Skills and their mechanics

[pedroph]

I am making this thread to be dedicated to discussing Skills and its mechanics.

The only other relevant thread that discusses it I could find is:
DD-475e-or-Old-is-Old-my-reaction-on-my-first.

some of the things regarding skills' DC that are there:

GreyWolfLord wrote:

The skill system seems to be more complicated and hard to remember in many ways. I think this is why Numbers that go up is actually been popular in the past, because a number is always easy to remember.

With PF2e you get the skill, but very limited upgrades. After that you seem to get better with skill feats. This makes a division that is hard to keep track of in my mind (as I said, it could be old is getting old).

I would prefer something a little simpler and easier to keep track of.

Escalating DCs. (for those who can't find this, the table is on pg 337 of the Playtest rulebook) This is what 4e did and it appears to be what PF2e is doing. So you want to climb that wall over there...great, that's a DC 14. Oh wait, your a level 12 character, sorry, that's a DC 30.

Sure, it's the same wall and all, but the DC just got that much tougher because you are a higher level...

Say WHAT!!!???

Yes, I know supposedly the task it represents gets harder, but that isn't necessarily how it gets played or seen.

Joshua James Jordan wrote:

I think you are misinterpreting the intent. The narrative would change to accommodate the increase in DC for the higher level challenge. The higher level wall has no handholds, is covered in grease, has thousands of wriggling worms coming out of it...whatever you want.

That chart on page 337 is great for coming up with DCs on the fly. If I think the wall is something that should be easy for someone at 12th level, I'd go with the trivial column.

The intent is to have the DCs escalate and the narrative behind the difficulty also increase to help the players feel "cooler".

John Lynch 106 wrote:

Here's a problem with that. 4th ed did the same thing and it wasn't enough to stop DMs from doing exactly that.

Cydeth wrote:

Difficulty Classes is a segment of the book 3 pages long. That's 6 columns of text. The chart is half of one column. On the very next page, two-thirds of the page (1 and 1/3 columns) is given over to examples like you asked for. The levels of these challenges only range from 0 to 5, but it shows what they should be, what sort of situations could complicate the task and make it more difficult, and at what level the challenges can be viewed as trivial challenges.

This other thread is also relevant:

Bounded-Accuracy.

Which has:

pauljathome wrote:

I think I'm missing something.

The difference between trained in a skill and legendary is only +3..

Which hardly seems worth much :-(

I get that it also opens up the ability to buy some special abilities. So, I'm looking at diplomacy as an example. At level 15 I can buy an ability to let me make diplomatic checks quicker.

So, my legendary diplomat is only a little more likely to succeed than the equally charismatic sorcerer who is barely trained in the skill. But I can do it quicker? And that is the ONLY effect of being legendary?

Doesn't feel all that legendary to me :-)

Oh, and my legendary diplomat STILL has a 5% chance of failing at a DC 5 check :-).

Pan wrote:

To bring this back to skills, in 5E with BA (bounded accuracy) your skill is training plus proficiency plus stat mod. The difference between high levels and low is pretty minimal. The GM decides when things are possible or not for various levels of training.

In PF2 you add level to all that so the gap between a level 1 and level 20 is much larger. On top of that, there are feats that are supposed to gate what different characters can do.

Really the issue isnt BA but the universal progression system. It is easier to understand and design for, but is overall homogenized and awkward in situations like this.

Benjamin Medrano wrote:

In the game master section on DCs, they actually talk about some checks possibly requiring certain proficiencies to even attempt, and give examples of which would be the most common. They do state that you should be wary of making the requirements too high for a level, though (i.e. not putting Master on a level 6 or lower check, or Legendary on a level 14 or lower). Also, if I remember right, there were some mentions of snares only being able to be seen if the opponent had certain a perception proficiency, but I don't remember exactly where.

Now my personal opinion:

The concept of bounded accuaracy is something I like: the discrepancy between high and low levels is not so big. In the context of skills, it makes building skill-based challenges easier.

Comparing to D&D 5e, go DC 5 for very easy and +5 up to DC 30 (nearly impossible). A 1st level will usually have a bonus of +5 in a skill, while a 20th level has +12. So, even a high level character has some trouble with doing a nearly impossible task without magical aid, but it is still feasible. This makes coming up with skill challenges very straight forward

I am found of this, because I like my skill challeges requiring players creativity to make it easier (lower DC, or have an advantage), instead on relying only on the roll of a dice.

The way the skills scale in PF2, you have to fine-tune each challenge according to players' level by consulting table 10-2 on page 336. By the end of the day a GM will be making the same task with arbitraly higher DC, even though it is advised against, because it is easier and faster. If you are carefully preparing the adventure, that may not be a big of a issue (nonetheless I find consulting tables very boring), but coming up with challenges on the fly mid-session will be a problem, making the game slower-paced.

What can be done? I suggest applying the bounded accuracy concept for skill. How can this be done (maybe a golden rule?).

[willwrk4internet]

I agree that the current system of "escalating DCs" in the playtest is overly-complicated. The table on page 337 is maddening. It reminds me of the THAC0 tables from AD&D2e. It's not well explained thematically, and it could lead to abuse (hopefully unintentional) abuse by the players or the GM. As other's on the linked posts have pointed out, the GM may not remember what DC was set for a task between play sessions, and the escalating DCs table is not flat enough to easily remember your logic in picking a DC. There's also the problem of making the DCs secret, which might cause some distrust among the players.

This system of escalating DCs makes me uneasy about GMing a game. The hardest part of GMing (in my opinion) is adjudicating rules that aren't in the rulebook. When your players ask to do something unconventional, you should be able to shoe-horn it in to an existing rule in a way that both rewards creativity and is fair to others who might try the same thing. But then you have to come up with a DC for this unconventional rule, and the system breaks down.

I do agree that a system of bounded accuracy would be nice. I don't love the way that D&D 5e lumps everything unconventional into an ability score check. In my opinion, ability checks are not the answer - skills (and maybe feats) are the answer. But the current treadmill of +1 per level is anything but bounded, and the system of adjudicating DCs suffers mightily because of it.

[Corwin Icewolf]

see, but there's already a system where low levels in large numbers are still a threat to higher levels. D&D 5th edition. If I wanted to play in such a system, I'd go and play 5th edition. alas, I do not wish to play in such a system. instead I would like to play in a system where at high levels I can do awesome and nearly impossible things like talk a dragon out of eating a town, or slicing clean through an iron golem with a greatsword.

And then I would punch a mountain! In space!

No, some of us would rather keep the over the top awesomeness in the system. It might be unrealistic, but what's so great about dumb old reality anyway?

[pedroph]

I've GM'ed 2 games in the last 10 days. My players told me something similar to what Corwin Icewolf said: high level play is about super-human, magic-like powers. The power to defy reality should not be unique to spellcasters. That is why there is a feat (for rogues, I think) to ignore damage from falling, no matter the height.

It all boils down to mentality. The purpose of the skill scaling is that the challenges should become more epic each level. That's why I disliked it at first. I prefer low-fantasy games, while pathfinder is presented as high-fantasy.

When a challenge is presented, those trained in the appropriate skill can overcome it, those untrained should find other means to overcome it, like casting a spell, using a magic item, etc. That makes an interesting game, forcing the players to use their creativity and to spend their resources.

I think that the present system makes it very easy to scale monsters and challenges. One must just become used to it (and have a quick table for reference - one that is easier to read than the one in the Rulebook).

In high levels the +3 of legendary seems to make little difference, compared to an untrained character, it is +5 points higher, which is significant. What truly bothers me is the excessive number addition and subtraction a player must do, but that can be ignored if one uses a digital character sheet.

[LordKailas]

I don't like the skill system as presented for many of the reasons mentioned above. If they are bent on keeping it though, I did see one suggestion that might help and that would be to limit the proficiency bonus you can have to a particular skill based on training.

Something like untrained +5, trained +10, expert +15, master +20 and legendary no limit. In this way dc's don't have to be scaled to the level of the PCs. It also adds some realism in that, sure over the course of your adventuring career you will pick up "a little" knowledge. +15 seems rather excessive if you are literally untrained in the skill, +5 on the other hand? ok sure that seems reasonable. You could probably even remove the bonuses and penalties associated with these ranks and it would re-enforce the idea that becoming trained or an expert in a particular skill "unlocks" your abilities in that skill.

[pauljathome]

I was in the process of creating my own thread but it is probably better to use this one.

The current skills system is very close to what I want in the game. But unless it gets some changes (many on the order of tweaks) it is going to be an abject failure, at least for me.

I think skills is an area where the whole is greater than the sum of the parts so I'm going to discuss various different topics in this post.

I'll explicitly note here that I think the current system a vast improvement on the PF1 system. But I think some tweaks are absolutely essential and I'd like to see those tweaks actually playtested.

The following is based on actual playtest experience together with years of PF1 experience (running and playing the game) and lots and lots of experience in other games (including D20 variants such as Starfinder and D&D 5)

I'm going to list what are my goals for the PF2 Skill system and how the current system needs to be altered to fit them. I Think that my goals are very close to Paizo's goals but I can't be sure. Note that the goals can overlap and somewhat conflict with each other

• Proficiency Ranks should MATTER.
  Its pretty clear that in the current design the main benefit for increasing proficiency ranks is to open up skill feats. That is fine as long as the feats are worthwhile. To me, in order to be worthwhile a feat has to

  Be cool and fun, Mechanically, be something that actually comes up a reasonable amount of time, Not be something that people would just assume to be part of a high skill, Be worth the cost

  Right now, the consensus on the boards seems to be that most of the feats are failing.

  I'll use Legendary Diplomat as an example. It IS cool. In Mutants and Masterminds, I happily spent 1 (out of 150) build points for something similar. But its not worth a skill feat. And it will rarely come up
  

• Characters who focus on a skill should ACTUALLY get better, but NOT trivialize skills

  This is one place where I'm not sure that Paizo and I have the same goal. I think that if a character totally focuses on a skill they should actually improve at level appropriate challenges. However, they should NOT get to the point where level appropriate challenges become trivial.

  Under the current rules this doesn't happen. Even in the one skill that the character is focusing on they can barely stay even. The skill DCs seem to be set with that hyper focused character in mind.

  Solving this is just a simple matter of slightly tweaking the DC numbers. They're currently CLOSE, but should go down a little. About 2 for High through Extreme difficulty seems about right to me.

• Characters who merely dabble in a skill should still be somewhat viable in that skill

  This has been stated to be a goal and the current system does a pretty good job of achieving it. I'm assuming that the vast majority of characters will leave most skills at "trained" since all their advances will be used on the 1-3 skills that they want to shine in. Presumably, even for secondary skills, they WILL still spend stat boosts and they WILL buy skill boosting items once the latter become cheap enough. The current skill DC table indicates that a character will have about a 50% chance on low DCs and about a 35% chance on high difficulty tasks. This seems about right to me

  They will not be able to do activities gated by proficiency rank but that is working exactly as designed.
  

• Highly skilled characters should not come off as bumbling idiots.

  The issue here is two fold. Primarily, the automatic critical failure on a 1 means that characters WILL come across as idiots. The character who has invested heavily in a skill will be taking every opportunity to use that skill and therefore, ironically, fumble more often than others.

  As currently written, the Assurance feat does almost nothing to address this. First, one has to choose to use it in advance. Second, when facing a level appropriate challenge the Skill DC table tells us that characters are only ever going to be assured when doing trivial or low difficulty tasks. And not always even then

  I'd suggest altering Assurance to something like : A character with Assurance does not automatically fail on a 1. In addition, the lowest result they can get on a d20 is 5 (any rolls lower than a 5 are treated as a 5). This would be worth taking, it would still mean characters COULD fumble if they tried something particularly difficult and it would NOT make it trivial to bypass level appropriate challenges.

  A secondary issue is that the difference between legendary and trained is, numerically, just too small. Even including stat modifiers and item bonuses the difference is rarely going to be as much as 5. Which means that the dabbler is, quite often, going to beat the expert at their own game. Its hard to feel like you're a great diplomat when VALEROS beats you (which happened in the level 5 adventure, despite my rolling a 10).

  The current system is quite CLOSE to what I'd like to see, the variation is just a little too low. I'd suggest that having something as a signature skill automatically gives you an additional +1 at expert and an additional +2 at legendary.

• Some classes SHOULD be better than rogues at some things

  It seems wrong to me that the Rogue is going to be able to match or exceed classes in their own schtick. The Bard should have the option of being the best diplomat, the Wizard the best Arcanist, etc.

  I'd suggest adding Class Feats that allow a bonus on skills. Class feats are valuable, so something like (for a Bard) - Gain a +1 in diplomacy AND a diplomacy skill feat.
  

[Quijenoth]

I think we need more proficiency ranks and with it more definitions on applying those to encounters.

Additional Ranks
Currently we have

• Untrained (1)
• Trained (1)
• Expert (3rd)
• Master (7th)
• Legendary (15th)

This is based on the idea that you gain a skill increase at every odd level after 3rd.

What I propose is to increase the frequency

• Untrained (1)
• Trained (1)
• Journeyman (3rd)
• Expert (7th)
• Master (11th)
• Grandmaster(15th)
• Legendary (19th)

(personally I think legendary sounds too high for 15th level so used Grandmaster and moved legendary to 19th, but you could move legendary back to 15th and make 19th Mythical, Divine, Fabled, or perhaps "legen - wait for it! ——— dary" :) )

This opens up a broader range of speciality for characters without sacrificing other skills in the process and would give those that specialise a much more defined and customized skillset.

It would also then become a major part of encounter building. Providing a proficiency rank to encounters helps GMs understand the difficulty of that encounter while also rewarding those who exceed that difficulty in the fields their characters specialise.
A Journeyman encounter with some goblin guards would give the expert stealth rogue a +2 bonus to sneak past them or an master deception bard a +2 bonus to distract them.
The goblin king encounter, However might be a Master encounter. The rogue would have more difficulty sneaking past than the bard has to distract him but both can still attempt it.

Finally skills should introduce trivial levels.
Every skill action has a trivial level, for example Treat Poison might have a trivial of master. Therefore any character who is at least master in Medicine receives a +2 bonus to Treat Poisons. As an option you could make the bonus only +1 and increase the bonus per tier you exceed the trivial but I think that would be too much on the record keeping and just slow things down.
Alternatively skills that increase with the DC by level could use the progression to section off the skill.
Craft: Identify alchemy

• Level of Item (Skill Difficulty)
• 1-5 (Journeyman)
• 6-10 (Expert)
• 11-15 (Master)
• 16+ (Grandmaster)

An Expert Crafter would get a +2 identifying an alchemical item of level 1 to 10.

Note this doesn't replace the Proficiency-Gated Tasks involved with skills and this should still be a thing.