I don’t think there is a right or wrong way to play as far as xp or milestones, but I do personally enjoy the transparency of XP as both a player and a GM. Like some GMs might not want PCs metagaming things like “we are 50xp of the next level, let’s go do something important but not story advancing important until we get that next level,” but for me, as a GM, ai love it when my players take that decision on for themselves rather than leave it to me to decide if the big boss fight is going to be too difficult or not, and I am always happy to let the story progress differently based on that choice. I keep a calendar in my games and will always have villains working on projects of their own that advance differently based upon how much time they get before the PCs stop them.
I should also say that I give a lot of XP for players accomplishing stuff with down time, wether it is player directed quests or campaign ones and it seems like XP helps players feel more invested in that kind of stuff since xp is shared. Without the XP I find players wanting to rush through down time, but they get into it when there are more kinds of rewards attached to it.
| 1 person marked this as a favorite. |
XP in PF2 is a constant game of manipulation to get it to where its supposed to be. Not having it eliminates all that math.That's true, but it has at least one use: to measure progress in some way. When you don't have a strict vision of how and when PCs should level up, it helps to know that they actually should about earn a level if you used XP. And so you use it in this way, only implicitly. You could use vibes to determine when to give a level, I guess? But I don't believe my vibes very much :)
It's pretty easy when running most* APs: every chapter is statted for a specific level. It's written assuming the players are level X during that chapter. It'll frequently tell you to get them to level X if they're not. If you're running milestone? The milestone is completing the chapter, which puts you always at the level the AP expects you to be at and eliminates the need to manipulate the XP number to get the expected value.
Hell, in one case, the AP flat out says "the players might level halfway through this chapter, don't let them even if they have enough XP." Which is the writer effectively saying to use milestone levelling without saying it.
The problem is that a level in PF2 is hugely powerful. Even one level drastically changes difficulty. So things are written targeting a specific level because not being that level can make things either trivial or impossible. At which point the GM has to start rescaling everything... or throw some easy XP at the party to catch them up, which is just manipulating the XP number to get it to be the value you want anyway.
If you're using Proficiency Without Level, the value of a level isn't nearly as high and being off by one doesn't skew things nearly so much. But standard PF2 is not a game that still works if the PCs skip over a bunch of stuff and arrive at something 2 levels below the expected level, because that changes a moderate encounter into an extreme encounter and a severe/extreme encounter into a likely TPK. So what ends up happening is people use XP but then game the XP to get the numbers they need to get for the game to run smoothly, which is just milestone with more steps.
In a fully homebrew game this all works better since the GM will be making up all the encounters anyway and can just set whatever difficulty they want relative to their specific party.
*There are a couple of exceptions. The less linear the AP is, the more likely XP will work better. Kingmaker is an example, though again if the PCs wind up underlevelled when stuff starts happening it will be a real bad time. But PCs can overlevel if they focus on exploration.
Abomination Vaults also has a bunch of floors with no real clear "end", so it's not obvious where the milestones should be.
| 1 person marked this as a favorite. |
XP-based leveling can lead to silly situations too. In one PF1 AP, there were many zero-obstacle interactions one could have with recorded events that led to enormous XP boosts (not merely topping off to finish off a level). I felt robbed by that module. JJ responded that the PCs were experiencing things therefore gained XP; maybe the only "meh" interaction I've had with him.
If I recall, a 1st level NPC could get to level 7 just by watching those. And while in dangerous territory where the PCs would have to protect said NPCs, there's was nothing in print to stop the party from shuttling in allies to level them up. All just because the writer had to pad the XP.
(I think there may have been deeper issues too, given the underdeveloped nature of many of the region's enemy factions.)
---
Isn't it silly to say using a different system than XP is denying players anything? Having specific goals lead to leveling should grant players more agency. There's less of a lure to milk the system as say murder hobos as mentioned above. No need to add or subtract encounters nor to pad or deny XP for power balance. Way in the back when, I knew of high-level parties that would go whale hunting because of their high hit dice (which then determined XP rather than actual threat) or rob from the weaker rich folk because gold provided XP too. I prefer being on the opposite side of that mindset, where the narrative successes level one up instead.
|
| 2 people marked this as a favorite. |
I think XP vs Milestone is also a matter of taste. And I can also see it being something you pick for one campaign, but not the other.
For example, if I was doing more of a West Marches style sandbox campaign without much overall plot, XP makes a lot of sense. You spot a cave with a really scary monster in it, you decide to leave that for later when you've leveled up a bit. I think Paizo also recommended that for Abomination Vaults, which has some opportunities to sneak to the next floor or leave some bosses to deal with later.
On the other hand, adventures with more of a story-based plot can work better with milestones. You *can't* make the bossfight easier by dragging your heels and going off on sidequests until you're a level higher than the bossfight was designed for. But you also won't be a level too low for it. A lot of APs have you leveling up after finishing a big event or dungeon, which does feel like they're the logical place for it.
---
Now, the locked door. If I was doing XP in this particular campaign, and enemies coming out is the fail-forward plan for if the party can't pick the lock, then I wouldn't give *more* XP for doing the fight than if you'd managed to pick the lock. Maybe these are enemies that you would have had to fight either here, or later. But if you can pick the lock you can attack them before they're properly organised: they still need to spend actions drawing weapons, moving in from another room, opening doors etcetera. If you fail to pick the lock then you lose these tactical benefits. But the adventure doesn't completely fall flat (fail forward).
| 4 people marked this as a favorite. |
This push and pull comes to a head when a GM wants to wrench the gameplay out of the players' hands and break their expectations to push the narrative in a preferred direction. When you have an appropriately-leveled combat, you give the players experience for that. If you don't, you took that away from them, and you'd better have a good reason. I don't think, "You failed to pick the lock I wanted you to pick" is a good reason to take that away from them.
What does this have to do with anything? A GM that wants to tell their own story no matter what the players do isn't stopped in the slightest by XP being involved. If you have an adversarial table, the GM is going to win that anytime they want to because they have an infinitely sized toolbox of ways to beat the players.
What you're describing here is an adversarial GM. It's got nothing to do with how level ups are determined. (Especially since even XP doesn't stop that. A GM can get around that by using slower progression or just under-reporting creature levels so the encounter is worth less XP than it should be if their goal is to do that.)
And since we're already off-topic, to touch on my issues with Milestone leveling for a bit - Experience is a useful, mutually-agreed-upon system by which players can measure progress and keep the GM honest about how much adventuring they've done, and how rewarded they should be in player progression. It is a points-based system for rewarding the players for following along with the narrative tying things together. Ditching it for "you level when I say so to meet the needs of the story" is an option, but in my opinion makes things feel a little arbitrary. It leads to the PCs being the toys of the GM to tell the story, rather than a cooperative game at a table where everyone plays by rules.
Well its fair to say I disagree with all of this, especially the idea that milestone somehow makes the players the GM's toy. That is not how it works at all and it's a misrepersntation of what it is to claim that.
That'd be roughly the same as me making a bad faith argument like "XP serves no purpose except to act as a dopamine hit for players that require a number going up to stay engaged." Which I'm not.
I feel there's a good design reason that AP's keep experience as the measure of success rather than strictly milestones, beyond just experience being a sacred cow. But I'd be happy to be corrected!
For some APs, there is. The more open ended ones it makes sense. XP works best in an unpredictable, non-linear campaign because it acts as a way of tracking progress that works in basically any situation. It also works great in drop-in games like PFS since it's predictable and people can go at their own pace. Since the early games where these systems were originally created were all homebrew and much more improvisational than an AP, it makes sense they'd have evolved in a way that works well for that. (And if you're running a homebrew sandbox campaign, XP is an effective system.)
The problem with a "typical" AP is that they're linear: they follow a set of events in a set order. If the book doesn't give enough XP to reach the next level or the players miss/skip stuff entirely and don't get the XP for it, they're not at the level the linear adventure wants to be at. The standard fix for that is to throw meaningless random encounters at them until they get enough XP and the story can continue.
And in one infamous case the book gives too much XP and tells the GM to block levelling (or rescale a bunch of encounters).
XP is a bad match there because the linear story being told is being run in a system with very heavy level scaling, and it expects a certain level. If the players aren't that level, the AP either gets much easier or much harder depending on which way they're off and it can make it impossible. So at this point the GM has to just throw XP at them to get to the required number, which is only necessary because XP adds an arbitrary number requirement before you can be the level the story requires.
I can't help but feel I still have some solid questions that haven't really been answered. How do you accommodate the binary tree problem? How do you avoid rewarding failure? How do you design a better game in the first place? And is it worth going through all that effort of implementing Fail Forward Design when it can lead to additional content the players may never see simply because they accidentally passed a check?
The fundamental premise of Fail Forward is "a roll should result in something happening." 'Something' is very broad. It could be good. It could be bad. It could be both.
It can't be "nothing happens, but you can just keep rolling again until you get a number that results in something happening." That is just a waste of time.
There's lots of games that already do this, it's not like I'm inventing anything novel here. Pretty much all of the narrative focused, rules light games work this way. Since you only pick up dice in those when trying to do something that changes the narrative state in some way. A result that doesn't change the state at all is no different than if you never tried anything in the first place. That's why my specific target issues were a plot-mandatory ritual that requires success before the plot advances and can be repeated until you get a success, and skill checks that result in failure with no consequence.
A failure that results in something going wrong did change the narrative state, and that's perfectly fine. Likewise with a failure that results in losing time in a time-sensitive situation.
But the plot should never be blocked until you roll a 16 on the dice and you have infinite retries. What's the point of making you roll 7 times until you get the magic number in that situation?
The latter I'm specifically reminded of with regards to an adventure I've only heard of second-hand, so my apologies of the details are a little fuzzy.
The players arrive at a wizards tower, and are expected to trust some character at the front door - he departs for a moment, or some skill check is necessary to convince him, I'm not sure which - after which he leads the players up to the tower to set up their meeting with the wizard to accommodate their quest.
But if the players choose to be impatient, or fail the skill check to convince him, they are rewarded with an entire wizard's tower of monsters to fight, experience to gain, loot to obtain, etc. All of this is skipped if they "fail" the social encounter (except wherein the players just gain the experience for having skipped the tower as if they had defeated it.) I see this as the end result of correctly implemented fail forward design, and I can't help but think how much more interesting it could have been if the face of this event was a roadblocking lock (which enables all parties to participate in this content) rather than a doorman.
And if you say, "Well, if it didn't result in a good thing, then it wasn't correctly implemented," it starts to sound like there might be no true Scotsmen left in the world.
If that resulted in a good thing or not depends on if the players want that fight, I suppose. I have a group that pretty much takes every way out of a fight possible, so avoiding the fight is a success to them.
A party that wants to loot the place can choose to do that no matter how the skill check goes. It sounds like what this story is really about is just giving the players an excuse to loot the place without feeling like the bad guys. And yeah, you can do that more easily with a lock than with a social encounter.
But that comes down to what kind of game you're running: a dungeon crasher adventure should set that up so that it's there to be fought. One focused more on diplomacy and such shouldn't.
I don't see how its much of a commentary on fail forward design except that there's no case in this where the players roll and the result is "nothing happens", which is really what this thread was originally about.
| 11 people marked this as a favorite. |
I feel like one of the easiest ways to implement Fail Forward that escapes the binary tree is simply by scaling the result--rather like a basic save, in a way.
If the PCs simply must get through this door (or past this obstacle, this NPC's approval) then the die isn't being rolled to find out "Did they get through the door or not" but rather "When they get through the door, what does it cost them in terms of resources or consequences?"
So, perhaps succeeding at picking a door means whatever lies on the other side are caught unprepared (making the encounter easier since they have to spend actions grabbing weapons before they can defend themselves, etc). Meanwhile, if they fail, rather than simply rolling the dice until they succeed, you tell them that it takes a few moments more than expected, and then when the door opens the enemies are ready. Same encounter, but one is easier because of circumstances.
(And scaling further for more results - critical success maybe they're also off-guard until they act and take a penalty to initiative rolls because they're so surprised; critical failure not only are the guards armed and ready, but maybe a few extra bodies from an encounter around the corner have joined this one. No extra enemies, but it's harder to fight one group of six than two groups of three.)
However, a fight isn't naturally going to be on the opposite side of every door. In this case, you have to think of another way failing can make the heroes' lives more tense without fully stopping them from engaging the story.
Let's say the PCs have to sneak into somebody's rooms at a lodge. One way you might do this is by adding a time element. They may only have so much time before somebody returns to the room, so a success means they enter with lots of time to look around. Maybe they collect an extra clue, maybe they have opportunity to steal a minor valuable. Maybe just they get out before anybody wonders where they are. Failure means they have less time, maybe they can't set everything back perfectly so their tampering is noticed even if they get away. If you choose to scale for crits, maybe they just get more than one bonus (especially if regular success asks them to pick between a couple available bonuses). Crit fail could mean they are caught and have to fight, but for this example I feel like if they PCs could confront the NPC directly they wouldn't be sneaking in rooms. Instead, they barely get out with the clue they needed and leave a mess; their target is suspicious and they're going to have to lie to somebody about where they were.
This design might look like "The PCs have the opportunity to find a few things when they explore the duke's room, depending on the result of a DC 25 Thievery check. Regardless the result, they find the Plot Journal open on the desk. If they have time, they can search under the bed (trunk of muddy clothes), the wardrobe (magic item), or the dressing table (jewellery box full of loot). Additionally, if they don't take a moment to tidy up after themselves, it's apparent to the duke that somebody searched the room.
- Crit Failure: There's no time! Grab the journal and go!
- Failure: Search one thing or tidy up
- Success: As Failure, but search one more thing
- Crit Success: As Failure but search two more things"
| 2 people marked this as a favorite. |
I confess, while the iconic locked door is, well, iconic as far as failure stalls go, I feel like it doesn't always make for the best example. While many players have experienced the sudden momentum drop of hitting a locked door and realizing they just need to keep dropping math rocks until it says "16", I also don't see that many adventures being designed where progression is entirely locked behind a literal door. It can be done (like searching the duke's room) but I feel like it's easy to see the locked door in isolation, divorced from context that might explain what it's doing there (and thus illustrate what options you have for failing forward)
For other examples that might crop up more naturally: A magic gate that can only be opened by a ritual. The players have some instructions, but it takes an Arcana check. It's realistic a GM might put the rest of the adventure on the other side of the portal without realising that failing the check is meaningless. Fail Forward options include letting them fumble their way through he ritual but unstable energies give them warp sickness for an hour after they arrive (Crit fail: arcane backlash vapourizes their internal organs deals damage when the gate backfires while opening)
A canyon stands between the heroes and their next adventure. It was on the map and it feels like it should be hard to navigate, but how to make sure they don't just end up stranded here? First stop: If there's nothing interesting to be done here, forget the check. They cross the canyon and it was super hard but they were so brave. No checks, good job everyone. Alternatively, everyone rolls Athletics. Count number of successes and determine what they lost or gained down there that matters (time, if they're in a hurry; hit points, if they need to go into a fight right away; a secret treasure).
| 7 people marked this as a favorite. |
I feel like one of the easiest ways to implement Fail Forward that escapes the binary tree is simply by scaling the result--rather like a basic save, in a way.
If the PCs simply must get through this door (or past this obstacle, this NPC's approval) then the die isn't being rolled to find out "Did they get through the door or not" but rather "When they get through the door, what does it cost them in terms of resources or consequences?"
Yeah, exactly. One of the original complaint points was the SoT book 5 ritual that must be done to advance the plot. This requires a success or better to advance the plot. Failure stops you cold. There's also some minor penalties, but since the plot can't advance until you succeed and there's no real time sensitivity, you can wipe them away before trying again, however many times it takes. It's just making a bunch of rolls until the dice say "the plot is allowed to resume now".
Later in the thread someone posted a Myth Speakers version of the same idea (a ritual that must advance the plot). This version addressed the problem: the plot mandatory thing happens no matter what. If you critically succeed there's an extra good thing that happens, and there's consequences for doing poorly. But there's never any case where it's "just try it again tomorrow and hope you get a different number."
The latter version there is what I'm looking for. You're only ever rolling for that ritual once, and whatever happens, happens.
To put this into picking locks: if getting past the door is plot mandatory then there should be complications if you do badly at it but just stopping the entire plot is a problem. Of course, the best way to fix that is to make the lock one option to get into the room rather than the only option, because failing at it now means the players need to try something else the may have wanted to avoid doing instead.
If it's not plot mandatory, then potential failure outcomes are "you set off a trap", "you alert the creatures in the room", and of course, "a failsafe activates and you can't get this lock open no matter what you do."
Failure is entirely allowed, here. What's not allowed is "nothing happens, roll again until you get a different number."
I do understand why the "pick lock" example keeps coming up in this thread. As has been talked about extensively earlier, it is an activity written to be accomplished in encounter mode and yet most parties will try their best to avoid trying to pick any locks in encounter mode, unless there is some pressing plot reason to do so. But the idea that a locked door is a "succeed or the plot stops" situation is pretty incongruent with my experience playing or reading adventures. You can almost always choose to try to break a door down, and it is really just generally a bad idea for anyone to put themselves in a room where there is only one possible way in or out, much less for that way to be locked in such a fashion that no one can ever get in or out without having the key. That is why we also earlier talked about how there are kind of two different ways of looking at "Fail Forward" as a GM tool and the advice presented in the GM core about it. Personally, I think the advice given on page 11 of the GM core about Fail Forward really needs to be looked at as an introduction to an optional GM tool and not a suggestion to make every possible roll matter, because the game is very, very intentionally designed not to put that much pressure on every individual roll.
"Nothing happens" is one of the most common failure states in the game for the vast majority of checks, and that is fine because time is a resource and most of the time checks are made, it is at the cost of time/ability to do something else. We see this very clearly with picking a lock.
In an encounter (especially a combat one, but also an environmental hazard encounter), having one person holding their lock picking tools, standing at the lock and spending 2 actions to try to unlock it is a massive resource commitment. The activity works perfectly well for this kind of situation, and if it didn't (because there was some kind of 1 minute or 10 minute lock out of trying again) that would mean you could never have encounters with locks in them.
SO the real dilemma that GMs seem to have around locked doors is that sometimes they don't know what to do with locked doors that are not a part of an encounter. Do you make the players roll just to roll, just to see if they get so unlucky they break all their possible picks before they get the lock? Do you skip the roll? Do you essentially change the lock into a different kind of VP style encounter (as most locks already are a kind of encounter mode VP system) that has different failure states than the default activity.
I have strong feelings about which is the "right way" to handle this, but those are just my feelings, not anything objectively provable. I think a lot of players might enjoy more formal alternative lock picking encounter rules being provided somewhere, but it seems like if it isn't an official variant rule in a Paizo product, some GMs will never use it. I would recommend those GMs look at the infiltration subsystem for some ideas, as I think that is probably as close to official paizo advice on the locked door situation as you are likely to get.
I also strongly recommend that GMs treat the last two sentences of the failing forward suggestion as maybe the most important sentences of the guidance, and that is not to put pressure on yourself to make every single roll meaningful when there is not an obvious consequence in your specific scenario. Like feeling like you have to add a brand new combat encounter that you haven't prepared just because the dungeon has a locked supply closet, with a key on a specific NPC/enemy the players might or might not have already found, is really not necessary. It is perfectly fine to let the players roll once or twice or even five times and then either decide to waive the rolling if there really is nothing campaign defining on the other side of the door and the lock is primarily a narrative/aesthetic setting piece that shows that food is locked up and might be a resource that can be utilized by an infiltrating party, or even just set an arbitrary limit if giving it away feels bad to you. "After 5 tries, you've gotten 1 out of 3 successes and broken 3 sets of replacement picks. The damage to the lock is obvious and starting to make it harder to pick. From now on every failure is going to add 1 to the DC." or "You think 3 more critical failures will break the lock and leave you having to break the door down to proceed" are both equally fine ways to play it as just letting the door open without a check, or just saying you open the door after breaking 5 more sets of replacement picks, to just letting the players decide "we are never going to get this stupid lock, let's leave it and come back later," like once they find the key on the NPC who has it.
The reason I think it is such a bad idea to try to codify exactly how all locks should be resolved in every situation that is outside of an active, time sensitive encounter, is because the roll of locks can be so different from each other. Having very tight rules for the time sensitive combat encounter version is really important. Beyond that, it is really on the adventure writer and GM (which might be the same person in a homebrew campaign) to decide if the lock is meant to be an encounter in and of itself, with consequences and rewards commensurate with an encounter, or it if is just a setting piece that could have some significant role to play in the adventure, but probably wont, and then can just be written off if it is slowing down the pace of player enjoyment for no later reward of significance.
I think this can often be true of rituals as well (that it is really up to the GM to decide if the party wasting a day on a failed result matter or not). For the SoT example, the party has spent years of downtime in this campaign. Letting it take a week or more to cast the ritual, and letting the rest of the players do other things during those down time days seems perfectly fine, but if it feels like something that needs more tension, I think the easiest way forward is to just say that in 2 days the moon will be full and the ritual DC will drop by 5, but the magic will become more unstable, so the door might open on a failure, but the DC of the save to avoid the drained condition goes up by 2 and the crit failure result of that fort save leaves the player drained 2.
Again, I think it depends a lot more on the party and whether they like rolling dice and letting chance play that role in their story or if they want the rails of the adventure to be faster paced and more tightly controlled, and the GM can respond accordingly.
| 2 people marked this as a favorite. |
I've been playing this game since the playtest and GMing it since release. I don't think I've ever seen a case of someone picking a lock in encounter mode. The circumstances where it makes sense to do that are so hyper specific that it almost never happens, vs the more obvious solution of "deal with the immediate threat so you're out of encounter mode." (This is as opposed to disabling a haunt/trap where the thing is an active threat itself.)
If someone actually does do it in encounter mode, then there's time pressure (presumably your team would like you have you helping with the encounter) and so failure doing nothing is fine. But almost all locks are picked in Exploration mode, where failure often doesn't mean anything. And that case is the problem.
We're like four pages into this thread and people are still hung up on the difference between the case where failure matters and where failure just means "keep rolling until you get a different result."
I've been playing this game since the playtest and GMing it since release. I don't think I've ever seen a case of someone picking a lock in encounter mode. The circumstances where it makes sense to do that are so hyper specific that it almost never happens, vs the more obvious solution of "deal with the immediate threat so you're out of encounter mode." (This is as opposed to disabling a haunt/trap where the thing is an active threat itself.)
If someone actually does do it in encounter mode, then there's time pressure (presumably your team would like you have you helping with the encounter) and so failure doing nothing is fine. But almost all locks are picked in Exploration mode, where failure often doesn't mean anything. And that case is the problem.
I went through my memories of the Paizo adventure paths I have run and tried to remember the locked doors. Some were in hostile zones, but no-one tried the door until the fight was over.
I do recall one situation in which a locked strongbox was opened under time pressure in Exploration Mode for a social encounter, but my players invented the situation themselves. During Prisoners of the Blight the players learned that each of the five pieces of a disassembled artifact called Dryad's Song could protect its bearer from the blight and that they should try to reassemble Dryad's Song. One piece was held by the ancient black dragon Naphexi. The party distracted Naphexi with praise and food while the thief-racket rogue sneaked undetected over to her strongbox, searched for traps, disabled the trap, unlocked the box, stole the piece, and then resealed the box so that Naphexi would not notice the theft.
The module intended for the 14th-level party to fight Naphexi and placed the ring-shaped piece out in the open rather than locked up, "Finally, a tarnished, gold ring hangs from a simple tin chain, high above her [Naphexi's] bed. This ring comes from the corrupted Dryad’s Song." I invented the trapped strongbox to make the rogue's skills matter. I figured that if she failed, then the party could still fight the dragon and smash the strongbox. Each step--Sneak, Search, Disable, and Pick a Lock--had a single skill check well within the rogue's abilities, because flavor rather than failure was my goal.
I gave the PCs full XP for defeating the dragon and the story award XP for obtaining a piece of Dryad's Song, even though they did not fight the dragon.
A caper in which the lock must be opened before a guard shows up on their regular rounds makes more sense than opening a locked door during a combat encounter. A failure would mean that the party retreats to a hiding place to wait for the guard to pass and then returns to the door.
The most common situation that I have seen picking locks happen in coma at encounters is removing manacles from NPCs. I have seen it in a sinking ship combat and in cultist temple type adventures.
The reason why the rules for picking a lock have to work for combat are pretty clear and there is tons of fiction about picking locks in time sensitive situations so the rules needed to cover that. Outside of time sensitive encounters (combat or not), there are other kinds of tension that can work rather than time, but they require pretty complicated adventure writing or GMs that work well on the fly. Like “how quietly can you pick a lock?” Is not a question the rules really handle at all. Combining party wide stealth checks into other exploration mode activities usually results in combat or has to be handled through a VP style infiltration type encounter.
Every example of an attempt to codify out of encounter mode lock picking I have seen put forward in this thread is either hyper specific to like 1 out of 15+ different types of situations where a party is trying to pick a lock, or is so vague, it is pretty much something out of a rules light rpg that is pretty much just “the GM is going to have to evaluate the situation and decide if there ias a good reason to add a consequence here or not,” which is about timing in a campaign and doesn’t work in any dungeon that is not a railroad where the encounters and consequences are all very linear.
That is why I think those last two sentences of the “Fail forward” section of the GM core are so important, because it’s much more of a “can you do this easily as a GM with what you have prepared? Great! If not, don’t worry about it. Just move on.” People arguing that every roll has to matter or it shouldn’t happen are putting way too much pressure on themselves as GMs or (worse) other GMs to always know, before any dice are rolled, what new situation is going to be created by all 4 possible outcomes. Sometimes those consequence are obvious, and then it’s easy. Other times, it’s not and trying to force it can create problems much bigger than “oh nothing happened except some time passed? Should we just try again?” The GM can then reassess the current situation the party is facing and wing it from there whether to give the success because there are no possible consequences that feel worth adding to the scene, or to pull in anything from random encounters, to a quick comedy of errors monologue to move things along and that is fine for the game. This is true well beyond the lock picking activity.
It is fun when die rolls matter, but making a big deal out of every die roll can get awkward and ruin fun much more quickly than just moving on.
| 2 people marked this as a favorite. |
People arguing that every roll has to matter or it shouldn’t happen are putting way too much pressure on themselves as GMs or (worse) other GMs to always know, before any dice are rolled, what new situation is going to be created by all 4 possible outcomes.
GMs are supposed to know possible outcomes? My players are too creative for that.
But I remember many situations where dice rolls were just for fun. In Spoken on the Song Wind, 2nd module of Strength of Thousands in which the PCs are students at the Magaambya Academy in Nantambu, I had the PCs deputized into the local police force as a work-study job. (I wanted a logical explanation why they did so much crime fighting in that module.) At the end of the semester, after the heroics of the module were over, I decided to get them involved with a speeding ticket. Nantambu lacks automobiles, but it has an extensive canal system. I needed an excuse why they were all together next to a canal.
Therefore, one windy afternoon their dorm-mate Haibram Thodja took them out kite flying next to a canal. He had quite a collection of colorful kites (copied from the internet). The players had a lot of fun simply making Athletics or Sailing Lore skill checks for how well they flew their kites. The skill checks had no effect on any plot elements, but the players liked them.
Later than I had planned since the kite-flying was going so well, two sailboats sped past recklessly with a police officer in another boat giving chase. The catfolk fire kineticist burned the sails off of a speeding boat and the dromaar champion of Cayden Cailean swam over and joined the drunken sailors for a beer before apprehending them. The tengu bard cast Levitate and used her kite to pull her over to a speeding boat. Etc.
| 2 people marked this as a favorite. |
Anyone else ever get kind of jealous about whatever's going on over at Mathmuse's table?
That aside, the only time I can recall anyone bothering with a lock during encounter mode was actually in the playtest adventure, Doomsday Dawn. Even then, I don't remember precisely what happened, but iirc after failing a couple times the player gave up on the lock and rejoined the fight.
That said, I think I'll have to agree that the baseline rules for the Pick a Lock activity don't inherently need to be redesigned to Fail Forward... but then neither do I think that picking a lock mid-combat is a common enough experience to justify that being codified as the default. Rather than designing locks so that even a simple lock requires 3 turns at a DC 20, I would rather see combat-scale lock scenarios (i.e. the manacles etc) attach onto the Disable a Device if it even needs to be a bespoke activity (also takes 2 actions, thematically appropriate) and the activity for getting through locked doors can take a couple moments in Exploration rather than rolling 2-6 times in a row even assuming only successes without interfering with the narrative need for quick lock-picking.
A caper in which the lock must be opened before a guard shows up on their regular rounds makes more sense than opening a locked door during a combat encounter. A failure would mean that the party retreats to a hiding place to wait for the guard to pass and then returns to the door.
Yeah, exactly. This makes perfect sense in an Infiltration challenge since "pick the lock" can be one of the ways to overcome the obstacle, and not getting past it that round increases Awareness. That works great.
Anyone else ever get kind of jealous about whatever's going on over at Mathmuse's table?
They do seem to be a great GM!
That aside, the only time I can recall anyone bothering with a lock during encounter mode was actually in the playtest adventure, Doomsday Dawn. Even then, I don't remember precisely what happened, but iirc after failing a couple times the player gave up on the lock and rejoined the fight.
That said, I think I'll have to agree that the baseline rules for the Pick a Lock activity don't inherently need to be redesigned to Fail Forward... but then neither do I think that picking a lock mid-combat is a common enough experience to justify that being codified as the default. Rather than designing locks so that even a simple lock requires 3 turns at a DC 20, I would rather see combat-scale lock scenarios (i.e. the manacles etc) attach onto the Disable a Device if it even needs to be a bespoke activity (also takes 2 actions, thematically appropriate) and the activity for getting through locked doors can take a couple moments in Exploration rather than rolling 2-6 times in a row even assuming only successes without interfering with the narrative need for quick lock-picking.
Yeah, this makes sense. Even if pick a lock itself isn't fail-forward inherently, any situation where it's been put in as a challenge should be. And since AP writers are often the ones doing this wrong, that's still a Paizo issue. It's just in the scenario writing rather than the rulebook.
| 3 people marked this as a favorite. |
That is why I think those last two sentences of the “Fail forward” section of the GM core are so important, because it’s much more of a “can you do this easily as a GM with what you have prepared? Great! If not, don’t worry about it. Just move on.” People arguing that every roll has to matter or it shouldn’t happen are putting way too much pressure on themselves as GMs or (worse) other GMs to always know, before any dice are rolled, what new situation is going to be created by all 4 possible outcomes. Sometimes those consequence are obvious, and then it’s easy. Other times, it’s not and trying to force it can create problems much bigger than “oh nothing happened except some time passed? Should we just try again?” The GM can then reassess the current situation the party is facing and wing it from there whether to give the success because there are no possible consequences that feel worth adding to the scene, or to pull in anything from random encounters, to a quick comedy of errors monologue to move things along and that is fine for the game. This is true well beyond the lock picking activity.
It is fun when die rolls matter, but making a big deal out of every die roll can get awkward and ruin fun much more quickly than just moving on.
If a roll doesn't matter, why are you rolling it at all?
Presumably the players are trying to do something causing that roll. If roll doesn't matter, there's no reason to roll anything.
Thus if you are rolling something, presumably its to make something happen. If the outcome is "nothing happened, roll again", all you did is waste time. You're going to roll it again so you're still relying on the dice to give an outcome, you just have a dead zone where those numbers on the dice don't do anything.
Like, you're making a big deal out of something that works literally the same way in the end: you're going to get a result that does something at some point. If it's fail forward, you get that in one die roll. If it's not, you might do four die rolls before you get that.
But you're acting like this somehow requires the GM to do a much more complicated thing. It doesn't make any sense.
| 1 person marked this as a favorite. |
Anyone else ever get kind of jealous about whatever's going on over at Mathmuse's table?They do seem to be a great GM!
Thank you for the compliment. I am retired and my weekly Pathfinder game is my main hobby. I have time to add flavorful details to the modules.
This also reminded me of a thread from almost 6 years ago: Becoming a great GM in PF2. I scanned through the first few comments to check for relevance before mentioning it here, and discovered that in the 10th comment Ascalaphus discussed the Fail Forward dilemma! Here is two small quotes from a larger post:
Looking at PFS2 scenarios that are great, and ones that are... less so, I think a lot of it revolve around understanding how to wield the new uncertainty in dice rolls. 1E had characters that could be supremely good in their field and rarely or never fail; in 2E, whenever you're asking for a die roll, there's a chance people may succeed and a chance people may fail.
- Characters are going to fail an "easy" check needed to advance the plot
- Characters are going to fail an important saving throw
- Characters are going to get a series of unexpected critical successes and complete a challenge much faster than anticipated
and
But being able to eventually win, doesn't always make it fun. You might tell players, "statistically you'll win this skill challenge in 40 rounds" and people are going to stare at you like you're crazy.
|
|
I think back then I was looking at it in comparison to the Onyx StoryPath game engine, where if the GM asks players to make a check and they fail, they collect Momentum points, which they can later cash in again for (...something, I don't remember).
But that worked as a brake on GMs willy-nilly asking for checks for things that you should be able to just reliably do. If the GM is asking you for checks just to tie your shoelaces you're statistically gonna fail more often than you'd do in real life. But oh, the amount of Momentum points that you'll collect!
Pathfinder and D&D have always been "success is good and failure is 100% just bad" games, while (not-so-small) indie games have experimented with "failure is bad, but there is compensation" mechanics. It's kind of adjacent to fail-forward because if you have a lot of say, "Momentum", you might be able to try some risky stunts to get out of a bad situation.
| 1 person marked this as a favorite. |
I remember running through an old Paizo AP where the PCs were looking for an item in a dungeon. It was hidden in the treasure room behind a somewhat difficult perception check, but the entire campaign based itself around the idea that you'd find and use the item, so failing the perception check essentially ended the whole adventure. The book didn't even really have guidance on what to do if the PCs failed.
I feel like even for people who are anti-fail forward for whatever reason they have to admit that designing adventures with single points of failure that don't even have good stakes attached to them is really lame.
Experience is a useful, mutually-agreed-upon system by which players can measure progress and keep the GM honest about how much adventuring they've done, and how rewarded they should be in player progression. It is a points-based system for rewarding the players for following along with the narrative tying things together. Ditching it for "you level when I say so to meet the needs of the story" is an option, but in my opinion makes things feel a little arbitrary.
I mean you say one is more arbitrary than the other, but who designs the pacing and rate of encounters in an adventure? The GM or adventure writer. Yeah, in milestone leveling you don't level up until you hit a specific milestone... but in an XP driven adventure you don't level up unless the GM gives you a sufficient number of encounters, which itself is essentially just an arbitrary milestone.
... Ironically I think the rest of your post kind of makes a good argument for milestones. The fact that failing to negotiate with the wizard and being forced to kill everyone is something you consider an objectively superior outcome because you get more XP is a sort of weird perverse incentive that wouldn't exist without the arbitrary rules of XP (and also maybe says something about how you view the game itself, which is fine but notable).
I mean you say one is more arbitrary than the other, but who designs the pacing and rate of encounters in an adventure? The GM or adventure writer. Yeah, in milestone leveling you don't level up until you hit a specific milestone... but in an XP driven adventure you don't level up unless the GM gives you a sufficient number of encounters, which itself is essentially just an arbitrary milestone.
Hurricane's Howl, the 3rd module in Strength of Thousands, is supposed to end with the PCs reaching 12th level. My PCs will probably finish it at 11th level because the 7-member party dilutes the XP. I decided that increasing the number or level of the enemies was not appropriate for the story.
And this will cause zero problems for my campaign. The oversized party can handle the next module at only 11th level because they are oversized.
However, the players want another semester at the Magaambya Academy, so they can experience being teachers rather than students before departing the Magaambya for the next adventure at temple-city Mzali in Secrets of the Temple-City. That semester, especially with the salvage mission I am stealing from Pathfinder Society Scenario 38 No Plunder, No Pay, might level them up to 12th level.
What is the point of leveling up? It gives the players new abilities to try out and lets them handle new, tougher challenges. Thus, leveling up keeps the encounters fresh for the players despite months of playing the same characters. It is also a reward system, but reward by level-up feels personal only if accompanied by a story of character development.
I had an interesting dilemma in Hurricane's Howl. The party was supposed to chase bandits call the Knights of Abendego from the village Kiutu on the edge of the Mwangi Expanse to the town Jula in the Sodden Lands. My problem is that my players mastered methods of fast travel, such as summoning Marvelous Mounts, and could have caught the bandits long before Jula. Halfway through the module, they would have finished the chase and had no reason to go through the rest of the encounters in the Sodden Land. Success rather than failure would have stopped the module. I rearranged the details of the chase to get them to Jula regardless. This is why we GMs meddle with the encounters.
|
|
| 2 people marked this as a favorite. |
@Mathmuse, re: players getting off-track
It occurs to me that we don't see nearly as much threads & rage about "railroading" as we used to say, 5 or 10 years ago.
But it's something I've gotten quite interested in as a GM: when should you railroad and when shouldn't you? Some of my players have very clearly come out and told me that they'd rather have a story that's clearly going somewhere, even if I have to give them a bit of a push to move in a direction.
There's also cases where I think "low agency" is totally fine. Chases for example: you really don't need a ton of agency in a chase. It needs to be exciting, have cool set pieces, and run fast. Deeply character-driven choices are not really what makes a chase good.
But it does feel *extremely* nice when you're running an adventure and the players are supposed to come up with an X-shaped plan for solving problem Y, and you don't really have to push them, but after a bit of discussion someone goes "why don't we try X?"
To me that feels less like "we're on rails" and more like "if you give this cart a nudge, it's likely gonna roll downhill". *Writing* a plot so that the players will themselves come up with the "expected" plan (or something sufficiently close to it) is a sign of good craftsmanship.
| 1 person marked this as a favorite. |
I’ll try to keep this post short but can provide detailed examples and discussions about each point.
1. Sometimes the significance of a roll doesn’t become known until well after it has happened. Players only know that “nothing happens” means nothing will happen if the GM gives them that information.
2. Time is a significant resource. If specific activities happen at a scale that feels off, change it on a case by case basis. For example:
2b. If you want a very simple house rule for picking locks specifically, I’d suggest this: let players decide how long they want to spend on each check. For each full minute committed before the roll, drop the DC by 1. If the process gets interrupted, they have to start over. 10 minutes is the longest you can take per check, to drop the DC by 10. The DCs of picking locks are too high to keep them the same and add a punishing failure results. With the lowered DC, make all critical failure results completely break the tools (not just the replaceable picks). This still makes the consequence time, but if 10 to 30 minutes to an hour to get a complicated lock isn’t enough to be meaningful, the lock is a pretty setting piece, not an adventure challenge.
3. Fail Forward as a narrative tool is a GM skill. GMs like Math Muse and me have decades of experience with doing this and often players who have decades of experience playing and often GMing as well. We should all be careful not to turn “this is something that can occasionally be used in a fun way” into “always do it this way.” Like having a failed lock pick attempt (or more likely the first crit failure) alert guards in the next room is a fun thing to do, but always adding a combat after every failure (something some people have already thought was being suggested in this thread) is going to change a lot of how that part of the game works, likely to absurd ends.
4. How long a leash to give players is table dependent. Players (even just one or two at a table) can make their own consequences much more easily than others. It is awesome when your players do (oh the stories!), but it’s important to check in and make sure it’s fun for everyone and to figure out if you players prefer a more railroady fast moving choose between 2 or 3 options adventure, or like everything is on the table for possibilities.
Which all gets to 5. Why are you rolling the dice? Isn’t a just a question for the GM, it’s a table discussion, and it’s one that can get awkward in multiple different ways if people are making contradictory discussions. A GM might have better options than having a player roll 10 times to open a lock, but that can also be 60 seconds of play time to resolve, where a GM fumbling to make something significant happen could be a full session side track if it turns into an unexpected combat. Which is better will be case by case.
|
Many of the scenarios I've played already incorporate this.
Frequently the setup is that you have multiple challenges that you have to deal with before the final encounter. The number of them that you succeed at determines how tough the final boss is. Scenarios involving Runelords are infamous for this, but they are far from the only ones.
In some cases, failure will induce injury and/or a handicap you will have to deal with. In others, success or a critical success may hand you a scenario bonus.
Sorry something here got deleted by my large fingers on my new phone.
4. How long a leash to give players is table dependent. Players (even just one or two at a table) can make their own consequences much more easily than others. It is awesome when your players do (oh the stories!), but it’s important to check in and make sure it’s fun for everyone and to figure out if you players prefer a more railroady fast moving choose between 2 or 3 options adventure, or like everything is on the table for possibilities.
should read:
4. How long a leash to give players is table dependent. Players with a decade of experience playing and GMing (even just one or two at a table) can make their own consequences much more easily than others. It is awesome when your players do (oh the stories!), but it’s important to check in and make sure it’s fun for everyone and to figure out if you players prefer a more railroady fast moving choose between 2 or 3 options adventure, or like everything is on the table for possibilities.
@Mathmuse, re: players getting off-track
It occurs to me that we don't see nearly as much threads & rage about "railroading" as we used to say, 5 or 10 years ago.
But it's something I've gotten quite interested in as a GM: when should you railroad and when shouldn't you? Some of my players have very clearly come out and told me that they'd rather have a story that's clearly going somewhere, even if I have to give them a bit of a push to move in a direction.
There's also cases where I think "low agency" is totally fine. Chases for example: you really don't need a ton of agency in a chase. It needs to be exciting, have cool set pieces, and run fast. Deeply character-driven choices are not really what makes a chase good.
But it does feel *extremely* nice when you're running an adventure and the players are supposed to come up with an X-shaped plan for solving problem Y, and you don't really have to push them, but after a bit of discussion someone goes "why don't we try X?"
To me that feels less like "we're on rails" and more like "if you give this cart a nudge, it's likely gonna roll downhill". *Writing* a plot so that the players will themselves come up with the "expected" plan (or something sufficiently close to it) is a sign of good craftsmanship.
My take as a GM who doesn't run APs is that the players are allowed to pick which hooks they want to bite on. I'll scatter some hooks over a session or two if things seem to be getting stuck, but otherwise let the players tell me what they want to do and work with that.
My session notes are usually vague outlines of plot points, some NPC names, and a few encounters. Dungeons and end-of-chapter battles get more prep. The in-between sessions are improv and very player-driven. It works for my group, who are happy to grab the spotlight, but I know other tables lean more on the GM to move things along.
| 1 person marked this as a favorite. |
My take as a GM who doesn't run APs is that the players are allowed to pick which hooks they want to bite on. I'll scatter some hooks over a session or two if things seem to be getting stuck, but otherwise let the players tell me what they want to do and work with that.
My session notes are usually vague outlines of plot points, some NPC names, and a few encounters. Dungeons and end-of-chapter battles get more prep. The in-between sessions are improv and very player-driven. It works for my group, who are happy to grab the spotlight, but I know other tables lean more on the GM to move things along.
That's awesome, but not everyone has got the skills to improvise so much, or the free time to prepare something new session by session.
Also, improvising is easier to do if you are playing pen and paper, much less when you are using Foundry or similar tools with maps and other details. The difference with actually well prepared scenes will be glaring.
My take as a GM who doesn't run APs is that the players are allowed to pick which hooks they want to bite on. I'll scatter some hooks over a session or two if things seem to be getting stuck, but otherwise let the players tell me what they want to do and work with that.
My session notes are usually vague outlines of plot points, some NPC names, and a few encounters. Dungeons and end-of-chapter battles get more prep. The in-between sessions are improv and very player-driven. It works for my group, who are happy to grab the spotlight, but I know other tables lean more on the GM to move things along.
That's awesome, but not everyone has got the skills to improvise so much, or the free time to prepare something new session by session.
Also, improvising is easier to do if you are playing pen and paper, much less when you are using Foundry or similar tools with maps and other details. The difference with actually well prepared scenes will be glaring.
This is very true. My ADHD is a superpower when it comes to improv storytelling, but a liability when it comes to planning and making detailed notes.
My style is also part of why I can't get into using VTTs.
| 2 people marked this as a favorite. |
I’ll try to keep this post shortYou FAILED...but you failed forward. ;-)
I only promised to try
| 6 people marked this as a favorite. |
I don't know that I care too much for the idea that PCs can just fail their way to victory.
That's not failure. That's coddling. That's not an adventure with trials. That's watching a movie until you get to the end.
I do agree that something should be done about those time wasting failed just to repeat situations. Unless there's a time constraint of some kind, they're just not interesting.
You don't understand what failing forward means. Progressing the plot does not mean it's in favour of the players, and success at a cost quite literally means there is a cost for your success. It's not coddling, it's avoiding boring time wasting that the mechanics of the game causes.
There's this idea of not asking for a check if the only consequence is that you need to try again, or take some minor consequence that doesn't really matter. The take 20 or take 10 rule basically just codifies this in the rules. There's no reason to roll to pick a lock if the only consequence is that you just try again. There's no reason to get people to roll to climb something if the only real consequence is that they fall in the worst case and take like 10 damage (which they could likely just heal with no issue). Not to mention how boring these binary challenges are.
| 2 people marked this as a favorite. |
The point being ignored here is that players don’t know if a check is going to matter or not until the GM makes that clear. If every time they roll, no matter what, something bad happens on a failure, the game has changed and players tactics will too. I think there is a range of player responses to that, some good, some bad. And probably most dependent on the situation. Sometimes it feel like people in this thread think the consequences of failure need to be immediately felt by the players and their characters in order for a dice roll to matter. I think that can be unnecessarily burdensome to the GM and to everyone’s enjoyment of the game. As a player, if you assume every roll matters and trust your GM to reveal how eventually in things that happen at the table, I think you can have a lot more fun than if you pressure the GM to reveal wether the roll mattered or not immediately.
Now, that said:
I agree that there are many times it is unnecessary to make your players roll as a GM. Like if the dungeon has basically been cleared of enemies and there is no time constraints, you might not even make your party go back to every single room they passed up and just say “with the remaining 4 hours of the day, you sweep the dungeon and find xyz.” That is totally fine and can save a lot of game time. You want your players to have enough treasure to move on and the remaining challenges might be so trivial and unimportant to the story they don’t feel necessary. Maybe you are using milestone leveling and don’t need to worry about it, or maybe you are using xp and you bundle that xp into a story award along with the loot. Alternatively, some GMs don’t want to give anything away and the players enjoy hearing all the room descriptions and figuring out for themselves the dungeon ecology. Then it’s a couple more hours of prepared play time. It is all about learning your table and what makes playing the game fun. None of these are the wrong way to play, so, as a GM, you don’t actually need to make every roll matter in a tangible, known way, even to yourself, to make every roll feel like it could matter in a way that eventually pays off.
The point being ignored here is that players don’t know if a check is going to matter or not until the GM makes that clear. If every time they roll, no matter what, something bad happens on a failure, the game has changed and players tactics will too. I think there is a range of player responses to that, some good, some bad. And probably most dependent on the situation. Sometimes it feel like people in this thread think the consequences of failure need to be immediately felt by the players and their characters in order for a dice roll to matter. I think that can be unnecessarily burdensome to the GM and to everyone’s enjoyment of the game. As a player, if you assume every roll matters and trust your GM to reveal how eventually in things that happen at the table, I think you can have a lot more fun than if you pressure the GM to reveal wether the roll mattered or not immediately.
Now, that said:
I agree that there are many times it is unnecessary to make your players roll as a GM. Like if the dungeon has basically been cleared of enemies and there is no time constraints, you might not even make your party go back to every single room they passed up and just say “with the remaining 4 hours of the day, you sweep the dungeon and find xyz.” That is totally fine and can save a lot of game time. You want your players to have enough treasure to move on and the remaining challenges might be so trivial and unimportant to the story they don’t feel necessary. Maybe you are using milestone leveling and don’t need to worry about it, or maybe you are using xp and you bundle that xp into a story award along with the loot. Alternatively, some GMs don’t want to give anything away and the players enjoy hearing all the room descriptions and figuring out for themselves the dungeon ecology. Then it’s a couple more hours of prepared play time. It is all about learning your table and what makes playing the game fun. None of these are the wrong way to play, so, as a GM, you don’t actually need to make every roll matter in a...
Sure but in these scenarios you can keep it to one roll, having a failure say cause to the lock to open, but someone that you aren't aware of notices. Another idea is to be open but vague with the consequences, say that they get the lock open but they think they could've tripped some type of detection in the lock, or they swear they feel like someone saw them. It may be a little meta, but that's fine.
Here is perhaps a more clear example of what I am talking about.
The party comes to a door. They approach it relatively quietly and one of the party goes up, inspects the door and sees that it is locked. Someone decides they are going to begin picking the lock right a way. They roll once, an 8 on the dice and fail. The GM says nothing happens. The party gets into a debate about whether to proceed to try to pick the lock, or continue and explore the hall. They make no effort to suggest the party does this debate quietly, and the GM knows that there are some creatures on the other side of the door that are being quiet and attentive, so they might have not noticed one attempt to pick the lock, but with the conversation afterwards they take a couple of rounds to have one person sneak up to the door to try to listen more carefully for a while, one person go and get some allies that will take some rounds (lets decide 8) to get prepared and get to the room, and then, if it still sounds like there are people on the other side of the door, the caster of the group plans on casting bless at the same time someone unlocks and opens the door from the other side (this would constitute round 1, so basically the encounter would begin with the door locked and the party doing whatever they were doing right before). So the GM decides to be kind and say that the party probably has 10 rounds from the end of their conversation about what to do before the enemies will open the door.
To the players, if they don't have someone carefully listening at the door, it might feel like nothing happened as a result of having tried to pick the lock, especially if they decide to leave before the enemies are ready to open the door. Or it might seem that way for the 8 rolls it might take to get the 4 successes necessary to unlock the average lock on the door, only to immediately not feel that way when the door opens right as the reinforcements arrive, because it took 8 rounds to open the door, and the GM decided at the outset that is how long it would take the reinforcements to get there. Maybe the party had someone listening at the door and hears someone approach or hears the movement of preparation inside, maybe they don't. There are just way to many variables to pretend like that can fit into a one size fits all box of how to handle every situation where the party encounters a locked door, and reducing the entire thing down to one roll limits the potential variations on what could happen down to a maximum of 4, when, playing by the standard rules, a party could have two people good at picking locks give it a try and get the door in 4 or less rounds, catching the enemies with one less person in the room at the start, but with reinforcements on the way, or maybe even do it quietly enough to not have raised an alarm at all, or to try a couple of times and then notice someone on the other side of the door and decide to move back and set up a trap. The GM is doing a fair bit on the fly with this, but it is in response to individual actions by the party so they can wait a few rounds of a few failures and then move things along at a pace that feels right for the table. Having to have all of that figured out before the first roll, and then ready to arbitrate which happens on each of the 4 results is making the GM have to do a lot more work at once, and prepared to implement right after the first roll. They don't have to do that if they are playing by the rules as even a critical success on the first check is only going to get the lock half way open.
Here is perhaps a more clear example of what I am talking about.
The party comes to a door. They approach it relatively quietly and one of the party goes up, inspects the door and sees that it is locked. Someone decides they are going to begin picking the lock right a way. They roll once, an 8 on the dice and fail. The GM says nothing happens. The party gets into a debate about whether to proceed to try to pick the lock, or continue and explore the hall. They make no effort to suggest the party does this debate quietly, and the GM knows that there are some creatures on the other side of the door that are being quiet and attentive, so they might have not noticed one attempt to pick the lock, but with the conversation afterwards they take a couple of rounds to have one person sneak up to the door to try to listen more carefully for a while, one person go and get some allies that will take some rounds (lets decide 8) to get prepared and get to the room, and then, if it still sounds like there are people on the other side of the door, the caster of the group plans on casting bless at the same time someone unlocks and opens the door from the other side (this would constitute round 1, so basically the encounter would begin with the door locked and the party doing whatever they were doing right before). So the GM decides to be kind and say that the party probably has 10 rounds from the end of their conversation about what to do before the enemies will open the door.
To the players, if they don't have someone carefully listening at the door, it might feel like nothing happened as a result of having tried to pick the lock, especially if they decide to leave before the enemies are ready to open the door. Or it might seem that way for the 8 rolls it might take to get the 4 successes necessary to unlock the average lock on the door, only to immediately not feel that way when the door opens right as the reinforcements arrive, because it took 8 rounds to open the door, and the GM decided at the outset that is how long...
I think this would be an example of a dynamic situation with severe time pressure where having multiple rolls would not be a waste of time necessarily, and the consequences of failure are more than just taking extra time, due to the time pressure.
The thing about my scenario is that, in a traditional dungeon, the adventure writer nor the GM can really know whether that specific locked door is going to the hinge around which a key dynamic encounter swings open or not, until the party chooses to make it so. That’s the nature of the big mapped out dungeons with multiple points of egress. The players end up making the choices that unlock how things dynamically fit together.
There are other styles of adventure writing, such as having a set list of encounters that happen in order and you either ignore intersections as de idiom points, or make it super obvious where each path leads at the point the party is making that choice. These styles of adventure writing allow the GM to prepare each set encounter much further in advance because they will always know where the party enters the encounter and what is going to happen after, but they can also feel like players are loosing a lot of the agency that defines how TTRPGs are different from computer games or choose your own adventure books.
Thus in the traditional dungeon model, it can be counter to all players )including the GM) to try to overly consequence every single die roll, because die rolls, especially ones that serve as hinges for narrative beats outside of encounter mode, don’t happen in isolation of the adventure or dungeon. It can be really fun when everyone knows the next roll is going to be really important. It can be draining for everyone when every single roll has to be weighted that way, and it is especially burdensome on the GM to require that they try to make every roll feel that way to the players because building tension takes time and when everything about it is too transparent to the players, the GM loses a lever that they can control in making sure the tension is ramping up at the right rate for their party.
I have said before that time always is an important and limited resource, and that is something akin to a religious belief for me, especially in the real world. When it stops feeling that way is usually a result of mismatched scale to circumstance, but circumstances change often, so you have to be very careful when setting scales for “loss of time as resource management” in games. That is why the lock thing keeps coming up here, and why additional tools for GM to make locks interesting in encounter modes beyond combat sounds like a useful tool to think about and create, but not to try to force over locks as a general rule.
| 4 people marked this as a favorite. |
I'm prepping SoT book 5 right now as a GM and also had an encounter last night as a player in Spore War that really hit this home: PF2 has a lot of "you fail to do a thing, nothing interesting happens, try again" mechanics and situations in it.
The spoiler-free version of the ritual is "you must succeed at this ritual to advance the plot." Failure has some minor negative consequences but there's no real time pressure at this point, so you can just treat it or wait it out. Running this RAW, PCs will continue attempting it until they succeed. Given how stacked ritual DCs are against PCs when secondary casters are involved, "try again" is the most likely outcome.
TBH, this is less about 'Fail Forward' and more about bad adventure design.
This ritual simply shouldn't be a dice roll to begin with, it should be an emergent result of the PCs' actions up to that point in the game. If the PCs did X many things, they succeed; if they did less than that, they do the ritual but there are Consequences (or they need to go and complete X many things, either-or).
Yes, this is a case of the writer following the standard DC numbers without considering how those numbers affect the adventure.
When the player characters finally travel, negotiate, and fight their way to the Doorway to the Red Star (which is the name of the module, too), simply walking through the doorway would be anticlimactic. So the writer added a ritual to open the doorway. That is reasonable adventure design.
The mistake was making the ritual short and difficult. The ritual rules, which are the same in both the Remastered Player Core and the pre-Remaster Pathfinder Core Rulebook, say:
Checks
At the ritual's culmination, you must attempt the skill check listed in the Primary Check entry to determine the ritual's outcome. Primary checks usually have a very hard DC for a level that's twice the ritual's spell rank. As with other downtime activities, fortune and misfortune effects can't modify your checks for the ritual, nor can bonuses or penalties that aren't active throughout the process.The GM can adjust the DCs of rituals, add or change primary or secondary checks, or even waive requirements to fit specific circumstances. For example, performing a ritual in a location where ley lines converge on the night of a new moon might make a normally difficult ritual drastically easier.
The writer Michael Sayre decided that the ritual was 8th rank, probably because the party had just reached 16th level, and followed the DC suggested in the rulebook without applying any of the GM-discretion adjustments allowed by those rules. Sixteenth-level DC is 35 and a +5 "Very Hard" adjustment made a DC 40 Arcana or Nature (master) primary check.
Those DC numbers in the rulebook were chosen because rituals are powerful and therefore need to be difficult. On the other hand, succeeding at this ritual is necessary to the plot and the ritual has no practical value anywhere else, so it really should have been guaranteed. The ritual would have still been impressive at 6th rank (DC 28) and merely hard (+2) for DC 30. At least one PC in Strength of Thousands would have gone to legendary in Arcana or Nature for a +29 skill bonus. Only a natural 1 would have failed the Primary Check, and natural 1s are considered legitimate bad luck that has to be worked around. It would have replaced the meaningless (due to retries) tension of "We have only a 50% chance of success" with the smugness of "We are Magaambya mages. Of course, we can handle this."