. You may automatically spend a focus point at the end of the duration to continue it for an additional 10 minutes indefinitely, even if unconscious.

Mask Transformation Feat 4
Archetype
Based on Magic Warrior Transformation, World Guide pg. 95
Archetype Magic Warrior
Prerequisites Magic Warrior Dedication
You can transform fully into the animal your mask represents. You gain the Mask Transformation focus spell.

Mask Transformation [one-action] Focus 2
Uncommon Manipulate Polymorph
Based on Magic Warrior Transformation, World Guide pg. 95
Requirements You are wearing your Magic Warrior mask.
Duration 1 minute
You transform into the animal from your mask. You transform into a battle form copying your mask's creature, following the polymorph rules. You gain the features and attributes in the creature's Animal Companion entry, except all the attributes are increased by 1, you keep your original hit points, and you do not gain the creature's Minion trait, Support Benefit, nor Advanced Maneuver. You gain the animal companion entry's initial hit points as temporary hit points. If your form has the mount trait, you do not lose an action when a character rides you. Handwraps of Mighty Blows still affect your unarmed attacks. You can Dismiss this spell.
Heightened (4th) Your animal form's Strength, Dexterity, Intelligence, and Charisma increase by 1. This can increase a +4 into a +5.
Heightened (5th) The duration increases to 10 minutes. You may automatically spend a focus point at the end of the duration to continue it for an additional 10 minutes indefinitely, even if unconscious. This automatic extension renews the temporary hit points.

Basic Magic Warrior Spellcasting Feat 6
Archetype
Based on Basic Wizard Spellcasting, Player Core pg. 223
Archetype Magic Warrior
Prerequisites Magic Warrior Arcana
You gain the basic spellcasting benefits and you can prepare two cantrips each day instead of just one. You no longer need to wear your Magic Warrior mask while preparing your spells. The 1st-rank spell from basic spellcasting benefits is in addition to the 1st-rank spell from Magic Warrior Arcana. You gain your 2nd-rank spell at 8th level and your 3rd-rank spell at 10th level. At 6th, 8th, and 10th level, add two common spells of rank up to your highest Magic Warrior spellcasting rank to your spellbook.

Magic Warrior Expertise Feat 6
Archetype
Archetype Magic Warrior
Prerequisites Magic Warrior Dedication
You've dedicated yourself to the warrior side of Magic Warriors. Your proficiency ranks for weapons and unarmed attacks in which you are trained increase to expert.

Masked Anonymity Feat 6
Archetype
Based on Nameless Anonymity, World Guide pg. 95
Archetype Magic Warrior
Prerequisites Magic Warrior Dedication
Your mask protects you further from divination. If you cast Don the Mask during your daily preparations, you also gain the effect of Veil of Privacy for as long as you wear your mask.

Ancestry Mask Feat 8
Archetype
Archetype Magic Warrior
Prerequisites Magic Warrior Dedication
Another Magic Warrior mask lets you present a different face. You make another Magic Warrior mask representing a common or uncommon ancestry or a rare ancestry from the Mwangi Expanse (Anadi, Conrasu, or Goloma) to gain yet another name and identity. Choose a language that that ancestry typically speaks and a Heritage of that ancestry. You can speak that language while wearing the ancestry mask. When you cast Mask Aspect while wearing the ancestry mask, you gain the features of that ancestry and that heritage and the mask is absorbed to change your appearance. Mask Transformation while wearing the ancestry mask is the same as Mask Aspect and is a morph instead of a polymorph. You can have only one ancestry mask and cannot change its associated identity.

Mask Kinspeech Feat 10
Archetype
Based on Kinspeech familar ability Player Core pg. 212
Archetype Magic Warrior
Prerequisites Magic Warrior Dedication, trained in Nature
Your mask is in harmony with animals. While wearing your Magic Warrior animal mask, you can understand and speak with animals of the species represented by the mask. You can also cast Speak with Animals once per day while wearing your Magic Warrior animal mask.

Cosmic Mask Feat 12
Archetype
Based on Kindled Magic said on page 43, "Animals are the most typical shapes for Magaambyan masks, but other shapes (particularly celestial objects or supernatural beasts) aren’t uncommon."
Archetype Magic Warrior
Prerequisites Magic Warrior Dedication
Your mask taps into the cosmos mystery. Decorate your Magic Warrior Mask (both masks if you have Ancestry Mask) with images of the sun, moon, and stars. Learn one of the following focus spells spray of stars, interstellar void, and moonlight bridge. You can cast it while polymorphed with Mask Transformation.
Special You can take this feat up to three times, each time selecting a different focus spell.

Expert Magic Warrior Spellcasting Feat 12
Archetype
Based on Expert Wizard Spellcasting, Player Core pg. 223
Archetype Magic Warrior
Prerequisites Basic Magic Warrior Spellcasting, expert in Arcana
You gain the expert spellcasting benefits.

Magic Warrior Mastery Feat 14
Archetype
Archetype Magic Warrior
Prerequisites Magic Warrior Expertise
You've mastered the combat techniques of Magic Warriors. Your proficiency ranks for weapons and unarmed attacks in which you are expert increase to master.

Follow the Founder Feat 16
Auditory Concentrate Downtime
Based on Follow the Expert, Player Core pg. 438
Five of the Ten Magic Warriors founded the Magaambya and you can lead in their example. Any ally making a downtime skill check with a skill where you are Expert or better and in regular communication with you, gains a circumstance bonus to the skill based on your proficiency (+2 for expert, +3 for master, and +4 for legendary).

Master Magic Warrior Spellcasting Feat 18
Expert Magic Warrior Spellcasting Feat 12 (Spellcasting Line)
Archetype
Based on Master Wizard Spellcasting, Player Core pg. 223
Archetype Magic Warrior
Prerequisites Expert Magic Warrior Spellcasting, master in Arcana
You gain the master spellcasting benefits.

Legendary Mask Feat 20
You are inspired by the Ten Magic Warriors. You learn one of the following mask-related feats: Cunning Trickster Mask, Emancipator's Mask, Grand Medic's Mask, Protective Spirit Mask, Sky Master Mask, Stalking Feline Mask, Storyteller's Mask, Thick Hide Mask, Tireless Guide's Mask, Vigilant Mask.

Weeks ago I grew busy and put this discussion aside, but now this thread is active again, I wrote down my thoughts for a rewritten Magic Warrior archetype. The work took two full days.

The major flaw in the PF2 Magic Warrior archetype is that it is almost entirely direct conversion of the PF1 Magic Warrior class archetype for magus.
PF1 Nameless Mask (Ex) became PF2 Magic Warrior Dedication feat 2.
PF1 Magic Warrior’s Aspect (Su), replacing the 3rd-level magus arcana, became PF2 Magic Warrior Aspect feat 4.
PF1 Nameless Anonymity (Su), delaying 8th-level improved spell combat to 14th level, became Nameless Anonymity feat 6.
PF1 Magaambya Spell Access (Su), replacing 19th-level greater spell access, was not ported.
And PF2 Magic Warrior Transformation feat 4 is new.
The Magic Warrior in PF1 relied on the spells and combat prowess of the Magus class to be a magical warrior. The Magic Warrior in PF2 has no options to provide spellcasting or martial proficiency if a player starts with a class missing either aspect. A Magic Warrior fighter would have no magic except for at most two focus spells, and not fit the "Magic" part of "Magic Warrior." A Magic Warrior sorcerer would be untrained in martial weapons, never expert in simple weapons, and untrained in all armor, and nit fit the "Warrior" part of "Magic Warrior."

Thus, the archetype needs feats that can provide the missing proficiencies. Magic Warrior Aspect could provide unarmed attacks that are as good as martial weapons. A new feat called Mask Armor would train in light and medium armor. And the archetype would have basic, expert, and master spellcasting feats. An archetype suitable as a free archetype needs 10 feats, one to take at each even level, but since some Magic Warriors will skip the feats that improve martial abilities and others will skip the feats that grant spellcasting, I needed to increase the Magic Warrior archetype from 4 feats to 14 feats. I went a little further and increased it to 16 feats.

A complicating factor is that I have a teacher at the Magaambya Academy whom I want to be a Magic Warrior. Noor Khan gives her lectures maskless. She will sometimes wear her mask for ceremonial occasions while still known by her maskless identity. Thus, the mask cannot be a constant identity. Instead, I will make a focus spell called Don the Mask that conceals the character's original identity to embrace a masked identity.

Have subsequent feats let you choose better martial training or spellcasting benefits, including flavorful options for combining both.

Include the mask feats as capstone options at 20th level, e.g. Cunning Trickster Mask, Sky Master Mask, Tireless Guide's Mask, and so on.

Thanks for the suggestion. I had skimmed the adventure in the 6th module, [i]Shadows of the Ancients,[/url] but I had not glanced at the Adventure Toolbox section that included those ten feats. They make a solid capstone to the Magic Warrior archetype. Without the Magic Warrior archetype they are weirdly out of place.

Arguably the least of the archetype's worries, but "Nameless Anonymity" is quite possibly the worst name ever conceived for a feat in Pathfinder. Anonymous literally means "nameless". It's like calling a feat "Quiet Silence" or "Powerful Power".

And the Ten Magic Warriors were not nameless. They had names such as Shifting Frog. I will change the name to Masked Anonymity.

Thanks to everyone for their help.

Mathmuse wrote:

Strength of Thousands never mentions the Magaambyan Attendent, Halcyon Speaker, and Magic Warrior archetypes.

Running Strength of Thousands, I should note the first two are very much untrue - SoT doesn't just mention them, they give them as free feats via the academic system! Assuming you're using the FA Druid/Wizard concept, I'd say you can take feats from those two archetypes as FA feats (once you reach the appropriate level) but even if that's not the case, getting the feats for free isn't nothing.

Ah, I had dropped the Life in the Academy system, so I had forgotten that it offered Magaambyan Attendent Dedication at the 3rd step on the Branch Benefits by Level ladder and some other Magaambyan Attendent feats later.

I don't see the Halcyon Speaker archetype there, though the last step in the ladder mentions halcyon spells. However, the teacher Mafika Ayuwari is a halcyon speaker in his stat block. On the other hand, that stat block is from his appearance in Fists of the Ruby Phoenix adventure path rather than Strength of Thousands adventure path.

I am using free archetype in my campaign, but I let my players take other magical archetypes besides druid or wizard. The two bards both archetyped to druid, the kineticist archetyped to wizard, the champion archetyped to magus, the wizard Idris archetyped to Magaambyan Attendent (which is why he has a mask familiar), the magus archetyped to Dragon Disciple, and the rogue with Eldritch Trickster (Sorcerer) racket archetyped to Gelid Shard to run two archetypes.

I'm wondering if one way to handle the Magic Warrior Archetype would be to have the dedication give the character the ability to make a special magic item (irrespective of crafting proficiency, etc) that would cost about half the cost of a spell heart, and would 'grant' certain abilities/activation. (cantrip and once per day spell perhaps) They have to invest the item, and they stick to the anathema/edicts about secrecy of identity, and it loses magic and would have to be re-crafted.

So the archetype might have room for a bit more power than an average archetype, because it has a GP and Investment cost to it that many archetypes wouldn't (in addition to the edict/anathema aspects)

I also wonder does the anathema of not taking off the mask count if you aren't in your Adventurer/Hero garb and around people you don't know so no-one would associate 'you' with the 'masked one'. Can you not wear your mask when you are wearing a disguise? (it seems really hard to 'successfully' disguise yourself as a random peasant wearing a big wooden animal mask in most cases, for instance) Or was it supposed to be more like the extremist interpretation of the Mandalorian never take your mask off in front of someone.

I think these are great ideas because they are what I have been planning, except that I am going to greatly diminish the importance of wearing the mask all the time. The magic item will be the mask itself, which can have extra magical powers added to it.

I have a very strong habit if analyzing game balance to an extreme. I played in a few PF1 campaigns and ran three, so I am familiar with PF1 class archetypes. Each archetype features either substitutes for a class feature or delays a class feature. For example, the PF1 Magic Warrior’s Aspect says, "This ability replaces the magus arcana gained at 3rd level."

The exceptions are archetype features that are very weak or that balance an ability with a liability. The 1st feature of the PF1 Magic Warrior archetype for magus does not swap out an ability:

Nameless Mask (Ex): A Magic Warrior gains strength by hiding her identity behind a mask that represents an animal, but also suffers drawbacks from doing so. While wearing her mask, a Magic Warrior gains a +2 bonus on saving throws against divinations in the scrying subschool. However, NPCs who are unfamiliar with or have a poor opinion of the Magic Warrior tradition have a starting attitude of unfriendly toward the Magic Warrior while she wears her mask. If using the influence system (Pathfinder RPG Ultimate Intrigue 102), she requires one additional success to increase her influence level over such an NPC.

The Magic Warrior cannot ever reveal her name or remove her mask in front of anyone who she has never met before. If she violates this important taboo, she must discard her mask and create or obtain a new mask and wear it before she can gain any benefits that result from wearing her mask.

Thus, the PF1 Magic Warrior balances the +2 bonus against divination against the awkwardness of having to wear the mask all the time and have strangers suspicious about that masked adventurer, rather than swapping some 1st-level feature for the bonus. Ironically, that potential suspicion would not be a disadvantage in the 1st two modules of Strength of Thousands, because everyone is familiar with the Magic Warrior tradition.

In contrast, a PF2 archetype dedication costs a class feat, so it does not need a balancing liability.

And for my roleplaying needs, I need the teacher to be able to remove her mask rather than always flaunting that she is a Magic Warrior. She was seen without a mask before I had the idea of her taking the Magic Warrior archetype. But she also is an Emerald Boughs spy, so having a secret identity under the mask makes sense. Noor Khan needs the ability to walk as herself without a mask, wear her Magaambyan mask as herself for ceremonial purposes at the Magaambya, and slip into a secret identity as the masked Swimming Dog for secret agent missions (her mask is represents a damibwa, also called diving dog). Magic Warrior is going to be a lot like a Vigilante that way.

Here is a scene involving Noor Khan that I wrote up for an absent player. Noor was teaching a class called, "Contraceptive Magic and Other Health Concerns."

I should have finished altering the Magic Warrior archetype before roleplaying that scene, but I was busy with other projects.

Please discuss fixing the Magic Warrior archetype in my Magic Warrior thread rather than here. I copied Loreguard's suggestion over to that thread: Has Anyone Tried Magic Warrior Archetype, comment #31.

I mentioned my Magic Warrior difficulty in the thread Multiclass dedication feats could be improved in a number of ways, and some people had suggestions. I hope to move the discussion back to this thread.

For example, Loreguard said,

Here is a thought for your issue with Magic Warrior

Magic Warrior Dedication 2
Prerequisites ability to cast focus spells
+ Trained in Magic Warrior Lore; Arcana or Nature
Access You are from the Mwangi Expanse.

Grants:
Magic Warrior Lore becomes Expert (this is really very minor boost for flavor, and should be negligible power cost)

Pick your mask and associated creature.

Gain basic Spellcasting abilities, including Add one Variant Cantrip from the choice below:

Tame (variant is instead of being limited to any domesticated creatures is limited to creatures of the type selected by the mask)
Gouging Claw (variant causes damage is per spell save the damage type is per the primary damage of the animal aspect chosen)

The Cantrip take the Tradition of your focus spellcasting from your prerequisite.

Not sure how one would translate the Divination bonus. Maybe an ability like:
Reaction
Trigger: a spell or magic ability is invoked that would reveal your location, name or information about your abilities.
Effect: Apply a +1 status bonus to the DC required for the spell, or +1 to any save to prevent its success.

Let me talk about the prerequisites, because that was a topic under discussion in the other thread. The proposal was that Multiclass Archetypes should have skills are prerequisites rather than attributes as prerequisites. If an archetype needs a skill, instead of training in the skill, it could require the skill as a prerequisite.

The PF2 Magic Warrior archetype as written before the Remaster is terrible about focus pools. It wants the Magic Warrior to gain its focus points to power its Magic Warriro focus spells from its base class. And some classes, such as Monk and Ranger, gain their focus points from class feats, which leaves fewer feats for archetype feats. Other classes never gain focus spells.

Fortunately, the Remaster fixed that problem. It simplified the focus pool so that each focus spell learned increases the size of the focus pool, up to a maximum of 3 focus points. Thus, the two focus spells that the Magic Warrior archetype gives, magic warrior aspect and magic warrior transformation now provide their own focus points. We no longer need the prerequisite, "ability to cast focus spells." Starting with a focus pool from the base class would be handy, but no longer necessary.

The new prerequisite of "Trained in Arcana or Nature" is fine. The Magic Warrior has expressed a pre-existing inclination toward arcane or primal magic.

The new prerequisite of Magic Warrior Lore and the granted ability of expert proficiency in Magic Warrior Lore is sad. The pre-existing inclination expressed by Magic Warrior Lore feels like a narrow childhood dream, "I want to be a Magic Warrior like grandpa." Unfortunately, the setting in Strength of Thousands does not support that dream, because no NPCs are Magic Warriors. More importantly, Magic Warrior Lore has no practical uses; instead, it was so that the character can recite the illustrious history of the Ten Magic Warriors for flavor. Expert proficiency adds nothing useful, either. Having a useless lore as a prerequisite is a skill tax.

1. Roll class DC training into the multiclass trait
Class DC is weird. Most spells requires saving throws against their effects, so spellcasters have spellcasting DCs for those saves. But martials have almost no abilities that require enemies to make saving throws. In the rare case that they do, they have a class DC. For example, an Animal Instinct barbarian with Spider as their chosen animal has the save to escape from their web versus barbarian class DC. The developers avoid using a skill DC for this purpose, because skill proficiency can vary too much.

The remaster merged the spellcasting DCs of all four magic traditions into one all-purpose spellcasting DC. I think that likewise, all classes should be at least trained in an all-purpose character DC that serves the current purpose of class DC. Some classes will raise the proficiency in character DC faster than other classes, but that would be okay with me.

2. Skill prerequisites, not attribute prerequisites or trained skills
An attribute prerequisite demonstrates inborn talent, but a skill prerequisite demonstrated previous interest. Previous interest seems better for character development.

In addition, qualifying for attribute prerequisites requires a lot more pre-planning than skill or skill feat prerequisites. Suppose a 1st-level party formed and the player had not coordinated and they had no healer. Multiclassing to cleric won't help much, since the cleric multiclass archetype does not offer a divine font and the character would not gain the Heal spell before 4th level, but it still has the hefty "Wisdom +2" prerequisite. A character cannot increase attributes until 5th level and that is only a +1. Champion multiclass would enable Lay on Hands at 4th level, but it has a tougher prerequisite, "Charisma +2; Strength +2." A character with Str +2 probably is already training in light and medium armor, so the Champion Multiclass Dedication is mostly a feat tax. In comparison, Medic Archetype has Battle Medicine and trained in Healing as its prerequisites, so a character with no advance planning can increase Medicine Skill, learn Battle Medicine, and take Medic Dedication all at 2nd level, which instantly increases proficiency in Medicine to expert.

3. Initial spellcasting benefits
More standardization of initial spellcasting benefits works when classes are built to only three standards, but some spellcasting classes don't fit those standards. The Summoner Archetype grants spellcasting via a 6th-level feat. An animist's spells are divine, but their apparation spells are not necessarily divine, and the Animist Archetype folds the non-divine spells into the basic divine spellcasting.

I personally have a conundrum about basic spellcasting myself. I have been creating stat blocks for the teachers at the Magaambya Academy in my Strength of Thousands campaign, because I added educational field trips supervised by a teacher, including trips that fight will-o'-wisps and zombies. I wanted to give a teacher the Magic Warrior archetype, an archetype that honors the Ten Magic Warriors who founded the Magaambya Academy. Alas, the archetype was badly ported to Pathfinder 2nd Edition and is unworkable. It had been a magus class archetype in Pathfinder 1st Edition and the magus class granted the martial and spellcasting abilities that a Magic Warrior needs. The PF2 version is only the feats that altered a magus into a Magic Warrior without granting martial proficienicies or spellcasting capability. See my thread Has Anyone Tried the Magic Warrior Archetype?

Thus, I plan to rewrite the Magic Warrior archetype to add feats to gain routine martial combat and basic spellcasting. But the Magic Warrior Dedication offers other features already so that adding two cantrips would make it overpowered. Therefore, the cantrips will have to come from a 4th-level feat. But two cantrips is too weak for a 4th-level feat. Should the feat offer three or four cantrips? Or maybe two cantrips and the standard Basic Spellcasting benefits? Having to wait to 6th level for a Magic Warrior's first 1st-level spell would be frustrating.

Reaction attacks are easy to set up in well built groups. We get them all the time.

The rogue has one of the easiest to activate reactions with Opportune Backstab.

I take Opportune Backstab on fighters and barbarians and nearly any martial since even at level 16 Opportune Backstab is great and easy to set up. I also take Gang up on almost all martials to make flanking easy so we don't have to position on each side of a creature.

So not sure why anyone thinks Reactions are hard to set off. In a coordinated group, reactions are very easy to set up and activate, especially Opportune Backstab which requires an ally to land a melee hit on the same target.

I have had zero problems getting Opportune Backstab to activate in battles. Not sure why anyone would consider it a difficult to use reaction unless they didn't know how to play the game or have much experience.

The rogue Roshan in my campaign has not taken Opportune Backstab, because Roshan is 7th level and Opportune Backstab is an 8th-level feat. She has taken Gang Up.

Deriven Firelion wrote:

a hasted rogue attacking multiple times and getting their reaction is the game working as intended and not some special set up.

I think you need to change that statement to

A hasted rogue attacking multiple times and getting their reaction is the game working IN ONE OF THE WAYS THAT IT IS intended.

Its very clear from your posts that your table works differently than a great many tables. Some of that difference seems to be that your table does a lot of optimization of both individual characters and of the group as a whole. There seem to be other differences but I haven't been able to figure out what they are.

And that is just fine. You presumably are all having fun and have found balance points that work for your hyper optimizing groups. ...

The player characters in my campaigns have faced combat as difficult as Deriven Firelion's characters have faced. Rather than optimizing for intense damage, they optimized for tactics. They will identify enemy strengths and then switch to a group tactic that nullifies that strength, reducing their opponent to a block of tofu (I like Inkfist's metaphor). Under this strategy, the versatility of a kineticist makes up for their low single-target damage.

For example, a recent battle was against the 8th-level tripkee crime boss Froglegs and her gang. The party was supposed to have this battle at 6th level, but I sent them after the final boss Salathiss early, so they had leveled up to 7th. The party gave themselves the extra challenge of capturing all gang members alive for a fair trial.

Froglegs had "Deny Advantage Froglegs isn't flat-footed to creatures of 8th level or lower that are hidden, undetected, flanking, or using surprise attack." Flanking, including Gang Up, did not work against her. Fortunately, the rogue Roshan was built differently than most rogues. Her player knew that a party of mostly spellcasters would seldom engage in melee to flank, so she had trained, experted, and mastered Athletics. She would typically Trip low-Reflex creatures and Grapple low-Fortitude creatures to render them off-guard to everybody, which was handy for those spellcasters and the archer magus making attacks from a distance. Thus, she Grappled Froglegs. Froglegs was slippery (pun about tripkees intended) and escaped the grapple, but that cost her an Escape action every turn.

With a seven-member party and a Chime-Ringer runesmith NPC who returned for this police arrest, the party can deal damage to a single target by massively outnumbering them. Their weakness is that individually they are squishy, so nerfing the single foe's counterattacks helps keep them healthy. Different tactics for different parties.

I was also amused earlier in the encounters to see Roshan get critical successes in Grappling an ordinary gang member. The other party members quicky slapped handcuffs and leg manacles on the restrained perpetrator to capture them without beating them unconscious. I had to add more Chime-Ringer police NPCs just to guard the captured gang members who were still awake. Watching unexpected clever solutions is why I gamemaster.

The issue with having a martial being buffed multiple times/ or enemies being debuffed multiple times before anyone else acts is while it is *a* solution to encounter, it's often not *the* best solution to encounters. For every time 'buff the VIP' works, there is often one where a sixth rank 'slow' or a 'chain lightning', or even something like a 'banishment' offers more impact or will speed up encounters faster than spending multiple turns buffing one party member.

What I learned from this thread is that a rogue is easier to buff than a kineticist. That is a significant advantage in a party that uses teamwork.

OceanshieldwolPF 2.5 wrote:

Deriven Firelion wrote:

What do you think of kineticist damage?

I don’t like it. Kineticists just don’t really do anything for me thematically nor do I enjoy the narrative they evoke. I tried to play a Chaokineticist (I think) once, and it just wasn’t very interesting.

Chaokineticist? Is that a 1e thing? I don't think anyone will argue with you for not liking a 1e class's narrative (Remember that 1e Kineticists were supposed to be invoking Firestarter, Carrie, and other Psychic Powers At A Cost fantasies? They really did not express that well at all)

It appears to be a Pathfinder 1st Edition kineticist with Void element: Kineticist Elements. I never saw a kineticist in my PF1 games. I think the class was unpopular.

The PF2 playtest kineticist Collin does not count as a choice, but my next two PF2 campaigns each had a player choosing kineticist. I count that as popular. The water-and-wood kineticist Monet was played by my wife. She is a grandmaster tactician, so battlefield control and narrative control matter more to her than damage. For example, at 5th level she added earth to her elements, and delayed the arrival of enemy reinforcements by summoning a stone cube to block a door.

Inkfist wrote:

It gets worse he also presumes to be getting haste and reaction damage in this scenario and got frustrated when it was suggested that enemies typically will punish squishy characters who conventiently start their turn next to them and who try to attack three+ times.

Haste aside, how weak are your parties that you aren't getting reaction attacks off constantly? Gang up, opportune backstab and preparation mean you should be getting 1 or 2 off per round with the rogue seeing as you don't need to worry about positioning and the constant +3 and +4 aids coming in to guarentee your melee partner hits, not to mention any of the other omnipresent resourceless buffs and debuffs available.

I think Deriven Firelion summed up the kineticist's damage well. My previous experience watching Collin and Monet in my games demonstrated them as great generalists rather than damage dealers. Fire kineticist Cara in Strength of Thousands has been trying to deal serious damage, but comes across more as a versatile damage dealer than a dealer of massive damage.

I don't know about reactions in Inkfist's party, but reaction attacks are non-existent in my Strength of Thousands party. The champion uses his reactions to defend the party. The rogue is running two archetypes, Gelid Shard and Sorcerer, that use up all her class and free archetype feats, so her only reaction feat is 1st-level Nimble Dodge. The kineticist Cara has no reaction abilities that I know of. And the starlit magus, two bards, and wizard keep their distance from enemies.

I have seen a water-and-air kineticst (playtest kineticist Collin), a water-and-wood kineticist (Monet), and a fire kineticist (Cara) in my campaigns. The first two had seriously helpful non-damage support, so no-one minded that their damage was weak. Weak damage on the fire kineticist was more surprising.

Actually, my observations and the numbers in the kineticist class are that the basic Elemental Blast deals as much damage as a bow in the hands of a martial that lacks abilities good for bow use, such as a ranger's Precision Edge. That is reasonable at range. But Cara is one of the tougher PCs in my Strength of Thousands campaign due to the spellcaster-friendly Magaambya School of Magic setting, so she often stands in front to protect the squishies. But Cara has no heavy-hitting option at melee range like Strength-based martials have.

She does have the option of a one-action blast followed by a two-action impulse that requires a save, such as Flying Flame, to avoid multiple attack penalties. She also sometimes consumes her kinetic gate with more damaging overflow impulse Blazing Wave and then gets a free blast in the action that restores her gate.

In conclusion, though Cara's individual-shot damage is lackluster, it is persistent with two reliable shots per turn.

Avoid Notice with Quiet Allies was very important in my PF2-converted Ironfang Invasion campaign. They spent a large part of the 1st module, Trail of the Hunted, stealthily scouting ahead in the Fangwood Forest to guide a band of refugees. They were also warning the known civilized residents of the forest about the invasion that had conquered their town. They did not want random encounters with forest monsters, so they traveled using Avoid Notice. And the 2nd-level rogue Binny had boosted her Stealth skill to expert and learned Quiet Allies to help avoid those random encounters.

The party said they were traveling with Avoid Notice following Binny's expertise, and one of the least-stealthy players--that is lowest Dexterity because they had all trained in Stealth--would roll with a +2 bonus from Binny's Stealth proficiency. If they rolled 15 or higher, I would skip inventing a random encounter.

The random encounter would have had the excitement of combat, but the planned encounter ahead would be more exciting, so skipping random encounters increased the excitement in the game session.

And when they reached the planned encounter, such as the Gristledown logging camp, they were not trying to avoid the place. Instead, they fought the monsters there as part of determining what had happened to the lumberjacks. Arriving in Avoid Notice mode meant that they could roll Stealth for initiative so that Binny and the other rogue Sam could get their Surprise Attack.

If you had a range weapon then you can just strike them from where you stand no problem and they would be offguard, The main benefit is that you are incredibly defended while being hidden with any effect targeting you only having half the chance to actually hit you compared to what it would be otherwise, Even moreso if you apply cover.

That was Binny's favorite tactic. She was a sniper. She would Hide behind a bush or a crate to catch opponents off-guard from her shortbow Strikes. After the shot, she would Hide again for defense on the opponents' turns. Binny's player was new and wary of risk, so she liked the extra defense.

But yes otherwise you would need to sneak or step to close the distance, but would only be able to remain hidden do if theres concealment near the enemy or you had a feat that allowed it. If you are a rogue you don't even care, Suprise Attack Feature has them offguard as long as you act before them so as long as any creature acts after you, you can just walk up to them without a care, and any who acts before you is going to either waste actions or take chances if they want to hit you.

I myself ruled--and this is a houserule rather than an interpretation--that if a character was hidden or undetected and took a successful Sneak action but had no cover or concealment at the end of the Sneak, then they were still hidden at the beginning of their next action that turn. It was more fun for the rogues.

10' squares will be very problematic for maps on Roll20. Sigh.

I had no problem with the maps in Roll20 myself. I put four 5' squares of the Roll20 grid into each 10' square on the map.

One of my players used to play on a tablet, which often slowed to a crawl whenever the Roll20 maps were bigger than 50 squares by 50 squares. So I split the map on page 40 into two maps, one for Level 1 and one for Level 2.

So, 40' wide stairs, and all of the bunks being 10' long and wide.

Does feel a wee bit strange.

@mathmuse pretty much everyone in the group has a composite longbow - so I sympathise with the possible close quarters issues, but with 20' high ceilings and communal airwalk being used, I will feel less bad using 5' squares (if I don't go back and redo the maps in Roll20).

@erucsbo With the exception of the mithral golem, the rest seem to have plenty of room - and the trolls are specifically called out as being in the courtyard, due to it being the only place that can hold them all. I can only find a single minotaur, Grimhorn, and he seems to fit comfortably in the archery area of the citadel - have I missed anything?

There is also a Huge adamantine golem in Room H33 on Level 3, and Zanathura herself is Large. Nevertheless, most of the foes are Medium hobgoblins.

My own PCs had gotten a map of the Onyx Citadel from Ziguch the Seventh Facet, so they knew they were heading to H35 via H33. Once the Ironfang Munitions Officers opened the windows in Room H28 to shoot the catapults at their hovercraft (I added that to the campaign), the party turned invisible and flew in through those windows. Thus, they skipped the first two levels of the Onyx Citadel, and I don't recall who was there.

Which is why you keyword any terms with specific meanings.

If you define what senseless means for your game, then senseless, every time it appears, will have one and only one meaning in terms of how it interacts with the game's rules.

Senseless was used in the definition of the keyword Stunned. That sounds like an infinite recursion in which the rules have to define keywords in order to define keywords.

A good example is Menace as an MtG keyword. I common parlance it's so broad as to be effectively meaningless as a game term. In MtG it means, "Can only be blocked by two or more creatures." Lifelink, it's not even a real word and could mean a few things. In game terms, it means, "When this creature deals damage you gain that much life."

I played Magic: The Gathering so long ago that I remember when Menace was not a keyword. Instead, the ability was written out. For example, the M12 version of Stormblood Berserker said, "Stormblood Berserker can't be blocked except by two or more creatures."

Furthermore, the original Magic: The Gathering rules were sloppy at defining their keywords. I remember banding, such as on Benalish Infantry. The keyword meant that the card could band with another attacking or blocking creature so that they dealt damage and received damage together. But eventually Wizards of the Coast had to formally define banding and then they dropped it because the well-defined ability was too lengthy to explain: Banding (Any creatures with banding, and up to one without, can attack in a band. Bands are blocked as a group. If any creatures with banding you control are blocking or being blocked by a creature, you divide that creature's combat damage, not its controller, among any of the creatures it's being blocked by or is blocking.)

I'm also going to be absolutely honest... and say that I absolutely despise the MTG rules. The comprehensive rules document is close to 300 pages. Most of which is just saying what triggered abilities does and cornercases it might create. Most of it feeling like its weird undefined interactions that has popped up in MTG tournaments and then been added to the document as a separate ruling.

Thats alot of work on a single document, but it also feels like they have to due to the competitive nature.

Originally, the Magic: The Gathering rules came in a tiny booklet small enough to be included in a box of cards and written in 6-point font. I tried to save those booklets as keepsakes, but they were so small I lost them. This is why the early keywords were sparsely defined: the booklets did not have the text space to define them precisely. Written in 10-point font on letter-sized paper, the rules would have been about three pages.

Magic: The Gathering needs short and precise keywords because each card's ability has to fit onto the card. Roleplaying games have much more space to define abilities. Instead, the importance of keywords in Pathfinder is that they serve as precise elements that can have well-defined outcomes in complicated battles.

However, keywords come at the cost of replacing narrative with mechanics. Instead of thinking, "Bruce the Brave swings his enchanted sword at the goblin snarling before him," the player might fall into thinking, "Bruce the Brave makes a +1 longsword Strike targeting the adjacent goblin." Pathfinder viewed this way is a boardgame like chess rather than a storytelling adventure.

Clear technical writing isn't that hard.

I used to write the documentation for the software that my team developed. Technical writing is hard. We mathematicians are good at making sure the documentation is complete, but perfect clarity is often beyond our scope. For example, some words have multiple meanings.

For example, we have the recent argument about the word "senseless." Words like "senseless" and "ruthless" often are not direct negations of their root words. Sense can mean a method of perception, but it can also mean thinking as in the phrase "common sense." Ruth means compassion or remorse, but we forgot that meaning so ruthless stands alone to mean without mercy. Wireless covers electronic communications that never used wires for communications, and a wireless communication device might be wired to a power source. Aimless means without purpose rather than not aiming. I happen to agree with NorrKnekten that the meaning of "senseless" is the same meaning as in the idiom "knocked senseless" which is unconscious or just short of unconscious but not blinded or deafened. But my main point is that the word is too ambiguous.

Senseless has a third meaning of lacking meaning, such as "a senseless rant." Fortunately, no-one has tried to apply that meaning to the introductory sentence of the Stunned condition.

Technical writing has to avoid such ambiguities. Which means looking at the text from a fresh viewpoint to catch alternative meanings rather than judging it from its intended meaning. That is difficult.

Are the squares in the Onyz Citadel really supposed to be 10 feet each?

All of the artwork is geared towards 5 foot squares.
The Vault keepers and builders were all medium sized.

I followed the 1 square = 10 feet, and the rooms felt very big. This gave my PCs an advantage, because they had specialized in ranged combat during the outdoor modules.

I was introduced to flavor text by Magic: The Gathering. For example, the Elfsworn Giant card has flavor text:

Elfsworn Giant

Mana Cost of 3 mana and 2 green mana
Picture of a giant in a forest.
Creature – Giant
Common
Reach (This creature can block creatures with flying.)
Landfall — Whenever a land you control enters, create a 1/1 green Elf Warrior creature token.
He upheld his vow to the elves by watching their woods and blazing their trails.
Power/Toughness 5/3

The card has three sections in italics. The first, "(This creature can block creatures with flying.)" is called reminder text. It reminds the player what the Reach ability does. The second, "Landfall," is an ability label. Landfall means that the ability that follows triggers when the player plays a land card. The ability says that itself, but the label makes it more obvious. The third, "He upheld his vow to the elves by watching their woods and blazing their trails," is officially called flavor text. It tells a one-sentence story about the Elfsworn Giant. If we removed all the text in italics from the card, the card would function exactly the same, but a player would have more difficulty understanding how to use its abilities.

“Flavor text” is not a recognized category in official rulebooks.

The official rulebooks do have stories. For example, page 4 of Player Core has

“I think it’s heading your way!” Ezren called out a warning from the nearby alley, but it was too late.

Without so much as a whisper, the translucent elven warrior charged from the shadows, its blade passing through Kyra’s side without meeting any resistance. It was almost as if it were just an illusion.

But the pain it caused was quite real.

“By Sarenrae’s light!” Kyra exclaimed, falling to her knees and clutching her wounded flank.

This was the third night they had spent searching the back alleys of Caliphas for the legendary ghost that had preyed upon the townsfolk for months ...

That is for flavor.

But I presume that we are talking about the first sentence of most feats and features. For example, Acrobatic Performer says,

Acrobatic Performer Feat 1

General, Skill
Prerequisites trained in Acrobatics
You're an incredible acrobat, evoking wonder and enrapturing audiences with your prowess. You can roll an Acrobatics check instead of a Performance check when using the Perform action. If you are trained in both Acrobatics and Performance, you gain a +1 circumstances bonus on Acrobatics checks made to Perform.

The first sentence, "You're an incredible acrobat, evoking wonder and enrapturing audiences with your prowess," is narrative rather than mechanical. But it is not a story about an acrobatic performer. Instead, it is instructions on how to visualize the ability. It is not a mechanical rule in which "evoking" and "enrapturing" are effects in the game mechanics, but it clarifies the roleplaying of the mechanical parts. The character shows off Acrobatic prowess to conduct a Perform action with an Acrobatics check. Mechanically, they can do this even when the performance is singing, which would be confusing without some narrative. But we can visualize this as like a rock star strutting across the stage to add dynamic motion to their songs or an orator pounding his fist on the podium for emphasis.

REACTIVE STRIKE [reaction]

Trigger A creature within your reach uses a manipulate action or a move action, makes a ranged attack, or leaves a square during a move action it’s using.
You lash out at a foe that leaves an opening. Make a melee Strike against the triggering creature. If your attack is a critical hit and the trigger was a manipulate action, you disrupt that action. This Strike doesn’t count toward your multiple attack penalty, and your multiple attack penalty doesn’t apply to this Strike.

The sentence "You lash out at a foe that leaves an opening," sketches the reaction the fighter takes. "Lash" means Strike and "opening" means the triggering effect. A Strike is pretty clear narratively, but making a Strike during a reaction requires additional narrative explanation. Once again, this first sentence is not a mechanical description but it is also not flavor text.

For a player who thinks of Pathfinder as a series of mechanical effects, the narrative first sentence looks like mere flavor. But its real purpose is to add natural-language narrative clarity to the following mechanical description in game terminology. It is more like the reminder text and ability labels on Magic cards than the flavor text on Magic cards. For the rest of us who read the narrative sentence, we know that it uses non-jargon language and thus is poor for rules discussions about mechanics, but it is good for adding narration to an action, "I strut up to the stage and dramatically throw my cape back over my shoulder as I begin my story about a cinder rat. I motion its scurrying over the fire planes with my whole body." (I recently played the Tell a Tale section of Spoken on the Song Wing in which the PCs participated in a storytelling contest. The rogue who told a story about a cinder rat would have done better to roll Acrobatics, but she lacked Acrobatic Performer feat.)

My general attitude toward subsystems is illustrated by the Caravan Combat Subsystem in The Brinewall Legacy, 1st module of the PF1 adventure path Jade Regent. It replaced the individual player characters fighting against raiders attacking their caravan with some dice rolls that represent the entire caravan fighting back. I did not see the point of teaching them a whole new subsystem so that they don't get to play their characters as designed. So I skipped the Caravan Combat Subsystem and just let their characters fight a few of the raiders directly, telling the players that the rest of the caravan defenders would perform as well or as poorly as they did. Later I read in these forums that the caravan system did not work properly.

The big issue is teaching the subsystem. I own the modules and the players don't. In the days when we gathered in person around around a game-store table, I could hand the module around, but that meant that the players had to memorize the subsystem and then pass the book to another player. These days, we play online via Roll20. I could copy subsystem pages out of the PDF and email them to the players, but technically that is a copyright violation. Paizo and the Archives of Nethys are kind enough to include the Life in the Academy subsystem online for Kindled Magic in Strength of Thousands, but I rewrote that subsystem to include attending classes and having adventures on class field trips.

On the other hand, Paizo writers do invent one-shot subsystems designed for non-combat challenges that are pretty easy to explain. In our last game session, I ran the "Tell a Tale" event in which some PCs participated in a storytelling contest about magical animals. We had two absences that day among our seven players, so it was a good event that we could justify players skipping. The module writer Quinn Murphy invented a special exploration activity for this event:

PRESENT A TALL TAIL

Auditory, Concentrate, Exploration
You present one aspect in your narrative about a meeting with a mystical animal. Each attempt takes 1 minute and requires you to attempt a DC 20 skill check. You can choose between describing the animal (Nature), inventing something supernatural about it (Arcana, Deception, or Occult), recalling past tales about the animal (Society), adding narrative drama (Deception or Performance), or connecting with the audience (Performance). You can’t take any of these options more than twice in the same story.
Critical Success You gain 2 Story Points.
Success You gain 1 Story Point.
Critical Failure You lose 1 Story Point (to a minimum of 0) and can’t retry the option you just attempted, even if you haven’t yet used it twice.

I had some trouble understanding the description of the activity, but I decided--based on the four NPC contestants having earned 7, 5, 4, and 2 Story Points each--that the storyteller gets up to five rolls: one for describing the animal, one for its supernatural powers, one for choosing an animal from familiar folktales, one for drama, and one for exciting the audience--using one of the skills listed. This might be wrong, because the line, "You can’t take any of these options more than twice in the same story," makes no sense in this interpretation, but the bard in the party rolled 8 Story Points, the kineticist and wizard rolled 5 Story Points each, and the rogue rolled 2 Story Points, so their results came out in the right range. (The magus did not participate, and the 2nd bard and the champion were absent.) Weirdly, many of the players' stories were retellings of encounters the party had with animals, but with the animals able to talk.

PRESENT A TALL TAIL was quickly-teachable Victory-Point subsystem and more fun than a single Performance check. I described the five rolls and the player had to give a sentence about the story for each roll. And it fit the Uzunjati storyteller theme of the Magaambya Academy in Strength of Thousands. The magus who skipped the contest was a combat-oriented character who attended the Magaambya Academy to learn to fight better, so the player not wanting to participate fit his character.

This is all good information, but I feel I may have failed to communicate my actual needs. I need to make sure I know WHAT more than HOW if that make sense.

People cannot tell you WHAT information to tell the players if they don’t know WHAT your campaign is about.

I agree with steelhead that we don't know enough about Dungeon Master Zack's campaign to be specific about WHAT. But I can recommend what TYPES of information to give.

A good example of campaign-specific information is the Player's Guides that Paizo publishes for their adventure paths, such as the Player's Guide for Strength of Thousands. (A direct link to the Strength of Thousands Player's Guide would start a download, so I provided a link to the Strength of Thousands page that contains the link.) These guides are meant to be read by the players to provide essential information and recommendations.

The guide starts with a cover page and table of contents.
The Magaambya, page 3, describes the Magaambya Academy setting and its traditions
Alignments, page 4, recommends alignments for the characters. That no longer matters due to the Remastering dropping alignments.
Classes, page 4, suggests the best classes for the characters, who are roleplaying students of magic at the Magaambya. It also recommends giving free archetypes, an optional rule.
Ancestries, page 5, lists the best ancestries for typical students, emphasizing common ancestries near the Magaambya but pointing out that students come from all around the world.
Languages, page 5, lists the local languages.
Skills and Feats, page 5, points out that this campaign will have more social interactions than most campaigns, so it recommends learning skills such as Diplomacy, Deception, or Intimidation. The recommended skill feats are magic-oriented ones, such as Arcane Sense and Recognize Spell.
Backgrounds, page 6, says that all common backgrounds are workable, because people come from all walks of life to join the Magaambya. But it presents four custom backgrounds especially for Magaambya students.

Session Zero would add further information for fitting the player characters into the setting. In my game, the tengu bard Jinx Fuun had a backstory that she was a tengu curse eater raised shipboard on the nearby Arcadian Ocean. She met Magaambyan teach Takulu Ot on shore leave and he persuaded her to study at the Magaambya, Her Sponsored by Teacher Ot background reflects this backstory. The elf magus Zandre is an Ekujae elf from the mountains of the Mwangi Expanse north of the Magamabya. She came to the academy to study how to battle the dragons that menace the area. The fleshwarp rogue Roshan is the daughter of a researcher at the College of Dimensional Studies in Katapesh whose kitsune physiology was merged with an ifrit in a lab accident that also gave her elemental powers. The Magaambya is a better school for mastering those powers than the College of Dimensional Studies. The anadi wizard Idris simply wanted to study wizardry. His player selected Magaambyan Attendant for his free archetype to emphasize that his priority is his magical studies (and the player wanted to play with the Mask Familiar feat). The players are creating this information, but the GM has to advise how to make it consistent with the setting.

The best way to handle giving players initial information is via Session Zero (link to GM Core on Session Zero). On the other hand, sometimes players need some information before Session Zero, such as when I ask my players which adventure path we should run next. Session Zero is a get-together before the roleplaying starts.

The main purpose of Session Zero is to ensure that the players create characters who can work together, preventing such problems as an evil thieving rogue in the same party with an evil-smiting paladin. But Session Zero also introduces the players to the setting and allows them to work with the GM to become part of the setting. Session Zero also provides time to assist new players with character-creation rules.

For example, in my Iron Gods campaign, set in the town of Torch in Numeria, the plot hook was the town council offering a reward for adventurers finding their town wizard, who had failed to return from an expedition in the caves under the hills. I had to describe the town itself, its government, its industry (usually neglected but the wizard's expedition was to save the industry), and its troubles. One player decided to play Boffin, a gadgetter who worked at basic smithing for the missing wizard; another player played Elric, a magus who ran errands for the wizard; and a third player player Kirii, a strix skald who met Elric on the road and accompanied him back to Torch. Later, we added Kheld, a caravan guard (human fighter) assigned by a local business owner to join the party. Further details at Iron Gods among Scientists.

The key is to give the players enough information so that they can start roleplaying their characters immediately, instead of asking questions about the setting. A GM could keep information about their quest secret until some patron tells them the quest, but telling them the quest in Session Zero lets them create characters who will be excited about the quest.

What my players look for in an adventure path is a theme, setting, and plot that lets them have fun in roleplaying. Some of my players like to roleplay social interaction, but others don't care for that. Some like problem solving. Nevertheless, combat is the common denominator for something challenging.

But I can adjust combat myself. I could bump a creature up a level, drop them down a level, remove one foe, or add more. The adventure path has labels such as MODERATE 5 to let me know the default at a glance. Because I typically run large parties, I consider whether to leave it easy and unchanged for story purposes or increase the challenge for the excitement.

A new GM might not have mastered adjusting combat difficulty. Thus, I like that the default is easy combat for new GMs with new players.

Theme and setting are harder to change. So those need to be good from the start, which makes that them most important part of an adventure path. I am getting better at adjusting theme. The theme of Strength of Thousands is that adventurers enrolled at the Magaambya Academy grow in status as they tackle the dangers around campus and on field expeditions. My players wanted to emphasize that they are students, so the reasons why they encounter adventures between their classes became more central to the campaign. I have had to generate over a dozen additional NPCs to balance the new theme. I usually leave the combat encounters unchanged, because the students are not supposed to have dangerous combat at a respectable school, but I add more field trips with teacher-anticipated dangers to make up the diluted XP; for example, River into Darkness Revisited.

Plot is often derailed. I have learned improvisation for that.

I feel like these sort of "gross errors" in adventures often come down to fine nuances of how you present them. Shade them a little bit differently and they make much better sense.

I give my players narrative control over the campaign, which often forces a lot of improvisation on me. In Strength of Thousands they have also assumed control of the tone. They want to roleplay as students rather than adventurers, though their characters all came from adventurer backgrounds and are better at the rough stuff than the typical student. Thus, I have been re-shading the tone in the modules.

Yet surprisingly, rewriting the modules to let them roleplay as students in the Magaambya Academy highlights how much the authors had to violate the school premise in order to let a group of players have adventures without teacher oversight or police involvement.

In this case:

* Law enforcement is not the same as pure (military) prowess. The chime ringers get much more training in de-escalation and solving issues civilly or nonfatally than a soldier or war-mage would get. The Magaambya prefers to let them do most of the policing because they're the experts.

My players asked for student jobs rather than getting a stipend for free.

[Supply manager] Xhokan gives each hero their initial stipend of 40 gp and some basic supplies, such as parchment, quills, and ink. He explains that, so long as they’re students in good standing, they receive basic school supplies and 4 gp at the start of every month. He gives them their first 4 gp as well but bristles at any suggestion that he should’ve simply given a lump sum of 44 gp—to Xhokan’s keen mind, the amounts are separate and should be given and tracked separately. This initial stipend seems like a lot of money, but it’s the bulk of the party’s treasure for this chapter.

For the Spring Season, they worked at tending the Leshy Gardens. For the Summer Season, they served as guides for visiting students from the Student Exchange module. For Fall Season, they were deputized to help the Chime-Ringers. For Winter Season, they will assist in a research project with visiting scholar Kassi Aziril.

The Chime-Ringers are notable in Spoken in the Song Wind for never appearing on scene. The PCs can go to them for information in their own independent criminal investigations, but that is all. I noticed the number of criminal investigation missions in the module and invented the Chime-Ringer assisting job to provide more justification. In addition, I decided to invent a Chime-Ringer character to work with them, and by lucky coincidence, this gave me a reason to add the runesmith playtest character to their adventures as their Chime-Ringer liaison. I presume that Spoken on the Song Wind put the Nantambu Chime-Ringer archetype in its Adventure Toolbox articles in case a GM wanted to add a Chime-Ringer NPC.

* But, this is also an adventuring world and "levels" is something that the Magaambya is aware of at least on an intuitive level. They recognize that Froglegs is more than the neighborhood watch can handle, but unlikely to be a cosmic level threat. They estimate that the PCs are in fact more powerful than your typical chime-ringer and reasonably placed to do this.

In a way, the PCs are a pretty reasonable choice - the way most AP encounters are balanced, it's extremely likely that the players will win, so yes, the PCs are indeed a powerful enough supernatural SWAT team to send in there.

The Nantambu Chime-Ringer archetype says, "The Chime-Ringers serve as Nantambu's elite town guard, keeping the peace while carrying on ancient traditions of crafting chimes in metal, bamboo, or glass." They are a police force rather than a neighborhood watch.

But I checked over the levels of Nantambu characters in this module and in Lost Omens Mwangi Expanse and decided that the Chime-Ringers would vary from 2nd level to 4th level. Sending a squad of 15 4th-level Chime-Ringers to capture 8th-level Froglegs and her gang would have worked, but sending the 7th-level deputized party was more sensible. (The module intended this mission for 6th level, but I moved the mission to take down the serpentfolk spies earlier, so the party leveled up.)

As for taking Froglegs alive - I would not interpret it as "it's okay to kill anyone else, just not Froglegs". You can read "survivor" also as "not knocked out".

"Not knocked out" means still on their feet and fighting. My party typically goes with unconscious but not dying, either by means of nonlethal damage or by casting the Stabilize spell on a target dying from lethal damage (most spell damage is lethal). One exception, however, is the fleshwarp eldritch-trickster rogue PC Roshan. She is a master in Athletics and grapples the enemy. After Roshan rolled a critical success against the gang member Hrroupo, the party put manacles on his arms and his legs while he was restrained but close to full health. I reduced the time to apply manacles from 10 seconds down to two Interact actions for gameplay convenience.

I'm looking at the adventure now and it doesn't say fight to the death, it says "surrender when X left standing". Knocking them out or otherwise disabling them should work too.

Not surrendering means to fight or escape. I had the burglar Yorulu try to escape, but she Hid in a tree without time to move from her square and the anadi wizard Idris successfully made the flat check to hit her with a Ray of Frost. Then Idris ran over to cast Stabilize on her (Idris knows primal cantrips via his Magaambyan attendant free archetype). Idris's player described a time she saw many raccoons hide in a leafless tree in winter, so she was okay with believing that Yorulu could climb a tree and disappear from sight.

And Froglegs does have a sad backstory that explains why she went bad, and the book discusses possible redemption.

That will be for a judge to decide. I added a courthouse map to my Roll20 game because the PCs ended up testifying so much.

This AP is pretty okay in having enemies that can be defeated without killing them. The fey are kinda silly and don't really realize life in the Universe is more serious than the First World. But they can be scared with iron bells much of the time too.

Student Exchange had some fey into burglary just for the sake of entertaining mischief.

I also think it can be a bit on the GM how this plays out. If the PCs capture a mook, does the mook suddenly turn into a prison break genius and it's a huge pain for the players? Is it really hard to hand off captured enemies to NPC lawfolk who'll take care of them? Or is killing mooks really the only way for your players not to get saddled with a huge amount of trouble?

Part of making the Chime-Ringers show up in my campaign is that I developed Nantambu's legal system, too. The PCs testified in court and mayor-magistrate Ananda or councilor-magistrate Owethu judged the defendants. The typical sentence has been a few years at a labor camp deep in the Mwangi Expanse, but the professional assassin Nairu was executed. The Crime & Law section in Lost Omens Travel Guide has a handy table on page 65 of common punishments for crimes. Most of the serpentfolk were sentenced to branding (as painful as flogging) and exile, but the leaders were given a one-year geas and hard, dangerous labor in the Sarkoris Scar. Serpentfolk vemon-caller Atathik who surrendered without a fight was sent home peacefully with the message that if the serpentfolk wanted the knowledge of the Magaambya then they could enroll in its classes.

I think I might homebrew a bit more nonlethal magic for the Magaambya though. They've had a couple thousand years to develop techniques that match their ethics after all.

The Nonlethal Spell spellshape is already in the Player Core and I houseruled that any Magaambya student can take that feat. A few spells also have the nonlethal trait and Illusory Creature also deals nonlethal damage. But my spellcasters are satisfied with Stabilize after lethal damage. The rogue Roshan and champion Wilfred like to punch targets for a knockout blow.

I found a lapse in logic in Spoken on the Song Wind, and would like to take a moment to rant. The Finding Froglegs mission in which the party captures the criminal gangleader Froglegs at her stronghold is loaded with contradictions. To let the party act like questing adventurers the module ignores that the city of Nantambu has laws and police.

The party finally gathered enough evidence to arrest Froglegs for murder. Therefore, the Magaambya, represented by Learned One Janiaimo, sends them to capture Froglegs. They have a question-and-answer session with Janatimo. The potential questions are bolded.
Why doesn’t the faculty of the Magaambya take action? “We are! We act in Nantambu through our trusted conversants, and you’ve certainly proven your trustworthiness to me.” (In other words, the Magaambya is taking action by risking their students who have demonstrated prowess in combat.)
Why don’t we just go to the authorities? “The Chime‐Ringers [the local police force] are too busy chasing petty criminals, ..."
Is our arrest binding? “Yes, but only provisionally. Once she’s in custody, ... we can investigate the scope of her network ..."
Can we kill Froglegs? "Murder would make you little more than vigilantes, and we can’t risk such a stain on the Magaambya’s reputation. We need to stop her, but we don’t want to become like her. She’s unlikely to come along willingly, but if you’re resolute and forthright, you can capture rather than kill her."

However, the party has to fight through Froglegs' minions in order to capture her alive. The 1st hut says about the criminals there, "They don’t surrender until there’s only one of them [out of 3] left standing." The 2nd hut says, "The [four] gripplis are dedicated to Froglegs; they fight until only one grippli remains, and that survivor tries to flee to warn Froglegs that the intruders were more than they could handle." And the dwarf Kolbo in that hut is described as willing to turn against the gripplis and kill one to prove that he is helpful. In the 3rd hut, fortunately, the grippli jinxer Grubush is willing to flee and Froglegs is willing to surrender if reduced to fewer than 30 hit points (out of her 135 maximum hit points). So 5 minions will fight to the death.

The module expects the party to kill a handful of minion thieves to reach Froglegs, but has to let murdering crimelord Froglegs live because killing her would be evil. (TV Tropes: What Measure is a Mook?)

Fortunately for my sensibilities, my party has tried to avoid killing intelligent beings ever since they had to chase gremlins out of a storeroom in Kindled Magic. They captured everyone in the 1st hut alive. And manacled them and delivered them to the Chime-Ringers waiting nearby. I declared that of the Chime-Ringers only my 6th-level playtest runesmith Virgil Tibbs is strong enough to face Froglegs, so the other Chime-Ringers hang back and guard the prisoners. We ended the game session before the 2nd hut. And the party has been officially deputized to help the Chime-Ringers for over a month, but the other powerful people in Nantambu, such as Magaambya faculty, have to stay away because they are not deputized.

In late 2023 I was considering running a major Starfinder 1st Edition adventure. For practice I tried a mini-campaign with the Free RPG Day modules Skitter Shot, Skitter Crash, Skitter Home, and Skitter Warp. I did not require the players to play skittermanders like the modules provided. Instead, they had fun chosing exotic aliens as their characters, such Entu Colony and Kiirinta.

Skitter Shot went off well, but trouble started in Skitter Crash. After the crash in a mining vehicle used as an improvised lifeboat, they found the Monitoring Station and were extremely disappointed that they could not, no matter how much Computer and Engineering skills they applied, use it to contact the Helix Lyceum Research Station nearby. Instead, they had to walk. And fight monsters. Starfinder is science fantasy, but they wanted more science. They complained that Starfinder should have a stronger science-fiction style. I added a research project at the Helix Lyceum for them to participate in, and that made them happy. In Skitter Home I had to rewrite the adventure so that the local authorities were more involved. In Skitter Warp a disaster opened up the same planet from Skitter Crash to the plane of Hell, and the party would not rest after defeating the local demons. Nope, the players insisted on evacuating the Helix Lyceum to safety on another planet.

The Starfinder rules were fine. The Starfinder adventures were not to their taste. We went back to Pathfinder, where the fantasy is supposed to be only fantasy.

A personal example of this is freezing rain. Many people whose opinions I trust say freezing rain is an excellent spell that can really mess up an encounter, and that they've used it to great effect many times. When I read freezing rain, it looks like one of the most finicky spells I can imagine and it looks like it would forever to really get value. I would probably never try it without being told to do so. It seems the effect is just that good when multiple enemies are involved, even if you're paying 3A for movable mudpit and 1A sustains for weak damage and slowed 1 on fail. The math just works out that you chew up enough enemy actions between difficult terrain and the sustain, even if you're spending a lot of them yourself.

I find that curious, because one player in my Strength of Thousands game really likes Cyclone Rondo, which is like a 3rd-rank version of Freezing Rain. Her bard Stargazer has to make a small sacrifice to cast Cyclone Rondo, because its 3-action casting means she cannot cast a bard composition that turn, but then her following turns are Stride, sustain Cyclone Rondo, and cast Courageous Anthem. She likes that Cyclone Rondo is both damage and battlefield control and preserves her spell slots. And if she Sustains Cyclone Rondo twice in one turn, it deals damage twice and can move twice (I think that is how the rules work).

On the other hand, Cyclone Rondo is a 3rd-rank spell and Freezing Rain is a 5th-rank spell, so I feel that Freezing Rain should be four times as awesome. It isn't. It covers 16 times the area of Cyclone Rondo, can move twice as far, and can avoid dealing damage if necessary. Both spells deal the same damage. Alas, bigger is not more awesome. Cyclone Rondo feels like chasing enemies with a mini-tornado, but I think Freezing Rain would feel more like a big area effect.

Stargazer is never going to learn Freezing Rain for comparison, because it is not on the occult spell list.

Let's just go back to the barbarian feats. Of all of those feats, I think Furious Finish is probably the most interesting. Is it a feat I'm choosing if I want to play the best possible barb? Probably not. But there's a real hook to the risk/reward, the damage, the way other people will want to buff you up for this juiced 1A attack. The way it plays off your rage timer is fun. It has cool things going on, interacts with other mechanics, has an element of resource management, and has good flavor to boot.

Furious Finish ends the barbarian's rage and leaves them fatigued and unable to rage again in exchange for dealing much more damage on a Strike if the Strike hits. Sounds dramatically awesome. It could also backfire dramatically on a miss.

I think that Witch of Miracles' risk/reward property is about being able to be awesome. I have a bias toward this conclusion, because I believe that my role as a GM is to give my players opportunities to demonstrate that their characters are awesome.

In a later comment,

I would generally say that PF2E trusts its math too much to sell an ability's power, and neglects selling abilities on playfeel and neglects trying to sell them to people who won't understand the math.

No, the math says that feats are weak, even class feats which are the strongest kind. Most of a characters power comes from ability scores, proficiency rank, adding level to proficiency, and its strongest class abilities. So the math is not selling an optional ability's power.

Witch of Miracles had talked earlier about the prevalence of +1 and +2 bonuses. On a d20, a +2 is merely 10% more. It can have a 20% effect, because it usually means a 10% chance of turning a miss into a hit and a 10% chance of turning a hit into a critical hit, but the most dramatic hits are against high-AC opponents in which the chance of critical hits does not increase because it requires a natural 20 even with a +2 bonus to hit. A temporary +2 bonus purchased with an action is practical efficiency rather than awesomeness.

Also, the more you play PF2 as a group the more fun it gets, in a way that PF1 couldn't really match. In PF1, when you have system mastery, GMing becomes functional as you learn the ins and outs of encounter design, but Players also learn more and more tricks to tip the scales, so you don't reach a point where all of the group's combined experience is making the system sing. There's a tension there.

This is the key to awesomeness that my players use. They can outsmart and dominate an enemy force through tactical teamwork.

Let me use Fort Phaendar from Ironfang Invasion as an example. The party is supposed to liberate the village of Phaendar, converted into a fort by the Ironfang Legion, in PART 2, HOMECOMING at 16th level in Vault of the Onyx Citadel. My players did it at 12th level. I reduced the level of the troops at Fort Phaendar, because I lacked time to convert them all to PF2 rules. I mostly replaced various enemies ranging from CR 10 to CR 14 with 9th-level Hobgoblin Formations. But I carefully ported CR 14 slavemaster Ettoran Phark and CR 15 commander Stabvistin. Stabvistin was a duel-wielding master of melee combat, good enough to be annoying against a 16th-level party but now he was going to face a 12th-level party.

I have written this conflict before at I love PF2E butttt.... #149, Each Encounter is its own Creature #12, and Full Party Optimization #3,because it was awesome. The party had hoped to avoid the commander, but the monk Ren-zar'jo ended up face-to-face with him. So the rogue Binny came in riding atop the sorcerer Honey transformed into a flying dragon. Honey got in a breathe weapon attack on Commander Scabvistin, but then she had to land 50 feet away to let Binny do her work. Closing in on Scabvistin to attack with dragon jaws and claws would have opened up a world of hurt from Scabvistin's counterattacks. The rules penalize a PC riding another PC, which is why Honey landed so that Binny was sitting on her rather riding. On the other hand, Scabvistin could not simply run up to them, because he had temporarily switched to his backup bow for ranged combat. Interacting to draw a melee weapon and two Strides to the dragon would have left Honey laughing as she flew into the air to move away, taking only a Reactive Strike rather than one of Scabvistin's specialty attacks. Thus, melee-master Scabvistin was forced to stand still and Strike with his bow. And Binny kept him perpetually off-guard with Precise Debilitations, so that she could deal sneak attack damage and two other party members--ranger and druid--attacking from range also had an easier target. Scabvistin was nullified and died.

Okay, Scabvistin was only a Severe Threat against four 12th-level characters, but those characters had been fighting other troops and were not longer at full spells and hit points. The druid was casting Ray of Frost cantrip rather than top-rank spells. At the party's current readiness, he counted as Extreme Threat, but against their tactics, he was Moderate Threat. I loved it.

Precise Debiliations is a good class feat designed to make the Debiliations class feature more versatile, but most rogues would be flanking and not need it. Dragon Form is a good 6th-rank spell, but ordinarily a sorcerer using it takes lots of damage. Used together, they crippled a powerful enemy. That is how my players display awesomeness.

Ryangwy wrote:

Have you... actually fought a decent variety of PF2e enemies? Because they do exactly that. Taking just from Abomination Vaults alone, the early floors have zebub and wisp (miss chance), golem and shadow (immunities and resistances), the bloodsiphon and the canker cultists (large hp pools via self healing), werewolf (regeneration) and of course oozes. Oh, and the mitflits have Bane, can't forget that.

Purely from a flavour perspective, PF2e has as much variety in enemy threat profiles for a given level as 3.PF. Actually, more, because they aren't bound by the false monster-PC symmetry 3.PF suffers from, meaning you can get fun stuff like gibbering mouthers without bizarre contortions. The only 'difference' is that barring extremely terrible build choices, a poorly matched PF2e PC can still wail away for 1/5th of their expected DPR whereas in PF1e it's 0

I may make a separate thread about this, but after talking about this with someone, I've come to the conclusion that many of my frustrations with the system have to do with the implicit principle that monsters of the same level are fungible for the purposes of encounter building.

Yes, please make that thread. I often make threads for side discussions like these, but the topic I would chose would probably differ from yours.

PF1E/3.5 says they're fungible, but everyone figures out almost instantly that's just false. Some monsters are clearly designed for more supportive roles in a combat, some are primary attackers, some are just nuisances. Some are high threat glass cannons with no defense, others are just weak damage sponges. And the thing is, these monsters will nominally be the same level and CR. Putting them in an encounter by themselves will result in widely variable difficulty. The primary attacker will be a much harder fight than the glass cannon, probably. And that's fine. The designs are meant to complement each other. (This is, in part, a natural consequence of having some amount of symmetry between monsters and PCs.)

PF2E makes a point of ensuring monsters of the same level are actually fungible, for the most part; and even when they're not exactly fungible, the way the xp allocation works keeps you from having much going on anyways. To explain:

-The game generally recommends you stick to moderate encounters, and save severe and extreme encounters for more climactic moments. So let's discuss moderates.
-The most enemies you can have in a moderate is 8 APL-4 enemies, each of which will individually crumple to AoE or crits in less than two rounds. APL-3? 5 enemies max. And the counts go down severely if you add in *any* enemies at party level. A single APL+0 enemy caps you to 5 total; the APL+0 and 4 weak APL-4 enemies. Having an APL+1 enemy caps you to 3 total, and a single APL+2 eats the budget.
-Let's say you want a buffing enemy and some lackeys. If you want to buff a lot of enemies, they have to be individually weak, to the point the buffs may not even affect their critrate; and yet if you want the buffing enemy to withstand any sustained fire, they start to eat up most or all of the encounter budget.
-Furthermore, let's look at the monster building tables. Even if we pick an APL-4 creature with a high or extreme to-hit for its level, it's only around the mark of a low to-hit for our current level. And monsters with higher damage will usually have lower to-hits.

I have repeated those Moderate-Threat recommendations in my thread Encounter Balance: The Math and the Monsters. But in my campaigns, I often set up back-to-back Moderate Threats or a stream of enemies that delivers a new Low Threat every two rounds. This lets me throw lots more Level-3 enemies at the PCs.

The fungibility of monsters is an interesting point. It means that we can calculate how difficult the monster will be to defeat. It does not mean that the monsters feel the same in combat. One number cannot describe much.

Futhermore, I have had variety in combat without variety in opponents. I repeatedly threw Hobgoblin Soldiers against the party in my PF2-converted Ironfang Invasion campaign. The 2nd module created some higher-level hobgoblins for fighting the party at 4th and 5th level, but that did not feel right, so I grouped 4 Hobgoblin Soldiers together into a Large 5th-level troop unit that I called a Hobgobli Troop. High hit points, low AC, fewer attacks with a higher attack bonus. Later I grouped 4 Hobgoblin troops into a 9th-level Hobgoblin Formation. It let the players feel like they were fighting the same army, but more of them at once. The goal of the adventure was not combat as a sport but combat to save their villages from invasion. So against the Ironfang Legion, most of their opponents were the same, except for three varieties across three ranges of levels, and the players liked that. But they had variety in ambushing patrols on the road, rescuing villages from occupying hobgoblins, taking down hobgoblin garrisons, defeating the assault on Longshadow by 20 Hobgoblin Formations with the aid of Longshadow archers they trained, etc.

The end result is it's hard to make a moderate encounter out of this that feels too threatening. It would be ideal, sometimes, to have more monsters that have to-hit and damage more closely matched to an on-level encounter, but also have worse initiative and die quicker. This is easier at low level. But since the monsters get beefier as you level, this becomes less and less plausible; APL-4 moderate HP becomes pretty close to the current level's low HP as the game progresses. The game just isn't designed with this kind of thing in mind. There are all kinds of threat profiles the game can't really create if enemies have to be roughly similar in threat level when standalone, and it frustrates me to no end.

Moderate Threats are not supposed to be threatening. They are the weakest threat that feels like a real fight, so the party has to stay on their toes during the encounter. An adventuring day can have several Moderate Threats because the party does not have to consume significant resources during the fight. A seriously threatening fight, such as Severe Threat, forces the spellcasters to use their highest ranked spells and the martial characters will be so low in hit points that they need healing magic. Too many Severe Threats would exhaust the party beyond what 10-minute rest and recuperation periods can restore.

But on a day when the party will have only one encounter, Severe Threat is fine.

If you want something like an APL+2 buffing monster with weak attacks for their level and good defense, and 4 lackies with the best threat profile you can muster, that's... an extreme encounter, and the lackies are still capped to APL-2. That's how restrictive the encounter builder is, here. You end up with a punching bag of a wannabe bard and enemies that can be hit by incap spells.

Putting 5 enemies against a 4-member party and one of those enemies is definitely more powerful that than any party member and the other enemies are not totally useless--what do you expect? It will be a tough fight that the party could lose.

But my players have defeated Beyond-Extreme Threats. They most often used a divide-and-conquer strategy after carefully scouting the situation. One Extreme Threat becomes two back-to-back Moderate Threats when divided. One Beyond-Extreme Threat can become a stream of Moderate and Low Threats.

The game puts up a lot of safeguards to get its set-and-forget balance, and I really feel those rails every time I interact with the encounter building.

I have dug deeper into the restrictive math and learned how to manipulate it for more freedom. And my players are quite clever about tactics.

I have been reading about investing lately, and "Set and Forget" is lamented as an underperforming investing strategy. Has the phrase been used elsewhere?

Mathmuse wrote:

Rich was a bad player. This is common for the first few sessions with a new player, but he stayed bad longer than most newbies. His goals for his character Kheld exceeded the abilities of a character of Kheld's level. He wanted Kheld to be able to do everything: fight, negotiate, solve mysteries, run a business, etc. He kept asking questions about character builds to optimize Kheld as a Renaissance man who was also the best fighter in the world. He multiclassing into investigator to get the skills he wanted. Eventually, I directed him to several build optimization guides on the internet.

Rich sounds like a poor fit for TTRPGs in general as no system would meet his goals for his character at that low level and while many systems could let him dabble in mutliple skills at higher levels, this seems like a player who wouldn't be satisfied unless he was the best at everything he wanted his character to do. I feel the issue here is with Rich, not any TTRPG system he interacts with.

Rich slowly became a better player year by year. And he liked my roleplaying campaigns.

I do not want to reserve Pathfinder 2nd Edition to a top tier which only good players can play. I want less capable players to enjoy it, too. Rich made mistakes, but he was bright enough to understand when the other players explained the flaws in his logic.

In The Divinity Drive, final module of Iron Gods, Rich's character finally reached his goal. The security chief of the spaceship Divinity lent Rich's character Kheld a suit of Powered Armor. That is an artifact-level item and it made Kheld as overpowered as Rich wanted. Security chief Bastion was supposed to be an enemy, but our campaign got weird and he became an ally: Make a roll for existential philosophy (Divinity Drive spoiler).

Our next campaign was GURPS Dungeon Fantasy, and Rich's character also messed up badly there. But the campaign after that was the PF2 public playtest and he did reasonably well. We switched to weekly boardgames until our first real PF2 campaign and Rich made a champion. A month later Rich developed brain cancer and he died 4 months after that at age 65.

Both Pathfinder 1st Edition and the beginnings of Pathfinder 2nd Edition made Rich happier in his last years, and I do not want to deny anyone that opportunity.

Tridus wrote:

This completely ignores the players that don't know how to do that, show up with something drastically underpowered because of all the trap options baked into 3.5/PF1, and expects the GM to just make it work somehow.

It seems like that player is the problem, and one easily solved with a season zero followed by a session 0.5 where everybody builds characters together from allowed sources. The better builders will help the new player get the character they want while keeping the power levels even.

Session Zero is not always enough.

A new player joined our PF1 Iron Gods campaign at 2nd level. He was new to roleplaying games but had plenty of design help from the other players, who encouraged him to play a Fighter because the party needed another heavy hitter. As I said in my chronicle, Iron Gods among Scientists, comment #21

Kheld was designed by committee of the other players to fit that background. The 20-point build gave Str 17, Dex 13, Con 12, Int 14, Wis 13, Cha 10. Rich chose Ancestral Weapon, a Numerian regional trait, but was more familiar with real medieval weapons than Pathfinder weapons, and wanted a lucerne hammer without the long handle. I said that the ancestral weapon could be a custom weapon. After more discussion, he decided against a hammer, but his ancestral weapon is still a custom weapon, a cold iron collapsible glaive. His other trait is Conspiracy Hunter, a campaign trait from Council of Thieves, to make Perception a class skill. Starfall is a town with a corrupt government, same as Westcrown, so it fits. (I have been using the Archives of Nethys website as my source of traits.)

Rich was a bad player. This is common for the first few sessions with a new player, but he stayed bad longer than most newbies. His goals for his character Kheld exceeded the abilities of a character of Kheld's level. He wanted Kheld to be able to do everything: fight, negotiate, solve mysteries, run a business, etc. He kept asking questions about character builds to optimize Kheld as a Renaissance man who was also the best fighter in the world. He multiclassing into investigator to get the skills he wanted. Eventually, I directed him to several build optimization guides on the internet.

At 7th level, my wife's character Boffin took Leadership in order to gain a robot companion. She was essentially inventing the PF2 Inventor class by means of PF1's strong customization. This worked so well that Rich and another player both took Leadership at 9th level. The strix skald Kirii gained a lyrakien skald apprentice Tay. The fighter/investigator Kheld gained a samsaran wizard Juran highly optimized for crafting magic items. That made sense, because the campaign allowed a lot of downtime for crafting. I suspect that Rich found the build on the internet somewhere.

The first magic item Juran made was a Belt of Giant Strength. But instead of putting the belt on his fighter Kheld, Rich put it on his wizard Juran. He put his two characters side by side on the front line. Samsarans have a -2 to Constitution and Juran almost died while doing very little damage. The other players taught Rich how Juran would play better as a support wizard in the back rather than as a frontline damage dealer.

Sometimes PF1 players persist in their bad builds.

Rich built a lizardfolk champion for my first PF2 campaign, but he got sick and had to drop out.

Tridus wrote:

The 3.x systems are actively new player hostile.

As is PF2, as evidenced by some people finding low-level play too difficult and the complete need for turn-by-turn optimisation in combat, too mentally taxing to be fun. You have a preference for the PF2 style, but that doesn't mean that 3.x is that much more unfriendly to the new player.

We had a whole thread on new players overwhelmed in PF2 back in March, The game doesn't do a good job at teaching new player's how to play. PF1 relies on builds, so PF1 players have to learn optimal builds. PF2 relies on tactics, so PF2 players have to learn optimal tactics.

(maybe) interesting observation after skimming over the replies:

Every single class with Int as key attribute has beeen mentioned. Some of them quite a lot. What is it about those intelligent characters that makes them disliked so much / weak?

I'm sorting the classes by their key attribute. I added the choice of attributes "Strength or Dexterity" or "Intelligence or Charisma", too, but reduced the rogue's "Dexterity or Other" to just "Dexterity."

The classes with STR, DEX, or STR or DEX are clearly martial classes. I guess Paizo knows how to design classes that hit with weapons well. The monk, which was weak in PF1 with "a flurry of misses," works better in PF2. The classes that cast magic spells generally work well, with some exceptions.

The remaining cases are Kineticist, Alchemist, Inventor, Investigator, and Thaumaturge.

Next, let's count mentions in this thread as worst classes. My count is inaccurate because some comments elaborate on mechanics rather than agree that a class is worst. Is that agreement, disagreement, or neither? And I counted every entry in a list for people who posted lists.

Alchemist (non-bomber) 4
Barbarian (fury, superstition) 3
Bard 1
Cleric (battle herald) 1
Druid 4
Fighter 1
Gunslinger 5
Inventor 6
Investigator 7
Magus 3
Oracle (battle) 1
Psychic 3
Ranger 1
Rogue 2
Swashbuckler 1
Witch 5
Wizard 4

The classes not in the running for worst class are Animist, Champion, Exemplar, Kineticist, Monk, Sorcerer, Summoner, and Thaumaturge. Thaumaturge has been mentioned as mixed feelings, but no definite vote for worst. Animist and Exemplar might be getting a pass for being new.

The classes with just one or two strikes against them can be set aside as possibly a matter of personal playstyle. That puts the ranking at:
Investigator 7 INT
Inventor 6 INT
Gunslinger 5 DEX
Witch 5 INT
Alchemist 4 INT
Druid 4 WIS
Wizard 4 INT
Barbarian 3 STR
Magus 3 STR or DEX
Psychic 3 INT or CHA

Investigator was criticized as too specialized, better for detective stories than adventuring stories. Inventor has unstable mechanics. Gunslinger and Magus have a rigid set of actions to take each turn. Druid, Psychic, Wizard, and Witch rely too much on their spell slots, and although Witch has the best familiar for extra flavor some people don't see the point of a familiar. Alchemist works as a bomber but its other options flop. And Barbarian is pulled down in its ranking by a few lame instincts.

All the INT classes ended up on the list, even the Psychic with INT or CHA. And the reason seems to be that for spellcasters Charisma offers more flavor for interesting alternatives to spellcasting, so the non-CHA spellcasters rank lower. WIS-based Druid was ranked low, too. INT also experimented with several martial classes: Alchemist, Investigator, and Inventor, which had their own individual failures.

WRT fictional tropes: Yes, there are tropes like "the villain grossly miscalculates or underestimates the protagonist," "the villain thinks the party is beneath their notice," and all kinds of standard contrivances. I personally find that when the main ways the players interact with things are skill checks and fighting and tactics, and they're expected to earn progress mechanically, they are very aware of when punches are pulled or when it feels like the GM is giving them a bone. Such freebies tend to make the players feel like their agency doesn't really matter.

In contrast, I find the villain underestimating the PCs to be natural. Furthermore, a closely related case is the villain throws their toughest sub-boss and their spare minions at the party, hoping that that is enough, because the villain is too busy to handle the party themselves.

I have an example from Vault of the Onyx Citadel in Ironfang Invasion. In Part 3 of that module, the heroes reach the Vault of the Onyx Citadel, a habitable region in the Elemental Plane of Earth. The Ironfang Legion based in the Onyx Citadel responds.

I could not not simply add enemies over 17th level at whim, because the Ironfang Legion would have sent those people to hunt down the party earlier. Therefore, the Ironfang Legion was running out of forces that could be effective against the party. Creatures that can overpower a 19th-level party would be legendary. The adventure path had forewarned about the five (and only five) high-level characters in the Ironfang Legion: General Azaersi (20), sorceress Zanathura (19 rebuilt as 20), cleric Azlowe (18 rebuilt as 19), cavalier Kraelos (17 rebuilt as 18), and spymaster Taurgreth (16 rebuilt as 17). They all had duties that kept them from hunting the party themselves and earlier in the adventure path Kraelos and Taurgreth were probably lower level. I did invent extra 16th-level enemies, such as Richelle Redrage (tiefling warrior bard 16) who led Ironfang troops that conquered towns.

On the main topic of comparing PF1 and PF2, building and rebuilding creatures is easier in PF2. Thus, I had an easier effort adding extra enemies such as Richelle Redrage to keep the party challenged.

Mathmuse wrote:

The situation with Aria, bard 17, would more likely be the opposite. Aria could be performing on stage when an assassin stabs the duke's son in the audience. The party rushes in to help and then suddenly Aria traps the assassin in a Qunadary spell. Next turn she casts Soothe to heal the son. The party has nothing to do.

This suggests that the very idea of levelled worlds is problematic, and we should either make systems with far flatter growth or ditch levels altogether.

I have played flatter-growth systems, such as Legend of the Five Rings Roleplaying Game, 3rd Edition and 4th Edition. In a system like that, the PCs seldom can win alone. The bigger group wins, so the PCs are the spearpoint of a bigger group, such as scouts of an army or detectives in a police force.

Or that the GM just needs to ensure that Aria has a reason not to prepare any spells not useful to her performance, perhaps a binding contract with the very Duke whose son now lies dead, forcing her to prepare or otherwise not use her potent magic to interact with anything beyond her stage. I would imagine cities and kingdoms hiring minor outsiders to create and enforce such contracts.

Witch of Miracles said, "I used that as the easiest example of a way the level gap could immediately become relevant. It could be just as easily relevant if they tried to lie to get into her graces, or interact with her in any number of other ways." She simply wanted an example of why a high-level bard would interact with a low-level party on neutral ground. I presumed that this bard Aria would be significant in a later module in an imaginary adventure path, and the GM decided to squeeze in an early appearance by Aria to give the PCs a chance to met her. Sometimes NPCs pop up as the plot requires and the PCs have to immediately acknowledge them as friend or foe. That is an awkward situation, and I have added early appearances myself in order to avoid it. Totally rearranging the society of the city to add weird magically binding contracts is an even more awkward situation. Would the PCs themselves be willing to sign such contracts when they accept a mission from the duke?

The statement about this being a difference in degree and not kind is important, here. Someone like Nualia would be a nightmare for first or second level PCs if she were directly translated to pf2e. You can say that Nualia would probably been designed differently if Rise were built from the ground up for PF2E, and that's true. But I don't really consider it positive that you need to weaken the main enemy of the first book in such a direct manner to prevent the question, "why isn't the current arc villain just killing us?" from popping up. She's still quite scary in 1E, but not to the same degree. (And god, I don't even want to imagine that stupid Barghest fight in PF2E. That's an encounter that simply would not exist. Shouldn't exist in PF1E either, to be fair.)

In my wife's D&D 3rd Edition copy of Burnt Offerings Nualla is female aasimar fighter 2/cleric 3 (Lamashtu). That would have to be changed for PF2, since multiclassing works differently. Maybe warpriest cleric 5.

As for the barghest, I was a player in that part of the campaign, and we players simply decided to not open the sealed door that led to the greater barghest.

Nualla was holed up in Thistletop and never directly met the party. She was probably more worried about the professional soldiers that Sheriff Hemlock was recruiting from Magnimar rather than the amateur mercenaries that he left behind. The villain Tsuto did encounter the party directly, but he was only CR 3 and was killed or captured with no chance to report back to Nualla.

The thing that bothers me is how consistently the problem shows up. 3-4 levels is very normal arc length; that's about one book of an AP and about the length of the module. Given that an enemy that's a moderate encounter at the end of the arc is going to be APL+5 or APL+6 at the start of it, why /don't/ they just solo the party early on? They have an exceptionally good shot at doing so. If they take even two lackeys, it's beyond an extreme encounter. Why don't they do that, even? This isn't even some wizard doing a scry and fry. This is literally Jim Sword-and-board ambushing and soloing four people.

A more relevant example is Jade Regent. The backstory is that the Five Storms carefully wiped out the royal families of Minkai so that they could rule through their lackey the Jade Regent. When they discovered a surviving branch in Varisia, they sent ninjas to wipe them out. They unluckily missed Lonjiku Kaijitsu, whom they assumed drowned at sea but had survived. Lonjiku's daughter Ameiko was a surviving heir to the Jade Throne a generation later. If the Five Storms ever learned about Ameiko's heritage, they would have sent CR 9 Kimandatsu and her Frozen Shadow ninjas to eliminate her. If she survived and made it to Minkai in the 5th module, Tide of Honor, then killing her would have been their highest priority. Instead, they sent the 2nd weakest of the Five Storms oni and lost interest after he was defeated. My party (Amaya of Westcrown) in contrast managed to keep the heritage of Ameiko and her sister Amaya secret until the very end. The original players did not let two new players into the secret, though those new players guessed that something was important about those two NPCs. Because the Five Storms never learned about the heirs, they did not make any special effort to wipe out a bunch of well-armed folk heroes.

My players love keeping their characters incognito. In Iron Gods (Iron Gods among Scientists) the party was known in the town of Torch as the young friends of Val Baine who had rescued Val's father from a failed cave expedition, but outside of Torch they changed their names to Gremlinsbane, Nightingale, Lifestealer, Cold Iron, and Jolt. The evil Technic League offered rewards for those illegal adventurers under those false names. For safe downtime to craft, the party simply went back home to their true identities in Torch. When they went to the capital city Starfall, home of Technic League headquarters, they entered under their real names totally under the radar of the Technic League (Inconspicuous PCs Unmotivated in Palace of Fallen Stars).

In Ironfang Invasion the plot of Trail of the Hunted had the party already hiding from Ironfang patrols trying to capture them. In the next module, they were too mobile for the Ironfang Legion to track down. When they finally settled down in the city Longshadow in Assault on Longshadaw the party killed several assassins sent to deal with them. Then the party vanished from the Ironfang Legion's view in two secret missions in the next two modules. The Ironfang Legion headquarters had a crystal ball in the Vault of the Onyx Citadel in the Elemental Plane of Earth to scry the party and send troops with sub-boss commanders to kill the party, but the party was strong enough to defeat the troops. The Ironfang Legion didn't realize that the once-weak party had grown unbelievably strong.

The secret BBEG of the 1st module in Strength of Thousands followed Witch of Miracles' plan. He wanted to murder some Magaambya students to emotionally harm the teachers of the Magaambya, and he developed a grudge against the PCs for foiling a minor plan at 1st level. He sent some gremlins to assasinate them, but my party had reached 2nd level before that attempt and won the battle (actually, they reached 2nd level early due to good roleplaying, so I leveled up the assassination attempt to more dangerous). The BBEG delayed a few weeks to conduct other sabotage and then it was too late as the 4th-level party caught up to him. They did find a plan to murder them asleep in their dormitory among his papers.

The villains do not expect an adventuring party to gain 2 or 3 levels in a few weeks, so they don't feel the need to hurry. Most people take at least a year to earn a new level.

Witch of Miracles wrote:

Like, I've had plot-critical high level NPCs show up just in the background early in campaigns. I will put them in places where it makes sense and the party isn't fighting them, yes, but they're still high level. Like, let's say Aria, the most renowned performer on the continent, is putting on a show; the party is second level and watching them. I'm not going to make this level 17 bard level 4 just because the party is nearby and some murderhobo might try to attack them, and then make the bard level 17 again when the party leaves. That's just awful. I want some semblance of coherence here.

I'm confused; this sounds like more of an issue with the party than with the system. If your level 4 party starts picking a fight with a level 17 character ... let them? At that point it's more coherent to let the chips fall where they may, and the party along with them.

If the issue is that the party would be too easily wiped out by this high-level threat, then why not try something like Proficiency Without Level?

The situation with Aria, bard 17, would more likely be the opposite. Aria could be performing on stage when an assassin stabs the duke's son in the audience. The party rushes in to help and then suddenly Aria traps the assassin in a Qunadary spell. Next turn she casts Soothe to heal the son. The party has nothing to do.

Thus, the GM has to avoid that outcome. A solution of moving the assassination attempt elsewhere would be troublesome if the plot requires that the party publicly rescue the son. A 2nd solution of Aria ignoring the action would be a black mark on her character and might inhibit future interactions with Aria when the party matches her level. A 3rd solution of giving Aria almost no combat skills, despite having a very high Performance skill, is more typical for NPCs. She would have legendary Performance +30 but otherwise count as a 4th-level bard. A 4th solution would be Aria aiding the party rather than acting directly. She could cast 7th-level Haste on the party and then start Courageous Anthem. Her excuse for such a half-hearted effort would be keeping her strongest spells in reserve in case the assassination attempt was a diversion. I prefer the 4th solution.

Like, I've had plot-critical high level NPCs show up just in the background early in campaigns. I will put them in places where it makes sense and the party isn't fighting them, yes, but they're still high level. Like, let's say Aria, the most renowned performer on the continent, is putting on a show; the party is second level and watching them. I'm not going to make this level 17 bard level 4 just because the party is nearby and some murderhobo might try to attack them, and then make the bard level 17 again when the party leaves. That's just awful. I want some semblance of coherence here.

That reminds me of an amusing second encounter in Ironfang Invasion. At 3rd level, the party encounters a friendly traveling Darklands merchant Novvi (N female svirfneblin rogue 2/expert 4). They meet Novvi again at 12th level, 3 modules later, as a female svirfneblin expert 4/rogue 5. The module takes a moment to explain that she had leveled up, "Like the PCs, Novvi has not rested on her laurels, and she has honed her skills as a spy and scout while growing her trading empire amid the fighting between the dark folk and fey." She gained only 3 levels while the party gained 9 levels. The module wrote the 2nd encounter as random chance in meeting Novvi again, but my players, planning on going down into the Darklands, contacted Novvi beforehand via Sending to ask for her guidance. She responded that she would meet them at a certain location--the location in the module.

In Strength of Thousands the teachers at the Magaambya roughly scale up with the modules. Their point of contact in the 1st module, teacher Takulu Ot, was only 4th level (my players insisted that I rewrite him as 6th level). They also meet teacher Zuma (CN male half-orc conspiracy theorist) and teacher Koride Ulawa (CN female human naturalist) without the players nor the GM learning their levels. In the 2nd module, teacher Janatimo (CG male half-elf storyteller 12), replaces Ot as their main quest-giver. They also are assigned a mission by teacher Lesedi (CG female elf summoner 13). In the 3rd module, the 8th-level PCs have become teachers themselves, but they go to Nhyira (NG genderfluid elf historian) for advice. The module avoids giving Nhyira a level. The 4th module also avoids mentioning the teachers' levels.

The 5th module lets slip that teacher Zuma is CN male half-orc conspiracy theorist 9, teacher Nhyira is NG genderfluid elf historian 11, teacher Tahenkot is NG female human defender 11, teacher Ahassunu is N female iruxi diplomat 12, teacher Lesedi is CG female elf summoner 13, teacher Izem Mezitani is NG male aasimar human archaeologist 14, teacher Koride Ulawa is CN female human naturalist 16, and teacher Mafika Ayuwari is NG male human martial artist 17. The PCs are 15th level at this time. Takulu Ot has not leveled up and is still LG male human teacher 4.

I wanted more faculty interaction in my campaign, so all these teachers were friendly to the party in the 1st module. Then I ran into the problem that the PCs would go to a high-level teacher such as Lesedi (13) or Izem Mezitani (14) for aid. See my thread Common Sense versus The Plot. Later I had to pull out my mathematics to figure out how to assign experience points after a Moderate 7 encounter in which Izem Mezitani tagged along to give healing and advice but not to fight.

These level-scaling problems are as much PF1 problems as they are PF2 problems. Both systems have the same power progression. But the calculated flexibility of PF2 makes me as a GM more likely to put helpful NPCs of different levels alongside the party.