Deadmanwalking's Main Problem With PF2


General Discussion

51 to 100 of 227 << first < prev | 1 | 2 | 3 | 4 | 5 | next > last >>

1 person marked this as a favorite.
Mark Seifter wrote:
It was actually when we removed glancing blows that suddenly the very lowest level monsters started becoming whiff-fests that usually did nothing even if it was supposed to be a Hard or Severe encounter, which led to lowering damage and upping accuracy a bit for those.

Can you elaborate on this?

According to the bestiary a L16 is a "low threat minion" and minimal creature worth counting for a L20 party. So "the very lowest monsters" depends greatly on who the monster is fighting.

Can you roughly quantify (ballpark) the degree of changes?

If we assume that a L3 monster is a very lowest for a L7 party, how did changing that monster's stats impact its interaction with L1 and L3 characters?

Thanks


Bartram wrote:
Mark Seifter wrote:


PC best possible optimized saving throw

My concern is that these "optimized values include receiving buffs from other characters or conditional situations (flat footed etc)

A 100% optimized character should be hitting around 80% in a vacuum. Before external conditions are applied. No buffs from the party wizard, the opponent isn't flat footed from being flanked etc.

A maximally skilled and geared (for his level) fighter that walks up and swings at an equal level opponent should hit about 75% of the time. Not a with a flanking buddy, not because the bard is singing, not because the wizard cast haste on him or slow on the opponent. Just using that characters pure skill. Otherwise it just doesn't _feel_ fun. The math can be adjusted to make any hit rate _work_ what is more important is to make the character's _feel_ competent to play.

You never know what party composition is going to be or what buffs you are going to have, and those shouldn't be factored in to the baseline. If a group of players wants to make an all fighter party, they shouldn't be incapable of feeling heroic because they don't have 3 other sets of class features to bring them up to the "baseline."

When you can't guarantee that a certain condition is going to be present, don't include it in your base assumptions.

I have full confidence that the dev team will be able to tweak the numbers and achieve the goals they set out. I just hope they are the same goals that I have. =)

Edited in because I had an idea and didn't want to double post:

Easy fix (which im sure has some trickle down effects I haven't considered and will need some tweaking): Move the baseline assumption of accuracy from 50% to 75% (by adjusting AC/DCs) and increase NPC HP by 25%. Same length of combat, but the PCs _feel_ more heroic because they hit more often, even though each round they are still doing the same percentage of an NPCs health in damage.

For my part, I want this to be true even if flat footed and circumstance bonuses need to be removed from the game.

For instance, Flat footed could instead just impart a scaling weakness to weapon damage and the bard’s Inspire Courage could add additional damage dice (though this adjustment may have some effects on how such abilities might interact with spells).


1 person marked this as a favorite.
Pathfinder Adventure Path, Lost Omens, Rulebook Subscriber

The other thought I have is that it is unfortunate that the gaming convention is that most skills come down to one die roll. Having a 50% chance of success each check means that most skills with no consequence for failure are going to be successfully completed in one round or two. Balancing skills that can be attempted over and over with those with immediate consequence for failure is difficult to do.


1 person marked this as a favorite.
Pathfinder Adventure Path, Lost Omens, Rulebook Subscriber
BryonD wrote:
Mark Seifter wrote:
It was actually when we removed glancing blows that suddenly the very lowest level monsters started becoming whiff-fests that usually did nothing even if it was supposed to be a Hard or Severe encounter, which led to lowering damage and upping accuracy a bit for those.

Can you elaborate on this?

According to the bestiary a L16 is a "low threat minion" and minimal creature worth counting for a L20 party. So "the very lowest monsters" depends greatly on who the monster is fighting.

Can you roughly quantify (ballpark) the degree of changes?

If we assume that a L3 monster is a very lowest for a L7 party, how did changing that monster's stats impact its interaction with L1 and L3 characters?

Thanks

I think this quote is specifically addressing 0th level monsters and why their numbers are skewed a little differently than the rest of the monsters (ie their accuracy is higher and their damage is lower.)


2 people marked this as a favorite.
Unicore wrote:
Bartram wrote:


A maximally skilled and geared (for his level) fighter that walks up and swings at an equal level opponent should hit about 75% of the time. Not a with a flanking buddy, not because the bard is singing, not because the wizard cast haste on him or slow on the opponent. Just using that characters pure skill. Otherwise it just doesn't _feel_ fun. The math can be adjusted to make any hit rate _work_ what is more important is to make the character's _feel_ competent to play.

I really think that the new action economy makes numbers like this very difficult to use. 75% on your first attack means 50% on your second and still a 25% on your third attack. (assuming the fighter is not using agile weapons or other feats to get lower penalties) I don't think first level characters need to have a 50% chance of hitting 2 times or more. I especially don't know how you balance the idea that every level 1 enemy will probably hit at least once a round and often hit 2 or 3 times. Parties that don't really optimize tactics that minimize # of enemy attacks and maximize their own are going to be in a lot of trouble.

I think something close, albeit a bit less, than this might make sense for a fighter, say 65%/40%/15%, because fighters are supposed to be the class that's all about hitting anyway. For non-fighters optimized for combat, I'm guessing something closer to 55%/30%/5% makes more sense. Note, even with agile, these numbers mean the second attack is still only critting on a natural 20 even for the fighter. Spells and circumstances should be able to up these numbers, but we also shouldn't be baselining on fighter and assuming that's the "only" way to be good in combat.

Also, I'm assuming this is against some average enemy AC. Against an armor specialist, dropping these numbers by 10-15% make sense. I'd honestly prefer to err on the side of hitting overly often rather than missing overly often...


19 people marked this as a favorite.
Pathfinder Starfinder Society Subscriber

I'm on board with DMWs analysis above; in the first week after the CRB and Bestiary dropped I did the calcs myself and saw as much. I think what really needs to be emphasized is the 'feel-bad' aspect of the current balance fulcrum.

"Silver-tongue Steve" the lying focused rogue, who by 7th level has shaped their entire character build around being the best gosh-darn liar ever, has a (+7+4+2+2) = +15 to Deception. He wants to lie to a very unintelligent (-2), not very wise (+1) harpy. That harpy has a +12 on perception and is a Level 5 creature. This is probably a chump, a goon serving the Level 9 boss of the area Steve is trying to infiltrate. And yet, despite his focus on this specialty, Steve has a 30% chance that his lie is unraveled!

This is a feel-bad moment. A PC should be challenged in the area of their focus by what is an appropriate matchup: The player should sweat when their super-liar rogue is up against the APL+2 dragon with all its great wisdom and cunning, not against the dragon's barely competent goons!

Liberty's Edge

17 people marked this as a favorite.
Cellion wrote:
I'm on board with DMWs analysis above; in the first week after the CRB and Bestiary dropped I did the calcs myself and saw as much. I think what really needs to be emphasized is the 'feel-bad' aspect of the current balance fulcrum.

Absolutely. This is probably the most important bit. It's not primarily that the math doesn't work. It works (though not without issues). The problem is that it feels bad.

It feels even worse if you're not optimized. I mean, imagine Steve was made by someone who didn't know the system well and only had a 14 Cha. Then his bonus is only a +13 and his odds of tricking the (Level-2) Harpy are only 60%. Vs. on-level foes he's pulling a coin flip despite having invested his only Master level skill and an Item into Deception.


2 people marked this as a favorite.

You know what might be an interesting bandaid to this situation? You could introduce a luck mechanic that allows for rerolls (or forced enemy rerolls) a certain number of times per session that is tied to proficiency. Experts in bluff could reroll a failed bluff checks twice a day and master spellcasters could force an enemy to reroll succeeded saves 3 times a day/session.

That way, specialists in certain tasks can suffer a resource drain for a single unlucky roll rather than feeling like total chumps.


Pathfinder Adventure Path, Lost Omens, Rulebook Subscriber

What if situations like the lying your way past a enemy were not blown after one failure? or What if most skill challenges (where a character is engaged in one skill for the whole round) gave you a way to make two or three rolls in a turn and take the best one? I think that would allow the numbers to sit close math-wise, without giving away critical successes, which is what happens at the 75-90% success rate that players are used to in PF1 for a single check.


1 person marked this as a favorite.
Pathfinder Rulebook Subscriber

So what is the appropriate percentage of success for it to feel good? Keeping in mind the differences between someone who 100% optimizes vs someone who dabbles? Should 100% optimization allow 95% success?

Excaliburproxy wrote:

You know what might be an interesting bandaid to this situation? You could introduce a luck mechanic that allows for rerolls (or forced enemy rerolls) a certain number of times per session that is tied to proficiency. Experts in bluff could reroll a failed bluff checks twice a day and master spellcasters could force an enemy to reroll succeeded saves 3 times a day/session.

That way, specialists in certain tasks can suffer a resource drain for a single unlucky roll rather than feeling like total chumps.

You could just be more generous with the hero point rules to accomplish this.


5 people marked this as a favorite.
Captain Morgan wrote:

So what is the appropriate percentage of success for it to feel good? Keeping in mind the differences between someone who 100% optimizes vs someone who dabbles? Should 100% optimization allow 95% success?

Excaliburproxy wrote:

You know what might be an interesting bandaid to this situation? You could introduce a luck mechanic that allows for rerolls (or forced enemy rerolls) a certain number of times per session that is tied to proficiency. Experts in bluff could reroll a failed bluff checks twice a day and master spellcasters could force an enemy to reroll succeeded saves 3 times a day/session.

That way, specialists in certain tasks can suffer a resource drain for a single unlucky roll rather than feeling like total chumps.

You could just be more generous with the hero point rules to accomplish this.

A character that has maxed stat, +skill items, highest proficiency and maybe a skill feat or two should probably expect at least 80% success against an equal level guy who hans't done any of this. Average creatures with a 12-14 stat and expert at best should probably have 50% versus each other (this includes monsters who don't specialize either).

Nothing weird about sneaking past noisy Orc sentries with great odds, but Elf sentries might make it tough for the party Rogue... But he is still an exceptional PC with a 18 who was trained the skill for years, so even 60% wouldn't be that generous.

The DC table should likewise be changed. First, most challenges should be in the Low-High range, but almost nothing level-appropiate is lower than High, if not Severe. Currently the lv1 row says a Trivial challenge is DC10, which a specialist (+5 on the roll) can do 75% of the time. I think "Trivial" Should be almost impossible to fail, like 90%. The "Sneaking Past Orcs" example from before that I said should have 80%+ would be something one would consider "Easy" (65% on that table), but apparently 80% is too good even for Trivial. Actual lv1 Orc warriors have +3 perception so it's even lower than 65%. I think that table is off by at least 2 in the early columns.


1 person marked this as a favorite.
Pathfinder Rulebook Subscriber
Unicore wrote:
What if situations like the lying your way past a enemy were not blown after one failure? or What if most skill challenges (where a character is engaged in one skill for the whole round) gave you a way to make two or three rolls in a turn and take the best one? I think that would allow the numbers to sit close math-wise, without giving away critical successes, which is what happens at the 75-90% success rate that players are used to in PF1 for a single check.

I think there's some ways that the rules can already support this, but they aren't spelled out super clearly. For example, if you are sneaking and have total cover, failing a stealth check shouldn't make you seen. It might make you sensed, but the enemies then need to roll perception against your stealth DC to try and find you. They may be on guard but the noise they heard in the night could have been anything and you have a chance to respond.

Also, circling back to an earlier conversation, I'm pretty sure I read in these rules that a GM can apply circumstance bonuses or penalties as they see fit. This is explicitly called out to let you modify the DC for Deception rolls, but you could also apply it for relevant stealth conditions like distance.

Dunno, just spit balling here.


Captain Morgan wrote:

So what is the appropriate percentage of success for it to feel good? Keeping in mind the differences between someone who 100% optimizes vs someone who dabbles? Should 100% optimization allow 95% success?

Excaliburproxy wrote:

You know what might be an interesting bandaid to this situation? You could introduce a luck mechanic that allows for rerolls (or forced enemy rerolls) a certain number of times per session that is tied to proficiency. Experts in bluff could reroll a failed bluff checks twice a day and master spellcasters could force an enemy to reroll succeeded saves 3 times a day/session.

That way, specialists in certain tasks can suffer a resource drain for a single unlucky roll rather than feeling like total chumps.

You could just be more generous with the hero point rules to accomplish this.

Well, I would prefer it if it were actually tied to skill proficiency in some way but expanding hero points would be less bookkeeping intensive. Perhaps you could achieve a combined effect by having hero point rerolls double the bonus you receive from proficiency.

To answer what I'd say "feels good" for on level challenges, I think it should be about matchups:

Optimal Character:
"Hard" challenge/Specialist opponent: 55% success rate
Average challenge/Non-Specialist opponent: 75% success rate
Easy challenge/Incompetent opponent: 95% success rate

Competent Character (Character using skills from a non-main stat or character merely trained in a skill of their main stat)
"Hard" challenge/Specialist opponent: 45%/40% success rate
Average challenge/Non-Specialist opponent: 65%/60% success rate
Easy challenge/Incompetent opponent: 85%/80% success rate

I expect someone to then say "but doesn't that get out of hand with various bonuses"?
Not if you reduce the bonuses people impart from aid another and/or further limit what bonuses stack.


4 people marked this as a favorite.

At the same time as making sure players who don't optimize stand a decent chance at success, I feel it's equally important to ensure players who do optimize fully don't trivialize challenges and usurp the spotlight too much.

For example, currently optimized Fighters have a 60-65% chance to hit vs. equal level enemies on their first attack. If we improve that chance to 70-75% or 75-80% to bring the unoptimized characters up to a decent hit chance, do we risk Fighters trivializing encounters by hitting too easily and critting too much?

What success rate do people want for decently optimized characters? Should those who are Expert in a skill but didn't invest into getting a 20 in the skill's corresponding stat succeed 55% of the time? 60% of the time?

The current system can have up to a 18 point difference between the most optimized skill check (+3 Legendary proficiency, +7 from stat mod, +5 item bonus vs. -2 untrained proficiency and -1 stat mod).

The difference between an optimized skill check using a main stat and another stat is usually 2 points (24 main stat vs. 20 other stat), a skill that doesn't benefit from an item bonus is 2-5 points behind a skill where you have a magic item boosting it.

There's still a large gap between unoptimized and optimized, and bringing the former up can make the latter too good. This is something that we must take into account when tweaking the math of the game.

Let's say the end goal is to make players feel better about their characters by increasing their expected success rate vs. an equal level opponent and give them a fighting chance vs. higher level opponents (up to 4 levels higher, since the bestiary assumes you can take on such challenges).

Say we want a decently optimized character (16-20 stat, Expert in a skill or attack, with an appropriate item bonus to the task the character is attempting) to have a 65% chance to succeed vs. an equal level opponent. And say we want a level+4 challenge to be severely difficult but not crippling to attempt, so we assign a 45% success rate there. If we did all this, what happens to our optimized character?

Vs. an equal level opponent, that same optimized character will have +1/+4 from stat, +1/+2 from proficiency, or 75-95% chance to succeed vs. an equal level opponent. Is that too high? I believe a 95% chance to succeed is too much.

Another problem that arises comes from PF2's assumption that PCs will always have an appropriate item bonus. The level+4 challenge won't just have +4 to its stats, it will have +5, maybe +6 because the game assumes you're getting an increase in item bonus, proficiency bonus or stat bonus from leveling. So realistically our 45% assumption is closer to 30-35%. In the actual game, it's even lower, along the 20-25% range. I dunno about you guys, but trying to fight a dragon that's hitting everyone easily and that can only get hit 25% of the time is not fun.

So how do improve fun levels (aka success rate) while preventing auto-hits and auto-misses? Well, we could get rid of where the differences are coming from. The biggest contributor to difference between an optimized and unoptimized character is Stat modifier, followed by Item bonus. The game assumes you'll get to +5 item bonus eventually in Attack bonus, AC and Skill checks, and scales monsters appropriately. The game also assumes you'll have a 24 in the stat used for attacks and skill checks, but each PC can only have 1 stat be this high, their main stat, and it certainly doesn't apply to most skills.

If we get rid of ALL item bonuses (or bake them into the leveling like proficiency) and reduce the maximum stat modifier from +7 to +6 (by getting rid of Potent items), then we can get to a pretty good spot where the difference between a decently optimized character and a fully optimized character is not big enough to skew the math and monster stats can be adjusted to allow a decently optimized character to have a decent success rate without a fully optimized character succeeding too much as to trivialize encounters.

But doing this would bring about complaints of "Everyone is too samey" and "An optimized character should succeed ALL the time". In the end, you can't please everybody, but with respect to this thread, I believe that reducing Item bonuses and maximum stat modifier is the first step to adjusting the monster math to make PCs succeed more in general and still maintain the game's balance.


3 people marked this as a favorite.
Pathfinder Rulebook Subscriber

It also occurs to me that sneaking and lying, the two skills we are mostly focused on here, both lack critical failure/success conditions. I wonder if that is specifically because they oppose perception DCs, and if maybe the general skill numbers don't need an overhaul so much as perception scores for NPCs/Monsters just need a big nerf?

Either that, or maybe add some critical success/failure conditions? So using Unicore's example, you fail a stealth check, you don't become seen or even sensed, and the enemy is just suspicious someone might be out there and may investigate. Which gives you an opportunity to react and try to fix it. But if you critically fail, you have blown it and the enemy knows exactly where you are.

For Deception, failure could just mean the person is skeptical but you may still persuade them. But a critical failure means they KNOW you are lying and will treat you accordingly. There's a big difference between being skeptical and flat out disbelieving someone, after all. Not sure how the +4 circumstance bonus they get against you works out there. Maybe it only applies on the critical failure? Maybe it is a +1 or +2 on a failure? Or maybe failure is +4 but critical failure and they just don't trust anything you say and don't have to roll? I dunno.

Liberty's Edge

6 people marked this as a favorite.
Captain Morgan wrote:
So what is the appropriate percentage of success for it to feel good? Keeping in mind the differences between someone who 100% optimizes vs someone who dabbles? Should 100% optimization allow 95% success?

At 13th level, an absolutely optimal character with level-appropriate items has a +23 (13 level, 5 Ability, 3 Item, 2 Proficiency). A High DC at that level is 32, meaning they still only have a 60% chance of success at that. That's probably not sufficient. It feels like a lot of investment for such a non-ideal percentage.

As levels rise past that, the odds get better (peaking at 75% at 20th level). A 75% is sufficient (indeed, a 70% is probably sufficient on high DC stuff)...but you don't get anywhere close to that for most of your career.

Equally importantly, someone with a minimal investment (a stat of 12 to start, rising every five levels, plus Trained and maybe a non-magic Master Quality Item eventually) should probably have a decent shot at low Skill DCs if not high. At 4th, such a person has a 45% chance of low DCs. At 13th, they've hit 50%, but by 20th, it's back down to 45%. Which is again a bit low.

Dropping the Low DCs by two or three and the high ones by about two pre-14th level (or thereabouts...a few low level Trivial ones would also need to be adjusted) seems like it'd be sufficient to me, but would need to be tested, of course.

This is aside from the monster skills and monster Perception issues noted previously.


Unicore wrote:
Bartram wrote:


A maximally skilled and geared (for his level) fighter that walks up and swings at an equal level opponent should hit about 75% of the time. Not a with a flanking buddy, not because the bard is singing, not because the wizard cast haste on him or slow on the opponent. Just using that characters pure skill. Otherwise it just doesn't _feel_ fun. The math can be adjusted to make any hit rate _work_ what is more important is to make the character's _feel_ competent to play.
I really think that the new action economy makes numbers like this very difficult to use. 75% on your first attack means 50% on your second and still a 25% on your third attack. (assuming the fighter is not using agile weapons or other feats to get lower penalties) I don't think first level characters need to have a 50% chance of hitting 2 times or more. I especially don't know how you balance the idea that every level 1 enemy will probably hit at least once a round and often hit 2 or 3 times. Parties that don't really optimize tactics that minimize # of enemy attacks and maximize their own are going to be in a lot of trouble.

If this optimized Fighter gets, say, an additional +2 from bonuses, and we count criticals as double, he's going to get the equivalent of 2.3 hits per round on average. (First attack 35% critical hit, 50% regular hit, second attack 10% critical, 50% regular, third attack 5% critical, 30% regular hit.)

That seems a bit much, though we could theoretically balance this out by doubling HP or reducing damage per hit by 50% or so.


4 people marked this as a favorite.
Pramxnim wrote:
I believe that reducing Item bonuses and maximum stat modifier is the first step to adjusting the monster math to make PCs succeed more in general and still maintain the game's balance.

I believe that the first step is fixing the tone of skill checks so that trivial activities are passed on a very low roll like 1, low could be 5, high (really normal) could be 10, severe could be 15, and extreme could be 20. This opens up the whole d20 instead of assuming a 50% failure rate.

+1 skill per level for PCs and Monsters may be another problem here. It doesn't play well with perception and stealth. Why is a bobcat so much worse at perception and sneak than a tiger? Small creatures tend to be low level creatures, but there is an intuitive understanding that small creatures usually match or exceed large creatures in perception and stealth. This helps keep a small creature alive when they encounter a large one. This feeling may be fueling unusual bonuses to low level creatures.

I hope that this gets worked through and an elegant solution is found ( I believe there is at least one). And it gets implemented through the whole rules set.


Pathfinder Rulebook Subscriber
Deadmanwalking wrote:
Captain Morgan wrote:
So what is the appropriate percentage of success for it to feel good? Keeping in mind the differences between someone who 100% optimizes vs someone who dabbles? Should 100% optimization allow 95% success?

At 13th level, an absolutely optimal character with level-appropriate items has a +23 (13 level, 5 Ability, 3 Item, 2 Proficiency). A High DC at that level is 32, meaning they still only have a 60% chance of success at that. That's probably not sufficient. It feels like a lot of investment for such a non-ideal percentage.

As levels rise past that, the odds get better (peaking at 75% at 20th level). A 75% is sufficient (indeed, a 70% is probably sufficient on high DC stuff)...but you don't get anywhere close to that for most of your career.

Equally importantly, someone with a minimal investment (a stat of 12 to start, rising every five levels, plus Trained and maybe a non-magic Master Quality Item eventually) should probably have a decent shot at low Skill DCs if not high. At 4th, such a person has a 45% chance of low DCs. At 13th, they've hit 50%, but by 20th, it's back down to 45%. Which is again a bit low.

Dropping the Low DCs by two or three and the high ones by about two pre-14th level (or thereabouts...a few low level Trivial ones would also need to be adjusted) seems like it'd be sufficient to me, but would need to be tested, of course.

This is aside from the monster skills and monster Perception issues noted previously.

So with 100% optimization, 70% odds of success on a difficult check seems acceptable? (Provided it kicks in at lower levels?)


Dasrak wrote:
Xenocrat wrote:
But this extends beyond combat. You only get once chance to cast a meaningful spell per round (and waste a scarce resource if they save). You only get one chance to deceive someone or sneak past them or to learn a spell (until next level) or...

I'd agree, and would go so far as to say these are where the issues are most acute. It makes spellcasters extremely swingy since it means their entire combat contribution could be riding on a single roll; flub that roll, and the party is effectively down a man. Skills are even worse, though. 1st level characters basically are incompetent at everything if they don't have at least a 16 in the associated stat. I want to focus on stealth in particular, because the way it's set up it's functionally impossible to succeed in any practical situation. Let's compare two equivalent 1st level characters for PF1 and PF2.

We'll take a character with 14 dexterity, 1 rank in stealth, a trait that gives a +1 bonus and class skill, for a total of +7 stealth. We'll take a group of three 1st level mook enemies with +0 perception. Because they aren't aware of your presence, they're just taking 10 on perception. If you keep 20 feet away (-2 penalty to their perception check) then you succeed even on a roll of 1. You can consistently sneak past these mooks, provided you have cover or concealment.

Now let's look at PF2 stealth. 1st level character, trained in stealth, and 14 dexterity. You have +3 to stealth in total. The 1st level mooks have +1 to perception, and each of them rolls separately (take 10 isn't a thing anymore). There are no distance penalties, either. So you need to beat all three of their rolls. The chance of doing that is... 37.5%
Ouch, this went from guaranteed success in PF1 down to a longshot. But it actually gets worse: you need to roll again every time you move. And since you're moving half your speed, that probably means you need to succeed three times per round. We're now looking at a chance of success so low that it no longer...

That is incorrect.

Only the active character rolls. So your PF2 character rolls with a +3 to beat a DC of 11 (the passive characters "take 10").


4 people marked this as a favorite.
Thorin001 wrote:
Only the active character rolls. So your PF2 character rolls with a +3 to beat a DC of 11 (the passive characters "take 10").

This has been pointed out to me. While it does help, the math is so skewed against stealth that it just moves it from "impossible" to "very unlikely".

Captain Morgan wrote:
It also occurs to me that sneaking and lying, the two skills we are mostly focused on here, both lack critical failure/success conditions. I wonder if that is specifically because they oppose perception DCs, and if maybe the general skill numbers don't need an overhaul so much as perception scores for NPCs/Monsters just need a big nerf?

Even if we presume lower numbers, it still doesn't work.

Consider a 1st level character with 18 dex and trained in stealth for a +5 check, trying to sneak past level 1 mooks with a perception DC of 11 (can't get much lower than that). That gives a 75% chance of success... on your first check. You need to make checks every time you take the sneak action, which is three times per round. That gives you only a 42% chance to succeed on the first round. In fact, you'd need to drive the perception of level 1 monsters deep in the negatives to get a reasonable chance to succeed at the 1st level. And that's with 1 PC, godsforbid a whole party!

Stealth at very minimum needs fewer rolls (even once per round is a bit excessive; once per sneak action is obscene). Bringing back the distance penalty also helps the skill better scale to different parties; noisier parties can still use stealth, they just need to allow more distance. However, the natural 1 rule probably needs to go as well, since it's very unsuitable to stealth.

The thing that makes the stealth skill different from most other skills is that it's kinda inverted. A successful stealth check doesn't mean you sneak past the enemies. It means they don't spot you. This is what creates the party conundrum; if we bluff our way into the duke's palace we need only one successful bluff check, but if we sneak our way in, everyone in the party needs to make multiple stealth checks, and any one failure means the jig is up. For stealthy actions over long periods of time, rolls stop mattering entirely because you will roll a natural 1. Because natural 1's are automatic failures in PF2, this creates a problem for stealth. With four PC's, it only takes 3 stealth checks before there's a 50/50 chance that someone has rolled a natural 1. That means that for a task calling for three stealth checks, you have a 50% chance of failure, no matter how skillful your characters are, no matter how easy the task, and no matter how well-executed your plan. I don't feel the natural 1 rule works for the stealth skill at all.


3 people marked this as a favorite.

This is a very interesting thread, the thesis of which I completely 100% agree.

As for stealth, I would like to suggest a codification of what someone suggested upthread:

When you make a Stealth check to Hide or Sneak, make a DCX (X being under intense discussion here), and change your visibility status as below.

Success: Decrease your visibility level by 1. (Seen to Concealed, Concealed to Sensed, Sensed to Unseen.)
Critical Success: Decrease your visibility level by 2.
Failure: Increase your visibility level by 1.
Critical Failure: Increase your visibility level by 2.

This mirrors a lot of other condition levels in the Playtest, and interacts with the wonderfully-streamlined Visibility rules. (Seriously, the 4-level visibility system works very very well.)

Also, this allows for nuanced multi-level reactions to Stealth. Not only can the Guard Captain have a different visibility level for the Stealthing Rogue than the Grunts, but you can actually adjudicate it and it's not binary.

If you include a requirement to Sneaking/Hiding of 'you must be at least Concealed' (this is already true, no?), that would mean a Success on Sneaking while Concealed moves you to Sensed, etc. etc.

Also, several spells (like Blur) and conditions (like dazzled) could be changed to say, e.g., Blur reducing your visibility level with everyone by 1. (Instead of 'you are concealed'.)

Etc. etc.

Liberty's Edge

7 people marked this as a favorite.
Captain Morgan wrote:
So with 100% optimization, 70% odds of success on a difficult check seems acceptable? (Provided it kicks in at lower levels?)

Yes. That's high enough to leave some leeway for people to not be absolutely optimal but still have better than a 50% chance of achieving such things. Some scope for 'pretty good', which is one thing there's a huge lack of at the moment.


6 people marked this as a favorite.

We should set some parameters, but I don't think they should be top down. PCs only get a few true optimizations due to wealth, stats, & advancement limitations, so maybe we should balance around another metric: like say somebody who's trailing just behind, but still investing regularly.

So on attack rolls for instance, a 16 Str Barb makes for a good baseline because that's the top stat for some ancestries and Barb's get few bonuses to attack, but still need to be hitting to contribute. Non-martial classes that wish to be martial have this same cap of 16, and investing that much in an off-stat should have some reward.
So even at 15th, this PC will likely have 20-22 Str & is just Expert w/ weapons, yet should still hit reasonably often.
AC, on the other hand, is really hard to advance past trained except for a few classes, so trained likely serves as the better baseline.

For many skills, a 14 Stat (going to 16, then 18) makes a good barometer for success. An expert w/ a 14 Stat before 5th should be that: an expert. Not the best expert, but reliable enough to be called one.
Arguably most DCs & gated checks should be based around Expert because otherwise party composition has major impacts at high levels, impacts that go beyond party roles.
In the same vein, PCs shouldn't be expected to have all the top gear for all of their skills. So maybe tune the DCs to players having gear several levels lower. Then having the top gear doesn't feel like an obligation, but the boon it should be.

I've already heard several complaints from my players that there seems no way to excel at something, only to hold steady or address a fault. Psychologically a loss has twice the impact of a non-gain despite mathematically being equivalent. Something the think about when phrasing these topics. If PCs can only fall behind the curve under discussion, we need to change the curve discussed, even if ultimately the numbers crunch the same.

Cheers


3 people marked this as a favorite.

Oh, and I like the idea that only a Crit fail in Stealth automatically reveals you.

It'd be neat if PCs had to hold still for several minutes as wary guards scan for whatever made that noise. It also helps with some genre tropes, such as giving a reasonable reason there was a noise, i.e. a cat (who in fact might be the Druid distracting on behalf of the rest of the party).
Since Stealth gives less of a combat advantage in PF2, this adjustment shouldn't imbalance play.


1 person marked this as a favorite.
Unicore wrote:
The other thought I have is that it is unfortunate that the gaming convention is that most skills come down to one die roll. Having a 50% chance of success each check means that most skills with no consequence for failure are going to be successfully completed in one round or two. Balancing skills that can be attempted over and over with those with immediate consequence for failure is difficult to do.

I think the underlying issue is the unification of mathematics for skills, attacks, AC, saves and DCs.

In Pathfinder, the ability to increase one's attribute gets progressively harder as you go down the list:

Skills (+1 every level, feats give +3/+6, items give bonuses at 100gp * bonus squared)
Saves (+1/3-2/3 per level, feats give +2, items give bonuses at 1000gp * bonus squared)
AC (feats give +1, items give bonuses at 1000gp * bonus squared, with several different types readily available)
Attacks (+1 every level (but iteratives come in providing -5s), feats give +1, items give bonuses at 2000gp * bonus squared)
Spell DCs (+1 every two levels (but of a very limited resource), feats give +1 (but only to a subset of them), no items exist that can give bonuses, with the exception of ability boosters).

And in Core Pathfinder, nothing* targets anything lower on the list: you never roll a skill check verses an armour class.

And the impact of being Really Good at something at the top of the list doesn't impact the system's mathematics as much. Sure, you might be great at Sleight of Hand, but when the enemy's attacking you, it's not going to come up.

The system doesn't break down because someone has +40 to stealth or perception (and if it does, it's the GM or Adventure Writer's fault).

If skill modifiers can be divorced from the rest of the extremely tight mathematics, it would solve many problems and result is a much better game.

* Barring oddities like Blinding Critical and Coup de Graĉe; Feint can target HD, but normally targets Sense Motive


Let's try and get a consensus on what the players want the numbers to be. I've put 3 polls below and will be posting them on other forums to gauge what players want with respect to this topic.

1. Poll on chance to hit same level enemies

2. Poll on chance to succeed at skill checks vs. high DC

3. Poll on chance to get hit assuming non-optimal AC values


1 person marked this as a favorite.

Eh. I don't put much stock in the accuracy of forum polls and neither do the developers. Hopefully, people get a chance to comment on this topic in a full playtest survey.


Mekkis wrote:
Unicore wrote:
The other thought I have is that it is unfortunate that the gaming convention is that most skills come down to one die roll. Having a 50% chance of success each check means that most skills with no consequence for failure are going to be successfully completed in one round or two. Balancing skills that can be attempted over and over with those with immediate consequence for failure is difficult to do.

I think the underlying issue is the unification of mathematics for skills, attacks, AC, saves and DCs.

In Pathfinder, the ability to increase one's attribute gets progressively harder as you go down the list:

Skills (+1 every level, feats give +3/+6, items give bonuses at 100gp * bonus squared)
Saves (+1/3-2/3 per level, feats give +2, items give bonuses at 1000gp * bonus squared)
AC (feats give +1, items give bonuses at 1000gp * bonus squared, with several different types readily available)
Attacks (+1 every level (but iteratives come in providing -5s), feats give +1, items give bonuses at 2000gp * bonus squared)
Spell DCs (+1 every two levels (but of a very limited resource), feats give +1 (but only to a subset of them), no items exist that can give bonuses, with the exception of ability boosters).

And in Core Pathfinder, nothing* targets anything lower on the list: you never roll a skill check verses an armour class.

And the impact of being Really Good at something at the top of the list doesn't impact the system's mathematics as much. Sure, you might be great at Sleight of Hand, but when the enemy's attacking you, it's not going to come up.

The system doesn't break down because someone has +40 to stealth or perception (and if it does, it's the GM or Adventure Writer's fault).

If skill modifiers can be divorced from the rest of the extremely tight mathematics, it would solve many problems and result is a much better game.

* Barring oddities like Blinding Critical and Coup de Graĉe; Feint can target HD, but normally targets Sense Motive

The issue with all of this is that the unification of rolls, modifiers, etc., is clearly a core design goal of PF2E. The above reads like, 'why can't we just do PF1E?' I'm not saying you're wrong. I frankly think that PF2E is a worse 4E, currently, and while I like 4E, why would I play this when I can play 4E, if I wanted?


2 people marked this as a favorite.
NemisCassander wrote:
The issue with all of this is that the unification of rolls, modifiers, etc., is clearly a core design goal of PF2E. The above reads like, 'why can't we just do PF1E?' I'm not saying you're wrong. I frankly think that PF2E is a worse 4E, currently, and while I like 4E, why would I play this when I can play 4E, if I wanted?

Yes, PF2 seems like it's trying to borrow only the good parts of 4E, but hesitate to go all the way because it fears association with some of that system that people irrationally hate. It's also keeping some archaic ideas that should be changed already (Touch Armour Class, for example. Vancian magic, for another).

Why have both saving throws and rolling vs. Reflex DCs? The system could be streamlined by picking one or the other and sticking with that instead of this hodgepodge thing that's just a legacy from olden days of D&D.


Pathfinder Rulebook Subscriber
Dasrak wrote:
Thorin001 wrote:
Only the active character rolls. So your PF2 character rolls with a +3 to beat a DC of 11 (the passive characters "take 10").

This has been pointed out to me. While it does help, the math is so skewed against stealth that it just moves it from "impossible" to "very unlikely".

Captain Morgan wrote:
It also occurs to me that sneaking and lying, the two skills we are mostly focused on here, both lack critical failure/success conditions. I wonder if that is specifically because they oppose perception DCs, and if maybe the general skill numbers don't need an overhaul so much as perception scores for NPCs/Monsters just need a big nerf?

Even if we presume lower numbers, it still doesn't work.

Consider a 1st level character with 18 dex and trained in stealth for a +5 check, trying to sneak past level 1 mooks with a perception DC of 11 (can't get much lower than that). That gives a 75% chance of success... on your first check. You need to make checks every time you take the sneak action, which is three times per round. That gives you only a 42% chance to succeed on the first round. In fact, you'd need to drive the perception of level 1 monsters deep in the negatives to get a reasonable chance to succeed at the 1st level. And that's with 1 PC, godsforbid a whole party!

Stealth at very minimum needs fewer rolls (even once per round is a bit excessive; once per sneak action is obscene). Bringing back the distance penalty also helps the skill better scale to different parties; noisier parties can still use stealth, they just need to allow more distance. However, the natural 1 rule probably needs to go as well, since it's very unsuitable to stealth.

The thing that makes the stealth skill different from most other skills is that it's kinda inverted. A successful stealth check doesn't mean you sneak past the enemies. It means they don't spot you. This is what creates the party conundrum; if we bluff our way into the duke's palace we need only one successful bluff...

You might be right on the natural 1 stuff. And the Sneak action being needed for every 10 feet is pretty unreasonable... If it applies outside of encounter mode. I think by RAW you are correct in your assessment, but it never occurred to me that would be the case, and I'm not sure that is intended by the rules right now. Specifying that you can just use a single stealth roll in exploration mode might suffice as a fix. I generally assume stealth checks carry over until something changes the situation, which can happen quite rapidly in combat but isn't likely when someone is on guard duty.

I'm not sure if you need "only 1 stealth check per turn" as a rule though. I think you are going to have points where you break stealth mid turn and need to re-roll it. IE, Sneak > Shoot > Sneak. I see that happening a lot more often than Sneak > Sneak > Sneak for example. Having a rule that you only need to roll stealth multiple times if you somehow break stealth midturn feels clunky.

Also, I'm not sure if making nat 1s not auto fail fixes much. Theoretically, any level appropriate challenge will probably be failed on a nat 1, especially for the non-specialist. I guess it can help it so that stealth specialists have 0% chance of failure on extremely underleveled checks, but that's a pretty fringe benefit. Reducing the number of checks in exploration mode will help offset this problem, but you've still got a pretty good chance of failure when the whole party rolls. I think to make up for it you need some way to insulate from a single bad roll, not just a nat 1. Didn't 4e or 5e use group skill checks, where as long as half the party were successes you were OK? Something like that might work for that problem. Or a critical success letting you cover for a failure, or something. Right now we just have Quiet Allies which isn't nothing but probably isn't enough.


Pathfinder Rulebook Subscriber
Castilliano wrote:

Oh, and I like the idea that only a Crit fail in Stealth automatically reveals you.

It'd be neat if PCs had to hold still for several minutes as wary guards scan for whatever made that noise. It also helps with some genre tropes, such as giving a reasonable reason there was a noise, i.e. a cat (who in fact might be the Druid distracting on behalf of the rest of the party).
Since Stealth gives less of a combat advantage in PF2, this adjustment shouldn't imbalance play.

Stealth is a sizable combat advantage since the first errata, actually. I think to make the critical failure thing work you probably need to build in some caveat for how it works in combat. Like, a failure should probably still make you seen if the enemy was already aware of you. As opposed to the scenario you outline. Maybe a 5th visibility level? "Unbeknownst" or something?


3 people marked this as a favorite.

First off I will say that most of the time I completely disagree with your posts 99% Dead Man Walking. However, in this instance you are correct. With a great stat you barely succeed in PF2 tasks regardless of what that task is (i.e. combat, skill challenge, saving throws, etc. ) This is a HUGE problem because the stats are set up that the majority of your roles are average unless you are hyperspecialized in one area or you main stat bleeds over into multiple areas such as dexterity or wisdom affecting class features, saves and skill checks. This creates very unfun sessions where an athletics or acrobatics check is easy for some characters and almost impossible for others. Heck, my sorceress had to save the druid in our party from the quicksand trap because the fighters had high negatives from armor check penalties that minimized their athletics rolls. Both she and the fighter ended up with a +1 to roll and she rolled high enough to succeed which lead to the odd result of a thin wisp of an elf pulling her friend out of quicksand when the body builder was helpless. Not my idea of each player playing to his strength.

This is really the fault of using 4E game design and Paizo had best steer away from it as it is a losing proposition. Every challenge at every level will feel the same because the system is set up to absorb high stats and with the maximum swing of +8 due to ability score plus legendary bonus every level except for the highest at 19-20 will feel like the same treadmill of I always barely just succeed, whether it be at combat, skills or saving throws unless the opponent is far below or far above the party level. This will lead to burnout and game fatigue really quickly as Paizo players will soon realize that most classes have an "optimal" build and other choices do not measure up. This is already quite clear in class builds for barbarian, paladin, ranger and sorcerer as well as ancestries and armor worn. Its an extremely disappointing result for me personally as I spent a whopping $125 to snatch up a collector's edition book, maps and adventure module. So yeah I put my money where my mouth was and Paizo rolled a critical failure on its skill check. Oh well, lesson learned.


Unicore wrote:
BryonD wrote:
Mark Seifter wrote:
It was actually when we removed glancing blows that suddenly the very lowest level monsters started becoming whiff-fests that usually did nothing even if it was supposed to be a Hard or Severe encounter, which led to lowering damage and upping accuracy a bit for those.

Can you elaborate on this?

According to the bestiary a L16 is a "low threat minion" and minimal creature worth counting for a L20 party. So "the very lowest monsters" depends greatly on who the monster is fighting.

Can you roughly quantify (ballpark) the degree of changes?

If we assume that a L3 monster is a very lowest for a L7 party, how did changing that monster's stats impact its interaction with L1 and L3 characters?

Thanks

I think this quote is specifically addressing 0th level monsters and why their numbers are skewed a little differently than the rest of the monsters (ie their accuracy is higher and their damage is lower.)

OK Thanks

Why is a L0 different with regard to a L2 than a L7 with regard to a L9?


Bryon: Presumably, because there's nothing lower than L0. 'I want a lot of mobs in a L2 encounter. That means L0.' As opposed to, 'I want a lot of mobs in a L6 encounter. That means L1, 2, 3...'

I don't buy the answer, as it makes things _ridiculous_ IMO. A better rule would be to just artificially increase the attack numbers when it's needed; at higher level encounters, not at level 1.


4 people marked this as a favorite.

Just to add some anecdote to this. In my DD game so far... I'll be honest, most of what we've done has been combat and Secret checks, so this is definitely coming at it from that angle. But I still have something to note. Our party has a Monk (me) and a Barbarian, both of us at 18 Strength, so we're at +5 attack bonus, the absolute maximum you can get without being a Fighter I believe. Not too long ago we were in a fight, I think it was the Quasits, where we (through keeping track of our rolls) found the enemy's AC was a 16. Fortunately we have a Bard in the party, and their Inspiration is the only reason we, the second-most optimized attackers we can get, could hit 50/50 odds on hitting on our first attack. Crits are almost a non-entity, and our third attacks are almost guaranteed to miss. Even with Bard Song up and using my Fist instead of my preferred Dragon Style for the Agile, I was still looking at only a 15% chance to generally make 2 of my 3 actions actually worth while. And since MAP applies to combat maneuvers too and we're only level 1, it's not like I actually have any viable alternatives, heck I'm less likely to succeed at a maneuver (since I can't apply Agile to them and I'm not sure if Bard Song applies or not.)

But hey, so far that attack has been one of the harder to hit so far. Unfortunately, I know the rest isn't much better. We had a discussion with the GM recently, where he revealed some of the average stats for the contents of the dungeon. The average AC according to him is 13-15, so we're looking at 50-60 hit chance with our best attack against the average foe, and for the Barbarian his third attack still is probably only hitting 5 to maybe 15% of the time (with my third and fourth only being better because Agile is, like, the best weapon quality in the game right now). Meanwhile, we also learned the average attack bonus is +5 (on par with us) to +6 (on par with an optimized fighter). Again, average attack. I'll admit my actual play experiences with that are skewed because both of us are kinda glass-cannon-y with horribly unoptimized AC, but it's still an issue that the average attack is on par with our optimized attacks.

There's also the issue that with the ~50/50 odds that generally arise... it really makes the value of the +/-10 system questionable. Most of the time it seems you're only going to crit on a nat 20 or crit fail on a nat 1 anyways, unless you're dealing with something well out of your league, with the exception seemingly being MAP and its increased chance of crit-failing (not that that usually matters anyways.)

As one final note... I am notoriously unlucky. There have been times I've had long stretches where I couldn't roll higher than a 10 (notably the Quasit fight was one of those, which just reinforced how miserable that AC was) where-as times when I manage to roll a 19 or 20 are generally note-worthy. So for me or anyone with luck like mine, we're pretty much doomed to fail far more often than we succeed with the math as laid out right now.


3 people marked this as a favorite.
Pathfinder Adventure Path, Lost Omens, Rulebook Subscriber
BryonD wrote:
Unicore wrote:
BryonD wrote:
Mark Seifter wrote:
It was actually when we removed glancing blows that suddenly the very lowest level monsters started becoming whiff-fests that usually did nothing even if it was supposed to be a Hard or Severe encounter, which led to lowering damage and upping accuracy a bit for those.

Can you elaborate on this?

According to the bestiary a L16 is a "low threat minion" and minimal creature worth counting for a L20 party. So "the very lowest monsters" depends greatly on who the monster is fighting.

Can you roughly quantify (ballpark) the degree of changes?

If we assume that a L3 monster is a very lowest for a L7 party, how did changing that monster's stats impact its interaction with L1 and L3 characters?

Thanks

I think this quote is specifically addressing 0th level monsters and why their numbers are skewed a little differently than the rest of the monsters (ie their accuracy is higher and their damage is lower.)

OK Thanks

Why is a L0 different with regard to a L2 than a L7 with regard to a L9?

Because at level 1 and above, there is room to go backwards but there isn't negative level creatures. Mark was talking about how when the party is level 4, you have more "easier" challenges to throw at the party, (level 0-3 creatures), but at first level you don't have that. If level 0 monsters are "wiff-fest" monsters for level 1 characters, they pose no threat to higher level parties, and the idea is that level 0 needs to be a worthy opponent (at least in numbers) for at least 2 or 3 levels until there is room for a more diverse array of monsters that the party might face.

Thus it was necessary to make them hit a little bit more often than other monsters, but do less damage so they were not over the top murder machines. They have to do more adventure design work than other monsters.


1 person marked this as a favorite.

Thank you


Unicore: If that is the design goal, I would submit that level-0 monsters are still overtuned for level-1 characters.

Any of the 4-6 monster encounters in the first playtest module would have been TPKs against my group if I hadn't fudged rolls left and right.


Pathfinder Adventure Path, Lost Omens, Rulebook Subscriber

I do think that PF2 might be a very different game tactically than PF1 and some PF1 strategy ideas are going to get parties killed pretty quickly. For example, moving twice to get in melee range with a higher level foe seems like an atrocious idea in PF2. It was a bad idea in PF1 at levels higher than 6 or 7 but it is a bad idea in PF2 from the beginning.

It also seems like melee stuff and the sorcerer is an even more terrible idea in PF2 than in PF1, as moving once or twice to make a melee attack is terrible idea, but it is an even worse idea if your AC is not anywhere near 15 or 16.

Relearning how to establish battlefield control and stop enemies from getting multiple attacks seems like something that parties are going to need to figure out.

Once they do, I wonder if some of these "killer" combats are going to get a lot easier.


Unicore wrote:

I do think that PF2 might be a very different game tactically than PF1 and some PF1 strategy ideas are going to get parties killed pretty quickly. For example, moving twice to get in melee range with a higher level foe seems like an atrocious idea in PF2. It was a bad idea in PF1 at levels higher than 6 or 7 but it is a bad idea in PF2 from the beginning.

It also seems like melee stuff and the sorcerer is an even more terrible idea in PF2 than in PF1, as moving once or twice to make a melee attack is terrible idea, but it is an even worse idea if your AC is not anywhere near 15 or 16.

Relearning how to establish battlefield control and stop enemies from getting multiple attacks seems like something that parties are going to need to figure out.

Once they do, I wonder if some of these "killer" combats are going to get a lot easier.

I actually disagree with a lot of your suggested tactical differences here. Moving twice in PF1 was a *much* worse strategy than moving twice in PF2 (and then attacking once). Honestly, that first attack is the one that counts, even for enemies, who also suffer iterative attack penalties. Yes, there's a chance the second one hits, but due to bounded accuracy, assuming you have a good AC, that chance is pretty low.

I think PF2 makes the Greatsword Single Big Attack build *much* more viable than it was in PF1. Things like Power Attack and the like mean that you can do this fairly easily. Yes, you still might go for a second attack, but the third attack? Nah, I'd *much* rather be doing something else.

NOTE: I had a Dwarven Sorcerer I was playing at level 4 (Magical Striker, Dwarven War Axe), and one round starting adjacent to an enemy, my moves were: regrip weapon in two hands, cast true strike, swing once. I could have swung 3 times, but this strategy seemed moer viable. Sure enough, one of the d20 rolls was a 16 and I ended up critting and rolling 6d12 damage (I had previously cast Magic Weapon). No way I was doing that with three regular swings, or even two swings two-handed without true strike.


2 people marked this as a favorite.

Unicore: How do you stop 12 Shortbow Arrows coming from the darkness at your party in A2 if the Goblin Warriors win initiative?

About the only thing the GM can do is either

1) Roll initiative separately for the Goblins so that some party members might go before some of the Goblins, but that's squarely against 'best practices for running lots of mobs'.

2) Make the Goblins attack less... why? And heaven forfend if you have one Goblin go get the others...

I mean, yeah, there are tactics to learn, but the concern about TPKs with large number of level-0 mobs is quite real.

Designer

4 people marked this as a favorite.
Deadmanwalking wrote:
Captain Morgan wrote:
So what is the appropriate percentage of success for it to feel good? Keeping in mind the differences between someone who 100% optimizes vs someone who dabbles? Should 100% optimization allow 95% success?

At 13th level, an absolutely optimal character with level-appropriate items has a +23 (13 level, 5 Ability, 3 Item, 2 Proficiency). A High DC at that level is 32, meaning they still only have a 60% chance of success at that. That's probably not sufficient. It feels like a lot of investment for such a non-ideal percentage.

As levels rise past that, the odds get better (peaking at 75% at 20th level). A 75% is sufficient (indeed, a 70% is probably sufficient on high DC stuff)...but you don't get anywhere close to that for most of your career.

Equally importantly, someone with a minimal investment (a stat of 12 to start, rising every five levels, plus Trained and maybe a non-magic Master Quality Item eventually) should probably have a decent shot at low Skill DCs if not high. At 4th, such a person has a 45% chance of low DCs. At 13th, they've hit 50%, but by 20th, it's back down to 45%. Which is again a bit low.

Dropping the Low DCs by two or three and the high ones by about two pre-14th level (or thereabouts...a few low level Trivial ones would also need to be adjusted) seems like it'd be sufficient to me, but would need to be tested, of course.

This is aside from the monster skills and monster Perception issues noted previously.

It strikes me going through it that Table 10-2 is affected by some (but not all three) of the factors that led level 13 monsters to be off in skills, as one of its bases was the same set of benchmarks that was off for monsters. So it wouldn't surprise me if Low, High, Severe, and Extreme had a 1-2 point drop at some levels (Severe, matching its definition on page 336, should be at 55% chance for a maxed character not benefiting from any conditional or circumstance bonuses, so any time it isn't, that row is likely to get a downward adjustment in Severe and Extreme at least (High and Low actually are not really related to the most optimized character in how they are determined, as seen in their definition on page 336, so it won't always move in lockstep, but they could be affected. Trivial is definitely still correct at all levels).

Designer

20 people marked this as a favorite.
NemisCassander wrote:

This is a very interesting thread, the thesis of which I completely 100% agree.

As for stealth, I would like to suggest a codification of what someone suggested upthread:

When you make a Stealth check to Hide or Sneak, make a DCX (X being under intense discussion here), and change your visibility status as below.

Success: Decrease your visibility level by 1. (Seen to Concealed, Concealed to Sensed, Sensed to Unseen.)
Critical Success: Decrease your visibility level by 2.
Failure: Increase your visibility level by 1.
Critical Failure: Increase your visibility level by 2.

This mirrors a lot of other condition levels in the Playtest, and interacts with the wonderfully-streamlined Visibility rules. (Seriously, the 4-level visibility system works very very well.)

Also, this allows for nuanced multi-level reactions to Stealth. Not only can the Guard Captain have a different visibility level for the Stealthing Rogue than the Grunts, but you can actually adjudicate it and it's not binary.

If you include a requirement to Sneaking/Hiding of 'you must be at least Concealed' (this is already true, no?), that would mean a Success on Sneaking while Concealed moves you to Sensed, etc. etc.

Also, several spells (like Blur) and conditions (like dazzled) could be changed to say, e.g., Blur reducing your visibility level with everyone by 1. (Instead of 'you are concealed'.)

Etc. etc.

Not that this is directly related to the main thesis of the thread, but I really like the idea of failing on a Sneak when you were Unseen only making you Sensed (so "Hey, I hear something moving over there! Let's go check it out" and then they roll a Seek) and only a critical fail making you fully Seen. I've added it to my list of cool ideas you guys have on the forums to bring up as possibilities (currently 85 entries long and growing).


Mark Seifter wrote:


Not that this is directly related to the main thesis of the thread, but I really like the idea of failing on a Sneak when you were Unseen only making you Sensed (so "Hey, I hear something moving over there! Let's go check it out" and then they roll a Seek) and only a critical fail making you fully Seen. I've added it to my list of cool ideas you guys have on the forums to bring up as possibilities (currently 85 entries long and growing).

:D

Thank you, but the one you should thank--and give credit to--is the person upthread who originally suggested it. I just tried for a codification.

And yes, the _exact_ scenario I was thinking of was the, 'Wait, I thought I heard something... is something out there?'

Designer

3 people marked this as a favorite.
NemisCassander wrote:
Mark Seifter wrote:


Not that this is directly related to the main thesis of the thread, but I really like the idea of failing on a Sneak when you were Unseen only making you Sensed (so "Hey, I hear something moving over there! Let's go check it out" and then they roll a Seek) and only a critical fail making you fully Seen. I've added it to my list of cool ideas you guys have on the forums to bring up as possibilities (currently 85 entries long and growing).

:D

Thank you, but the one you should thank--and give credit to--is the person upthread who originally suggested it. I just tried for a codification.

And yes, the _exact_ scenario I was thinking of was the, 'Wait, I thought I heard something... is something out there?'

Sorry Captain Morgan, credit where credit is due!


11 people marked this as a favorite.
Deadmanwalking wrote:
So, I like a lot of things about PF2. The math is much cleaner, character creation is a breeze....

DMW, I completely agree with everything you said, with one exception. I firmly believe that the core problem of PF2 is, in fact, the +10/-10 crit range paradigm. I'll try to explain my reasoning below.

Firstly, I would like to state that, as a huge fan of World of Darkness games, I really dig the concept of four degrees of sucesses. However, on that game, the degrees are dependent on a specific system: 5 or more successes (for critical successes) and 1 on a chance die (for critical failures). In order to achieve a crit success in WoD, you either have to be lucky (each D10 in your "pool" has about 30-35% to earn you a success, a character particularly good at something might have a pool of 5 or 6 dice) or have a huge pool (a pool around 12-15 D10s is necessary to consistently achieve crits, a huge number even for Ancient Vampires). For crit failures, you had to rely on a chance die, basically when your pool would equal 0 or less (due to circunstancial penalties, for instance), you had the chance to throw a chance D10, on a 10 you would suceed, 2-9 you fail and 1 you crit fail. I like this system because while it's possible to achieve a critical success, specially if you're really good at something, it's not something overwhelmingly present.

In PF2, however, a +10 result above the target DC is not very difficult to achieve, specially against a weaker opponent or with a few bonuses stacked together. The fact of the matter is that it makes the math EXTREMELY tight, and without extreme controle the crit chances just escalates too quickly. This leads to a whole lot of problems, but most notably is the problem of To Hit bonuses vs. AC. In the current state of the game, we have an optimized fighter with 55% hit chance (5% crit chance) against a similar leveled monster. Add flanking and Inspire Courage and suddenly he hits 70% and crits 20%, which is huge! But that means that non-optimized characters like a Cleric with good strength will only hit, at most, 30-40% of the time, and would not benefit from the increased accuracy in their crit chance.

If, on the other hand, Monsters ACs are toned down to fit the regular cleric as the base (around 50-55% hit chance), then the fighter would be criting regularly at 20% and would hit 35% crit chance with small bonus, and would most certainly just ROFLSTOMP the encounters. Then monsters HP would have to UP to compensate, and etc... You get the idea. But the core of the problem is still there.

For me at least, it's plain as day that almost every other decision in design was tweaked to fit the +10/-10 crit paradigm, but given the problems that are constantly arising in the forums, it seems that it's in the detriment to the game experience. From a game design perspective, I think that the current state of crits just doen'st work in practice, and it should be change completely. Maybe reverting to crits on 20/1s is the best alternative, maybe there'se another way. I'll think about it some more and share my thoughts later.

Cheers!


7 people marked this as a favorite.

Chiming in here, and agreeing with OP. I've since swapped places with my GM since he doesn't like running prewritten, but when I was playing as a player, one thing I noticed a lot was that challenges vs "not completely optimized" characters seemed completely FeelBad all around. Our first attempt at Lost Star crashed and burned, and when we tried it again we replaced half the party roster with optimized Big Stupid Beatsticks to ensure maximum odds of success. Deciding to make roleplaying-focused character-building decisions can actually endanger your character if it interferes with any of the stats that the game assumes you've maxed out (usually your corresponding offense and defense stats). On the larger meta-level, it also feels like the game expects you have a specific party composition as well, reducing flexibility even further.

It's almost like we need two versions of the game - one where the players are assumed to have optimized to the greatest extremes the system allows, and another where "reasonably good" is the baseline. As the system stands currently, tuning the system for one of those two groups breaks the system for the other group, because every +1 matters a lot more in PF2 than it did in PF1. Two versions of the game obviously isn't feasible, however, so I'm at a loss. Given that loss, I'd rather have a system tuned closer to "characters are reasonably good" standards than "characters are 100% optimized" standards.

Liberty's Edge

4 people marked this as a favorite.
Mark Seifter wrote:
It strikes me going through it that Table 10-2 is affected by some (but not all three) of the factors that led level 13 monsters to be off in skills, as one of its bases was the same set of benchmarks that was off for monsters. So it wouldn't surprise me if Low, High, Severe, and Extreme had a 1-2 point drop at some levels (Severe, matching its definition on page 336, should be at 55% chance for a maxed character not benefiting from any conditional or circumstance bonuses, so any time it isn't, that row is likely to get a downward adjustment in Severe and Extreme at least (High and Low actually are not really related to the most optimized character in how they are determined, as seen in their definition on page 336, so it won't always move in lockstep, but they could be affected. Trivial is definitely still correct at all levels).

Ah! That definitely helps a fair bit, especially since High DCs are the default for a lot of checks. I'm not sure it goes quite far enough, but it's a huge step in the direction of characters feeling and being more effective.

Monster Skills should probably mostly follow the same rules since they can wind up the DC of PC rolls IMO, which they currently do not. For example, a +23 skill at 13th level results in a DC 33, or Severe DC, for PCs to roll against, this seems like something that should be rarer than 'every monster of this level has a DC like this' based on the DC level descriptions.


NemisCassander wrote:

Unicore: How do you stop 12 Shortbow Arrows coming from the darkness at your party in A2 if the Goblin Warriors win initiative?

About the only thing the GM can do is either

1) Roll initiative separately for the Goblins so that some party members might go before some of the Goblins, but that's squarely against 'best practices for running lots of mobs'.

2) Make the Goblins attack less... why? And heaven forfend if you have one Goblin go get the others...

I mean, yeah, there are tactics to learn, but the concern about TPKs with large number of level-0 mobs is quite real.

1. Yep. Four is not a mob IMO, and you've shown you understand bunched initiative and focused fire are issues (as they likely should be in group v. group battles). So yes, you should've made that adjustment.

2. Yep. Because they don't start with weapons out since they're working and most shouldn't have line of sight with those pillars. Plus, book says they're itching to enter melee. And being Goblins, they very much might waste time yelling, laughing, or trying to Intimidate if that's what best suits the game & story.

If one runs for reinforcements, it wouldn't be to the scary boss and the other Goblins would have to leave a tactically sound position, so may just stay. The rushing Goblin would (spoiler re: noise) so a wounded party would know to retreat and a healthy party could set up a great defense in A2, which favors them much better than (redacted for spoilers).

In asking the questions, you've practically answered them. Why you want to put the onus elsewhere when there are GM solutions, I don't know.
Cheers

51 to 100 of 227 << first < prev | 1 | 2 | 3 | 4 | 5 | next > last >>
Community / Forums / Archive / Pathfinder / Playtests & Prerelease Discussions / Pathfinder Playtest / Pathfinder Playtest General Discussion / Deadmanwalking's Main Problem With PF2 All Messageboards

Want to post a reply? Sign in.