Pathfinder Is Still Bad

[tussock]

With count-down stealth failure, how do I interact with the enemy who hurts my stealth points? Is that healing? Do I stab the lone guard to regain stealth points? Does my Eddie Murphy routine get me bonus stealth

Stealth should be a shallow hp pool that people make attacks against when they try to detect you. The pool should regenerate when it's not being damaged so getting out of sight for a round or two can be a big boon.

If a guard hurt your stealth points your options are: try to leave before being discovered, continue on and hope he doesn't hit you again likely discovering you, find a place he can't detect and spend some time staying there quietly to regenerate your stealth points before trying again, or kill the guard. Attacking the guard would force another Stealth check making it more likely that he and any others around detect you. If you pass the check the guard is still aware that he was shot (or whatever) and would react accordingly likely raising alarm or calling other guards. If you killed him in one shot you've now got a body to hide (interacting with an object) so that no one finds him but once you've done that you can just keep going on your mission with one less guard around.

Wait, so I know when people see me but don't see me? OK, I was thinking DM tracking and .... That's ... how are we describing that? It's like the guard's suspicion or alert state relative to your real position goes up (or something, in-game).

So when you're low on stealth points, the guards are poking the bushes you're in (in an abstract sense). But per-character so them "searching your hiding spot" or whatever doesn't hurt your friend at the other end of the corridor. Not looking behind the curtain in game. Well, sort of. It's none too abstract if I can face-stab that one guard.


But if I'm gaming it, it seems like I just try and back off repeatedly until I get the success, and then repeat at each juncture until I can walk out with the MacGuffin. So obviously penalties for retries, and then you just have to kill enough guards each loop to counter the penalties, so ramping penalties for repeats. Then they start crit-fishing, so no crits.

The benefit I see there is degrees of failure on stealth checks, where most of them do not reveal your position or aggro the guard, letting you try something different. Like a game of D&D or something. Levels of stealth if you will, between completely ignorant opponents and them knowing it's you and exactly where you are.

Which I admit is probably easier to write as an abstract pool than a condition track or other mini-game, and those other things also have flaws (too many conditions, cross-referencing everything, mini-games in general). But as I mentioned, I totally tried damage in place of condition tracks for stuff and found easy math like that is just a huge boner-killer for most people.


But thanks, I think I got the idea of it at least.

[...You Lost Me]

Maybe I'm missing something, but I don't think making stealth deterministic is such a bad thing. I'll explain my reasoning, you tell me where I'm going south or why rolling is better.

• Less Frustrating: In a similar vein, having that one peasant notice you just because he rolled a 20 and you rolled a 2 isn't fun when you've put all your resources into hiding.
• Less Exclusionary: Stealth is not a game everyone gets to play, so minimizing the spotlight on it is probably good. More rolls > More Uncertainty > More time, and I worry the party paladin will leave the room to play Mario Kart.
• Encouraging Parties: When there are more perceivers / more listeners, there will be more rolls. This increases the chance for outliers, which is probably going to hurt the players.
• More Intuitive: In D&D, when player A has Hide +6 and player B has Hide +2, player A hides better than B 70% of the time. This is confusing, because having Hide +6 intuitively indicates that A is better than B, yet B is showing off regularly. More than one of my player has objected to this, and it seems to bother people in a way that attack bonuses do not.

[Lago PARANOIA]

@...You Lost Me:

Less Frustrating/More Intuitive: That's just part of the game. Do people get upset when someone rolls really poorly on an important opposed Bluff/Sense Motive check or blows a really important Saving Throw? You could make an argument that Stealth using a d20 has too much variance, but that's different from saying that the problem is that it's deterministic.

Less Exclusionary: Is that necessarily the case? Stealth is reasonably full-party at low levels of 3E and 4E D&D; while a halfling rogue can completely school anyone in the party in the 'sneak past the sleeping ogre' challenge, it's not out of the question for the unarmored-and-sheathed paladin or the party wizard to try to sneak past, too. And heck, there are games like Shadowrun where if someone complains that they're being excluded from more delicate runs because their Infiltration and Stealth skills suck, they get a withering glare and told to build their characters better.

Encouraging Parties: That one is genuinely a problem, where 20 Hobgoblins make a better fixed sentry than a single beholder. Hence the point of the past few pages.

And at any rate making stealth deterministic isn't going to solve any of the underlying problems. It just shuffles it around. Instead of people getting angry that they lost a stealth contest to someone with 4 less points than them, people are going to get angry that they will always lose a stealth contest to someone with one more point than them. Having stealth being deterministic will cut down on the time spent rolling, but it won't translate to more in-game screentime. And people getting upset that sometimes they lose to people of someone with less skill is just special pleading; that sort of jealousy certainly doesn't apply to debates, cooking contests, acrobatics, seduction, forgery, etc. The only one I've seen it regularly come up in is with feats of opposed strength and 3E D&D already offers an out for that anyway.

[ishy]

In my experience, people don't want to split the party in D&D. So the fact that the paladin can potentially fuck up the stealth check for everybody, just means nobody tries to stealth past the encounter.

[Lago PARANOIA]

Sure, but making stealth deterministic won't change that.

[...You Lost Me]

Intuitive: People do not have the same reaction to attack rolls and saves that they have about Stealth/Social checks. I think that's because you tend to make 1 roll to determine stealth/social, whereas a combat is decided by multiple saves/attacks.

Exclusionary: Some people wear stone plate. Some people don't put ranks in hide. Those people shouldn't feel like they need to open their DS because the rogue is going to spend 10 minutes rolling and checking and subtracting and referencing.

I disagree that the problems will be shuffled. My players would be significantly less annoyed about losing to someone measurably better than them, and have said so (if you want to include the niche case of running 20 stealth contests, then I agree a single d20 + stat would be better). Reducing rolls will speed up play, which is what I was talking about, and that "jealous certainty" can easily apply to debates, cooking, acrobatics, forgery even if I didn't explicitly mention it.

[tussock]

@ishy, I'm sure the reason no one uses stealth in 3e/Pathfinder is because it doesn't work.

Players are adept at noticing when their characters try something, get normal dice results, and then get stabbed in the face a lot. Like, people were surprised by how stealth didn't work back in August/September Y2K, but then they rolled a new character and didn't bother adding stealth skills again, let alone try to use them.

I presume the reason it's continued that way is because designers don't actually want stealth to work. "You see Monster, you Fight, stealth check for surprise attack". The basic 3e design principle was to never say no to players, just give them rolls that fail most of the time and let them take the hint. Such is stealth beyond the surprise round.

[Dean]

You Lost Me have you been reading the thread? Because it doesn't seem like you have. Your complaints are all things I have directly addressed.

20 defenders is a perfect defense: In D&D 20 Hobgoblins are way more likely to detect you than a Dragon is. That is a problem but it's a problem I've discussed fixing. In D&D Hobgoblins have a +2 to Spot so 20 of them rolling d20's mean they basically get a 22 but it doesn't have to be modeled that way. In my system Hobgoblins have a passive Notice Defense of 12 (10+2 from their spot skill). 20 of them together add +4 to their Defense so altogether they have a 16. So any round you roll a 16 or higher you can sneak around them.

Only the Rogue gets to play: In my system everyone gets Stealth automatically and I think that would be true of anyone's stealth system designed on these boards. Every character should be given the ability to perform competently in combat, stealth, movement, and social challenges purely by leveling if levels mean anything at all. So the same way everyone gets a BAB everyone should get an automatically rising Stealth check. Paladins can still wear plate mail to drop their check which in D&D's rolling model means you leave him at home but it doesn't have to be that way. I would model group stealth checks much like my group notice checks. In my group Stealth checks everyone rolls and applies a penalty to their stealth check based on the groups size but every member of the group uses the best member's roll. So a group of 5 takes a -4 to their check but gets to pick the best roll of 5 d20's. By modeling the dice mechanic correctly you can make small groups not mind a Paladin.

Every roll is a step closer to game loss: I covered that. I like an HP based system where regeneration occurs when you sneak successfully to a hidden area so you can actually botch several rolls and still succeed at the mission as long as those botches are far enough apart to never drop you to 0.

Stealth is just one totally deterministic die roll and isn't interactive: I've covered that! Your Stealth hp (notice points) let you know when the guards are close to finding you and think they heard something and you can choose whether to push through, retreat, or attack. Obviously if you fleshed out this system you would also create many different feats, abilities, and spells that offered people powers that were useful in the Stealth minigame like chucking rocks or using illusions to make people investigate places you weren't giving you a reprieve.

Finally @tussock: Making unlimited stealth attempts to avoid game loss wouldn't work. If you're in an area where you're being hit with Stealth damage retreating from that area is also something which may deal you stealth damage and possibly reveal you. So retreating isn't an auto-win it's just the fastest way to try to get out of a bad situation. I can see how if you were moving towards the outskirts of a bandit campfire at the limit of their detect range and made a bad roll and took some stealth damage that you might want to just sneak away and try to reset the encounter before trying again but that's perfectly in genre. If you step on a branch before you're even close it makes perfect sense to just sneak away and try again when suspicions have died down. When you're actually failing checks in the thick of it retreating is no kind of win button.

[...You Lost Me]

I've certainly read your post, but I assumed it was spitballing because it has all sorts of weird holes like two independent people both reducing your stealth despite not communicating with each other, or hobgoblins spotting together for +4 despite that being the least optimal thing to do.

That all aside, you saying "I made a thing too!" does not actually resolve my question. I was wondering what attributes of static stealth values were unnecessary, not whether or not you could make a system that satisfies 4 criteria I was mulling over this morning while simultaneously opening ugly rule holes.

[Username17]

However frustrating it is to lose because you rolled badly, it's a million times worse to lose because you presented a course of action you thought should work and the MC told you that you automagically fail without rolling. That is why we roll dice. Because otherwise we're playing Cops & Robbers where someone other than you has the final say as to whether you succeed or fail, and that's insanely frustrating when you fail (and rather trivial when you succeed).

Beholders have a spot check of +22, literally 20 points higher than a Hobgoblin Warrior's +2. Every roll a Beholder can get is higher than every roll a Hobgoblin can get. If a Beholder rolls a natural 1, he still gets a 23, which is higher than any number a Hobgoblin can get. To the extent that Hobgoblins are better at spotting people in any numbers than a single Beholder is in D&D (only true in very large numbers for characters with stealth in a range the beholder is likely to see in any case), it is because of the property of each sentry forcing the sneaker to roll again. You're fishing for 1s from the stealth character, not just fishing for 20s from the Hobgoblin. There is however, no reason to do that. Four hundred Hobgoblins would be virtually certain to spot a character with a Hide of +17, while a Beholder would only spot them 70% of the time. But if you just made one stealth check for each character sneaking rather than one per sneaker per sentry, this issue would not exist.

-Username17

[tussock]

But Dean, if it's about 2 checks to get out, and I can only take 2 fails worth of damage at that point, then I get out. Every time. If it's 5 checks to get out, it was also 5 to get in and so the stealth game fails a lot, and tends to fail deeply surrounded by enemies, so players can't use it like that and expect to survive.

There's a really fine line there you're trying to walk. Shallow challenges will be trivial for retries and even slightly deeper challenges are stupid to attempt because your penalty for failure is growing fast as the stealth points wear down, which is also making escape much less likely.

Maybe the trick is just stealth to the first guards and assassinate them all, then back out and rest, do it again the next day with your next nova. Depends on the rest of the system, but there's got to be a penalty for retries.

[Dean]

hobgoblins spotting together for +4 despite that being the least optimal thing to do.

It's neither optimal or suboptimal because their is no choice involved but if there was a choice then grouping would be optimal because without grouping those 20 Hobgoblins would just have their basic Notice DC of 12 and be no more difficult en masse than individually. You're still thinking that they roll to detect and the better way is to have detection be a set DC modified by circumstances and then have Stealth have to pass it. In this model groups get bonuses because groups are intrinsically harder to move by because to be undetected you must be undetected by everyone. This makes groups more difficult even if no member of the group is allied with any other.

If it helps, try imagining the stealth game as working like climbing walls in D&D. There's no choices on the part of the wall. It has a DC to climb and circumstances can change it's base DC making it higher or lower. If the wall is slippery it's DC goes up and if it has handholds its DC goes down, the climbers difficulty is set by the problem in front of him and then he rolls to make progress round by round.

I don't know what you mean by two independent people reducing your stealth so I don't have a response to that.

[hogarth]

Sure, but making stealth deterministic won't change that.

And at any rate, you can take 10 on Stealth checks in Pathfinder so you already have an option for deterministic Stealth within PF.

[ishy]

Sure, but making stealth deterministic won't change that.

It may change it. If you can tell the paladin that you could have just sneaked past some of the bullshit low level encounters if she just spend a skill point in stealth, she might just do that on the next level up.
While, if it is determined by the RNG, the paladin will argue that they're better off murdering everyone than risk failure and being surprised in a bad spot. Stealth is not like combat where a single bad result doesn't matter that much; a single bad stealth roll can turn into a tpk.

And at any rate, you can take 10 on Stealth checks in Pathfinder so you already have an option for deterministic Stealth within PF.

You can only take 10 when you're not in immediate danger or distracted and when there are no threats. So if your DM decides that being discovered in the middle of a bandit camp is enough of a threat, you can't use take 10.

---

Only the Rogue gets to play: In my system everyone gets Stealth automatically and I think that would be true of anyone's stealth system designed on these boards. Every character should be given the ability to perform competently in combat, stealth, movement, and social challenges purely by leveling if levels mean anything at all.

This is why I favor mechanics in which sneaking around is some kind of savings throw (or equivalent) and your level bonus (or equivalent, like Edge or whatever) applies as a bonus to it. It is, like not being killed by a stray arrow at Agincourt - one of those things that is successful because you are a hero.

Indeed, this is why characters like the Hulk, or comparable giant brute characters, always have a part where they just suddenly fucking start sneaking around.

---

However frustrating it is to lose because you rolled badly, it's a million times worse to lose because you presented a course of action you thought should work and the MC told you that you automagically fail without rolling.
-Username17

And when you roll a natural 20, but the beholder spots you anyway, people are going to be even more upset.
Your players need basic information to make meaningful decisions. Otherwise you might as well abstract all decisions to the roll of a die.

[Ghremdal]

Most of the complaints are based on the d20 method of resolution. In dicepool systems spending a few points on stealth actually makes you noticeably better at it, to the point that sneaking past a mook goes from 0% to 75%+.

Also people need to get over that rolling 20 is something special and you should autosucceed because you rolled great. It might be true in shit games like 5e, but it should be true in general.

[...You Lost Me]

hobgoblins spotting together for +4 despite that being the least optimal thing to do.

It's neither optimal or suboptimal because their is no choice involved but if there was a choice then grouping would be optimal because without grouping those 20 Hobgoblins would just have their basic Notice DC of 12 and be no more difficult en masse than individually. You're still thinking that they roll to detect and the better way is to have detection be a set DC modified by circumstances and then have Stealth have to pass it. In this model groups get bonuses because groups are intrinsically harder to move by because to be undetected you must be undetected by everyone. This makes groups more difficult even if no member of the group is allied with any other.

This is why I called it an "ugly rules hole". Because hobgoblins automagically are worse at noticing things when they work together instead of when they refuse to talk or when some of them have slightly different numbers.

If it helps, try imagining the stealth game as working like climbing walls in D&D. There's no choices on the part of the wall. It has a DC to climb and circumstances can change it's base DC making it higher or lower. If the wall is slippery it's DC goes up and if it has handholds its DC goes down, the climbers difficulty is set by the problem in front of him and then he rolls to make progress round by round.

Hobgoblins are alive. Walls are not alive. I would assume this is a very simple distinction.

I don't know what you mean by two independent people reducing your stealth so I don't have a response to that.

You have a stealth HP. Person A tries to see you, reducing that HP. Person B refuses to communicate with person A in any way, but reduces the same HP. Despite the two of them being completely uncooperative, they can still find you in half the time. That makes no sense.

[Dean]

Only the Rogue gets to play: In my system everyone gets Stealth automatically and I think that would be true of anyone's stealth system designed on these boards. Every character should be given the ability to perform competently in combat, stealth, movement, and social challenges purely by leveling if levels mean anything at all.

This is why I favor mechanics in which sneaking around is some kind of savings throw (or equivalent) and your level bonus (or equivalent, like Edge or whatever) applies as a bonus to it. It is, like not being killed by a stray arrow at Agincourt - one of those things that is successful because you are a hero.

Indeed, this is why characters like the Hulk, or comparable giant brute characters, always have a part where they just suddenly fucking start sneaking around.

The Hulk has a dedicated stealth form where he turns into a small quiet supergenius who is an incomparabe hacker and who carries around teleporters and shit that he makes. The Hulk has a dedicated Combat/Movement form and a dedicated Stealth/Social form. Hulk is a high level PC so he is expected to be able to play every minigame which he totally fucking does.

You Lost me stop instinctively arguing and listen because I am definitely right and you are definitely wrong. Stealth is the pre-awareness game. The stealth game begins before the defenders are aware it is occurring. The game ends when the defenders are aware it is occuring. You keep placing the action on the defenders side and that is wrong because once the defenders are declaring actions the game is over. Look...

Person A tries to see you, reducing that HP. Person B refuses to communicate with person A in any way, but reduces the same HP. Despite the two of them being completely uncooperative, they can still find you in half the time. That makes no sense.

Even if the two people you are sneaking past are deaf and mute and literally incapable of communicating the existence of more than once person's sight lines are harder to avoid than one person's sight line. Action you take to avoid person 1 makes it possible that you reveal yourself to person 2 so they attack the same hp. Both of their existence makes stealthing harder on you. "But why isn't it even harder when both of them are working together to find you?" I hear you say but that question makes no sense in context and that's what you need to realize. No one is working together to find you. No one is doing anything at you. You are undetected. When the stealth game is happening you are not there in the eyes of the defender. The Hobgoblins exist and will react when they see you but until you run out of stealth hp they do not see you. The cooperative nature of the guards when they know something is happening doesn't matter in the state where they definitionally don't know something is happening.

Hobgoblins are alive. Walls are not alive

It doesn't matter because neither is allowed to make choices. They are both static defenses and they have to be that way if the game is to work. Imagine this from the opposite side. If the PC's have made camp and set someone on watch the stealth game will function if an enemy Warg makes stealth checks against a PC but it will absolutely not function if you tell the PC "You are being stealthed by something you can't see yet, what do you do". That sentence is the end of the stealth game, not the start of it. To summarize; The defenders in the stealth game can NEVER make choices in it if the game is to work.

I understand that I haven't written a complete rules system here. I never said if people get notice penalties when they are sleeping or bonuses if they're "On Patrol" or "Alerted" or whatever. It is definitely true that I have not written out an entire working subsystem with all details covered but when we're just talking about getting the basic concept of a thing across that shouldn't be required yet. It's as if I described d20's basic attack and damage rolls to you and you've declared them nonfunctional because I didn't describe what happens if you attack from higher ground, or if you attack someone's eyes. Some of those questions would need to be answered in the rules final forms, sure, but some are willing sacrifices to abstraction. You are upset that the rules I've outlined don't create a mechanical difference between two unallied Hobgoblins looking in the same direction and two allied Hobgoblins specifically watching each others back but I don't want there to be a difference between those things because that means tracking the facing and sight lines of every creature you ever come across and fuck that noise. D&D doesn't have facing and sight lines anymore and it's better for it. An abstract stealth mechanic that just lets you roll against all the Goblin's notice DC is infinitely better than one that makes you track every goblins facing in the Goblin Fortress even though the second one would let you definitely know that creating a triangle of central facing goblins in every room is better than not doing so.

Edit: I also think its super reasonable to break this off into a separate thread just to talk about Stealth systems. I'll write up specifics for you to attack me on that way it doesn't feel like I'm using some quantum rules defense and we don't have an 8 page Stealth rules tangent I can never find again in our 300 page thread about how Pathfinder blows.

[Orca]

If anyone cares, the mystic theurge has now been discontinued in official PF.

[hogarth]

If anyone cares, the mystic theurge has now been discontinued in official PF.

The whole "spell-like abilities are different from spells, except they aren't, except they are" FAQ was fucking stupid in the first place.

[Antariuk]

That, and the problem becomes worse with all those "this (Sp) works like spell XXX, except that YYY" abilities where some people argue it counts as having spell XXX for all intends and purposes, and some poeple argue it doesn't. These discussions have been around since the CRB's release, for fuck's sake.

What frustrates me to no end is that the whole FAQ system is so needlessly byzantine. Unless you dig through the messagebord like a Google bot with too much time on its hands, you have no idea why certain rules changes are made (or where you would find the reasoning, if any was given) and have to F5 the FAQ page every time you sit at a "100% official PF rules" table and you will still get into arguments because nobody is aware of all FAQs ever.

[Lago PARANOIA]

The funny thing about the change is that early entry with spell-like abilities into Pathfinder PrCs turned out to be surprisingly small beer. Most PrCs already had a hard level lock with skill and/or feat prerequisites. And the builds where you could shave off a few prerequisite levels ended up being Not A Big Deal. Character level 4 Mystic Theurges and CL 3 Eldritch Knights ended up being usable, not overpowered. This change pretty much eliminates no overpowered character concepts and ends up nerfing several marginal ones.

But that's the Pathfinder Touch for you, yo.

[Antariuk]

I actually looked into the numerous threads sparked by that FAQ backpedaling and it seems a lot more people are getting upset about this than you think. Reminds me a little bit of 4E's gnomes to be honest, but I suspect part of the feelings right now exist only because that FAQ came totally out of the left field, without any explanation or anything (unlike the former FAQ it replaced).

And yes, MT and EK have become useless (again), which is a shame because both offered something not easily emulated by existing classes or archetypes. Several people proposed to change the core PrC's entry requirements in general and that is probably the best idea, other than dumping them alltogether.

[RobbyPants]

The funny thing about the change is that early entry with spell-like abilities into Pathfinder PrCs turned out to be surprisingly small beer. Most PrCs already had a hard level lock with skill and/or feat prerequisites. And the builds where you could shave off a few prerequisite levels ended up being Not A Big Deal. Character level 4 Mystic Theurges and CL 3 Eldritch Knights ended up being usable, not overpowered. This change pretty much eliminates no overpowered character concepts and ends up nerfing several marginal ones.

But that's the Pathfinder Touch for you, yo.

It's good to see them sticking close to their 3E roots. We can't have non-full casters being remotely useful now, can we?

[fectin]

Mystic theurge has always been worthwhile. Just enter via Ur-Priest, and you'll have no problems.

[TOZ]

Dragons can use AMFs now.