| 6 people marked this as a favorite. |
http://www.5mwd.com/archives/comic/ostog-the-untenured
As has been mentioned by many, many people on these boards, adding +1/level to everything creates some weirdness and some serious problems.
I looked forward to this playtest for five months, and often defended it against the dubious, but unless it changes substantially on a structural level, I'm not even sure I can house rule it into something I want to play.
But I did think of a houserule I could live with to solve the +1 to everything treadmill, and I'd love to hear critique or suggestions from anyone.
One popular solution seems to be just stripping +1/lvl off everything, monsters and PCs, and just using proficiency bonuses. I think this is too 5e for my tastes, which are closer to PF1/3.5.
So here's my proposal.
The +1/level bonus is granted and capped by your proficiency level, as follows:
UNTRAINED
No penalty but no +1/level either.
TRAINED
+1/level (MAXIMUM +5)
EXPERT
+1/level (MAXIMUM +10) +1
MASTER
+1/level (MAXIMUM +15) +2
LEGENDARY
+1/level (no maximum) +3
The above would apply to skills, armor proficiency (including unarmed defense, but only monks & barbarians would be trained in that), weapons, perception, spell rolls. (I'm tempted to leave saving throws as is, but they might also use the same rules as everything else.)
You would still be granted, for free, all of the skill boosts and proficiencies of your class; but there would be a new type of feat, the Cross-class Feats, which anyone can take but require you to use a class feat on them (not a skill or general).
These feats would include all of the armor and weapon proficiencies (trained, expert, master, legendary, individually, with prerequisites for the higher ones), as well as skill proficiencies, perception, and maybe even saving throws. The price for this is you're using your precious class feats for this stuff so, yeah, your wizard can become legendary in dagger but at the price of all the cool spell feats.
The multi-class feats would have to be adjusted, given that they'd no longer be the sole way to get extra weapon proficiencies, say.
I don't think this would change the game much at low-levels, and probably not much at mid-levels either. By high levels, though, there would be some serious disparities in the abilities between characters. You would not have situations where the wizard outfights the fighter, or the barbarian tutors the cleric on religion. The naked unequipped wizard would not have a monk-like ability to avoid harm.
As written, this would make high-level adventures a lot more difficult--which I like--or the high level monsters would need to toned to accommodate the character specialization. Or the monsters could be rewritten so as to be explicit what exactly they're trained in.
This would be a way I think I could live with the +1/level.
| 1 person marked this as a favorite. |
The major problem I can see with that is it still keeps the issue of the rest of the party twiddling their thumbs while the Rogue goes on a solo adventure, possibly even worse than PF1e had it. Granted, that problem still does exist to a certain extent in the playtest (mostly because monster skills are so inflated that it's detrimental to do things without being optimized), but that's a tuning thing and not a basic systemic issue.
| 6 people marked this as a favorite. |
The untrained 20th level barbarian might have a +18 in Arcana. However, he cannot:
-Identify magic items
-Read magic
-Automatically detect magic auras just by looking
-Trick a magic item into working if not normally usable by him
The untrained 20th level Wizard might have a +18 in Athletics. However, he cannot:
-Use Athletics to disarm an opponent
-Make a jump without a running start
-Fight effectively underwater
-Climb a slope one-handed
-Parkour up a wall via Wall Jump
-Swim like a fish via Legendary Swimmer
So even without a rules change, an untrained PC is still not better than someone trained (or higher) in a skill.
As far as your house rule suggestion, it could certainly work. The problem for me is that it adds back the same bug to me, but which was a feature for you: namely that some characters will still wind up completely sitting out on certain sections of a scenario. Perhaps the spellcasters could assist with magic, but if the spellcaster is unavailable, or doesn't have the spells needed to compensate, or the PC for whom a certain skill challenge is the best fit is unavailable for a game, then you have a challenge which is undefeatable by the group. This happens in my groups more often than one would think - a player is sick and has to miss out, or the spellcaster didn't prep the right spells today and doesn't have scroll backups, etc.
It also reintroduces something which, to me, felt just as unrealistic as the +18 Athletics wizard: The character who at 7th level is untrained in something, takes all their skill points at 8th level and spends a few weeks training, and next thing you know they're just as good of a linguist/blacksmith/Waltz champion as the person who had been putting points in since level 1. I can handle the Barbarian who had picked up bits and pieces of Arcana over his years of hanging out by the campfire with the wizard; It's harder for me to swallow the guy who crams overnight and is as good with a skill as an experienced practitioner.
| 3 people marked this as a favorite. |
So, my 19th level arcanist, Expert Stealth, but taking Deception, Sleight of Hand, and Arcana as her only legendary skills (you simply can't get more)... is unable to stealth past any level appropriate encounter, and probably anything five levels below her too.
PCs already have a horrible time keeping skills up with monsters in PF2. This would ruin them even further, making them utterly incompetent at anything past 3 skills at high levels.
unable to stealth past any level appropriate encounter
That would depend on whether the monsters also had their bonuses capped.
| 1 person marked this as a favorite. |
If you carefully read the rules (which I am only about 75% through) a lot of the issues raised about Ostog the Untenured are people playing the skill system wrong and assuming that bonuses alone determine likelihood of success.
Most examples people give are things that are already gated by a feat or a proficiency tier, or else they easily can be made so without much trouble.
The +level bonus to proficiencies is probably an all or nothing system as long as the game uses its +/-10 critical system. You just can't have the kind of variance you are used to in PF1 for attack rolls, Defenses (saves and ACs), or even skills that might have life or death consequences (like acrobatics and athletics).
The numbers are still not right and PF2 is definitely skewed to hard mode right now, but fixing it is a matter of soft adjustments, not drastic ones, at least not with scrapping the critical success system and possible the three action economy.
The other important thing that's locked behind trained competency: Practice a Trade or Stage a Performance. Standard uses of Perform (untrained) are good enough to impress someone or prove that you're decent, but it's not good enough to actually put on a performance if they have to look at you for more than a few seconds.
Similarly, Read Scripture, Read Esoterica, Treat Disease, actual Crafting, and Tracking. If you're not trained, doesn't matter how high level you are, you can't do those.
| 4 people marked this as a favorite. |
The other important thing that's locked behind trained competency: Practice a Trade or Stage a Performance. Standard uses of Perform (untrained) are good enough to impress someone or prove that you're decent, but it's not good enough to actually put on a performance if they have to look at you for more than a few seconds.
Similarly, Read Scripture, Read Esoterica, Treat Disease, actual Crafting, and Tracking. If you're not trained, doesn't matter how high level you are, you can't do those.
One man's "trained competency" is another's "arbitrary gatekeeping."
One man's "trained competency" is another's "arbitrary gatekeeping."The other important thing that's locked behind trained competency: Practice a Trade or Stage a Performance. Standard uses of Perform (untrained) are good enough to impress someone or prove that you're decent, but it's not good enough to actually put on a performance if they have to look at you for more than a few seconds.
Similarly, Read Scripture, Read Esoterica, Treat Disease, actual Crafting, and Tracking. If you're not trained, doesn't matter how high level you are, you can't do those.
So what I'm hearing is that you'd have a problem either way - either you complain because everyone can do everything at full competency, or you complain because "arbitrary gatekeeping".
One man's "trained competency" is another's "arbitrary gatekeeping."
The GM always had power over when an outcome could be determined by a die roll or not. Arbitrary or non-critical situations suffer under PF2 because it is less easy to have competency determined purely by bonus number. Critical situations have benefited drastically by not having variance that leads to massively higher fail rates (and thus an even larger shift in critical fail rates) for unoptimized characters for things characters have to do. Personally I think they made the right choice prioritizing important situations over arbitrary and less game defining situations.
| 2 people marked this as a favorite. |
So what I'm hearing is that you'd have a problem either way - either you complain because everyone can do everything at full competency, or you complain because "arbitrary gatekeeping".One man's "trained competency" is another's "arbitrary gatekeeping."The other important thing that's locked behind trained competency: Practice a Trade or Stage a Performance. Standard uses of Perform (untrained) are good enough to impress someone or prove that you're decent, but it's not good enough to actually put on a performance if they have to look at you for more than a few seconds.
Similarly, Read Scripture, Read Esoterica, Treat Disease, actual Crafting, and Tracking. If you're not trained, doesn't matter how high level you are, you can't do those.
Then you might need your hearing checked. I never said that I do or would have a problem either way.
Either the final 2nd edition looks like fun to me and a majority of those I game with, and so we play it, or it does not, and we do not. Either way I have zero problems.
My post was simply pointing out the argument against it. There are myriad ways of handling skills, and in a system with such tightly controlled mathematical bounds, some sort of artificial enforcement of boundaries is necessary.
I personally would have preferred to see a system that handles increasing skills and their more difficult uses with an increased opportunity cost at higher ranks, but that isn't what is in the playtest. I am not a huge fan of bounded accuracy at all, as it introduces a level of gamism with which I am uncomfortable.
With the math as it is now, it's not playing great in my experiences, but I can see it being a usable skill system if the numbers are tweaked, even if it is not my cup of tea.
| 1 person marked this as a favorite. |
thflame says this very well, IMO, and I concur.
I've been in four playtest sessions so far and every one rolls for every single check, because untrained people's competency is so slightly behind the allegedly trained. Not once so far has a cleric rolled highest at a religion check, and we had an amusing but implausible role-play session where the goblin alchemist tutored the cleric of Saranrae on the tenets of her faith.
Amusing, but not really what I'd like as the norm . . .
What if there were skill feats like this:
Expert Swimmer
You have expert proficiency in all Athletics skill checks pertaining to swimming. You remain untrained in all other aspects of Athletics.
thflame says this very well, IMO, and I concur.
I've been in four playtest sessions so far and every one rolls for every single check, because untrained people's competency is so slightly behind the allegedly trained. Not once so far has a cleric rolled highest at a religion check, and we had an amusing but implausible role-play session where the goblin alchemist tutored the cleric of Saranrae on the tenets of her faith.
Amusing, but not really what I'd like as the norm . . .
What if there were skill feats like this:
Expert Swimmer
You have expert proficiency in all Athletics skill checks pertaining to swimming. You remain untrained in all other aspects of Athletics.
I have been playing this way too, but I think that I have not been paying attention to the way exploration mode is supposed to be handled. It has been made clear to me that things like perception are supposed to only allow one roll per party, and given that, it will be the character that has the highest bonus that the party wants making the roll.
Also, where there is a consequence for failure, it is probably better to have other players assist and increase the odds of success of the most likely to succeed, rather than laugh when they fail, and have 4 characters keep rolling their own checks until they succeed, because there is some kind of consequence otherwise.
It is nice that the +2 from aid another is usually going to apply in situations where it radically increases the odds of getting a critical success or avoid a critical failure, which are bigger deals on skill checks than attacks.
| 3 people marked this as a favorite. |
As far as your house rule suggestion, it could certainly work. The problem for me is that it adds back the same bug to me, but which was a feature for you: namely that some characters will still wind up completely sitting out on certain sections of a scenario. Perhaps the spellcasters could assist with magic, but if the spellcaster is unavailable, or doesn't have the spells needed to compensate, or the PC for whom a certain skill challenge is the best fit is unavailable for a game, then you have a challenge which is undefeatable by the group. This happens in my groups more often than one would think - a player is sick and has to miss out, or the spellcaster didn't prep the right spells today and doesn't have scroll backups, etc.
If you have a Dex12 Dwarf fighter with no background in stealth and clanky armor on, and he needs to get past a set of reasonably competent guards, then your narrative has a situation which is undefeatable. You can change the rules so that the results declare success despite circumstances. But that just makes the rules blind to the story they are controlling.
I'm not saying there are not people who complain about being stuck at the guards. But I am saying that it isn't simply a matter of "bug to you / feature to me". You are actively replacing one problem with a different problem. And, honestly, I predict that you will find this wonkiness to grate on you more and more as time goes by.
If something was undefeatable before, there was a reason for that. And if the rules say "you can now defeat undeaftable things 'just because'", then you may as well say "any time you are doing anything against a more or less level appropriate foe, roll D20 and you succeed on an 8+". I realize that sounds like hyperbole. But, seriously, walk up to "the man on the street", describe the scenario and ask them the odds. If they say it can't be done then that should be meaningful to the story when it involves characters not built to do superhuman things in that arena.
It is a real poke in the eye to "storytelling" as the centerpiece of gaming. And, again, I'm not saying everyone has to want storytelling as the centerpiece. But a lot of people really do. And this casual blow-off response seems to disregard that.
It also reintroduces something which, to me, felt just as unrealistic as the +18 Athletics wizard: The character who at 7th level is untrained in something, takes all their skill points at 8th level and spends a few weeks training, and next thing you know they're just as good of a linguist/blacksmith/Waltz champion as the person who had been putting points in since level 1. I can handle the Barbarian who had picked up bits and pieces of Arcana over his years of hanging out by the campfire with the wizard; It's harder for me to swallow the guy who crams overnight and is as good with a skill as an experienced practitioner.
Seriously, wouldn't a house-rule limiting the dumping of skill points have less system-shaking impacts than everyone is good at everything?
| 3 people marked this as a favorite. |
The major problem I can see with that is it still keeps the issue of the rest of the party twiddling their thumbs while the Rogue goes on a solo adventure, possibly even worse than PF1e had it. Granted, that problem still does exist to a certain extent in the playtest (mostly because monster skills are so inflated that it's detrimental to do things without being optimized), but that's a tuning thing and not a basic systemic issue.Part of this could be fixed by giving out enough resources for learning skills that players that WANT to be able to contribute can spec into it without hindering their character.
It's one thing when you have to sit on the sidelines because the game doesn't give you enough resources to be able to contribute. It's a completely different feeling when you COULD have reasonably invested into a particular skill, but chose not to in favor of something else.
Indeed. Perhaps some kind of modular or granular point-based skill system would work, where investment into a given skill is what gives you bonuses instead of a flat per-level boost.
Nah, that's just silly.
| 1 person marked this as a favorite. |
Indeed. Perhaps some kind of modular or granular point-based skill system would work, where investment into a given skill is what gives you bonuses instead of a flat per-level boost.
Ah, if only Paizo had had some exposure to this idea before starting development on PF2 they might have thought to include it! Perhaps they could also have a similar system for attack bonuses, where classes that are more adept at combat would have a higher likelihood of their attacks hitting, where classes that are less adept would have a lower likelihood?
| 1 person marked this as a favorite. |
Indeed. Perhaps some kind of modular or granular point-based skill system would work, where investment into a given skill is what gives you bonuses instead of a flat per-level boost.Ah, if only Paizo had had some exposure to this idea before starting development on PF2 they might have thought to include it! Perhaps they could also have a similar system for attack bonuses, where classes that are more adept at combat would have a higher likelihood of their attacks hitting, where classes that are less adept would have a lower likelihood?If only Paizo had made such a system instead of copying it from someone else, warts and all.
If only there were some sort of system in this new edition whereby at certain intervals, say every two levels, there were ways of specalising in chosen skills. This hypothetical feature could open up new uses for the skills or making you better at existing features of the skills. If only Paizo had thought of such a novel feature for their new edition of their game, to tie in with the rest of it playing by different rules.
That sounds nice in theory, but... that doesn't actually address my criticism of it - the fact that two characters of similar level are rolling at relatively close bonuses even though it doesn't narratively make sense that that should be the case. Skill feats allow new uses - although some like Pickpocket would be better folded into proficiencies - but that doesn't change the fact that everyone gets better at all skills as they level no matter what.
| 1 person marked this as a favorite. |
That sounds nice in theory, but... that doesn't actually address my criticism of it - the fact that two characters of similar level are rolling at relatively close bonuses even though it doesn't narratively make sense that that should be the case. Skill feats allow new uses - although some like Pickpocket would be better folded into proficiencies - but that doesn't change the fact that everyone gets better at all skills as they level no matter what.
Honestly I'm not too torn up about the leveling aspect of it, a Barbarian who's made it that far in life should have picked up a few tips and tricks for dealing with the city guards in a diplomatic manner, and the Wizard probably learned a few tricks from hanging around with the Rogue for too long.
The problem is that people are passively *too* good at things. The Barbarian should be able to hold a conversation, but he shouldn't be able to out-diplomacizing the Bard. The Wizard should be able to pick a lock in a pinch, but he shouldn't be disarming incredibly difficult traps and stealing siege weapons in plain sight (barring magic, of course).
| 1 person marked this as a favorite. |
And that is why proficiency levels are more important than just the numerical bonus they add to your roll.
| 2 people marked this as a favorite. |
The untrained 20th level barbarian might have a +18 in Arcana. However, he cannot:
-Identify magic items
-Read magic
-Automatically detect magic auras just by looking
-Trick a magic item into working if not normally usable by himThe untrained 20th level Wizard might have a +18 in Athletics. However, he cannot:
-Use Athletics to disarm an opponent
-Make a jump without a running start
-Fight effectively underwater
-Climb a slope one-handed
-Parkour up a wall via Wall Jump
-Swim like a fish via Legendary Swimmer
That's the standard defence of this, that "Proficiency is a gate for certain activities which is more important than your +bonus" but looking at the actual rules, I don't see that limitation.
The skills section lists the activities for each skill, and some of them are listed as trained only. None of them, in the entire chapter, are listed as expert only, master only or legendary only. That means it's on the GM to say, "No, you can't do that," without being able to point to a good reason why not.
| 1 person marked this as a favorite. |
The GM always had power over when an outcome could be determined by a die roll or not. Arbitrary or non-critical situations suffer under PF2 because it is less easy to have competency determined purely by bonus number. Critical situations have benefited drastically by not having variance that leads to massively higher fail rates (and thus an even larger shift in critical fail rates) for unoptimized characters for things characters have to do. Personally I think they made the right choice prioritizing important situations over arbitrary and less game defining situations.The other important thing that's locked behind trained competency: Practice a Trade or Stage a Performance. Standard uses of Perform (untrained) are good enough to impress someone or prove that you're decent, but it's not good enough to actually put on a performance if they have to look at you for more than a few seconds.One man's "trained competency" is another's "arbitrary gatekeeping."
I have the distinction of GMing a game where the PCs actually staged a performance. I think this event is rare.
The 4th-level module Lords of Rust in the Iron Gods adventure path mostly takes place in Scrapwall, a shantytown of refugees and bandits in the middle of a gigantic junkyard. The party needed to investigate the Lords of Rust in Scrapwall, but they chose a very low profile method. They moved into Scrapwall, pretending to be refugees.
I had to create the daily life of Scrapwall inhabitants, and even held a beer festival with drinking, brawling, and gambling. The party skald decided to reciprocate and the party put on a concert. The crafters in the party made a stage out of scrap, the barmaid in the party arranged for beer and grog sold at cost (yes, she had a rank in Profession(Innkeeper)), the fighters in the party served as perimeter defense (Scrapwall has wandering monsters) and crowd control, etc.
Um, under Pathfinder 2nd Edition rules, would those other party members have been able to help stage the performance of the skald? Let me check the wording of Stage a Performance ... oh, Stage a Performance is simply a variant of the Lore skill Practice a Trade. It has nothing in it about the work involved in holding a concert. Since the party did not make a profit on their concert (most attendees were dirt poor), the official Stage a Performance skill would not apply to the performance they staged.
The Stage a Performance gatekeepers are guarding a gate to nowhere. Also see Ediwir's thread: The (dancing) Elephant in the Room: Performance.
| 1 person marked this as a favorite. |
I suppose we can now term the OP's argument as the "Ostog Fallacy"? :/
Though personally I do think that gating should be more explicit in the rules. Perhaps anchor it with:
Training: things that require a period of discipline, learning, and practice. This could be attained through schooling, some adventuring.
Expert: a distiguished professional in your field. For some skills you are considered the best in a small town that people go to. This can be attained through exclusive education (such as training under an expert) or an extended period of experience. In modern life you are the person who reada books on chess and competes regularly in tournaments, or gets a PhD training under a professor, or becomes a professional athlete.
Master: you are the premier person in your ability in your nation or region. People travel from the countryside to get your services. You may have a rival who also is a master in your large city or metropolis.
Legendary: you are one of a kind. You may have a rival or two in the world at your ability. Tales of tour exploits are told about you in our region/nation after you die.
I'd like to see the untrained penalty half the bonus from level either with or instead of the -2. It's small, simple and hammers down that untrained characters aren't as good as even lightly trained ones. Plus, having more clear rules on what level TEML level allows and doesn't allow would be nice.
The slightly more extreme option for me would be having Untrained be +1/2 level, trained to be +3/4 level, and expert be +level (with higher levels getting more flat bonuses) but I'm not sold that this option would be better or easier.
I think I know the best solution for this. It will never be taken up, but I think I'd house rule PF2 this way.
Change 1: Eradicate Level Bonus from all parts of PF2 except for Resonance (in its current incarnation).
Change 2: Change the Proficiency Modifier to as follows:
- Untrained: -2
- Trained: +0
- Expert: +1
- Master: +3
- Legendary: +6
Change 3: Change the d20 roll to a 2d10 rollThis will allow for many different enemies being relevant at many levels, make proficiency more important than ability modifiers, and provide a much better distribution in dice result in order for the bonuses to be much more impactful than they are on a d20.
Just my thoughts.
Change 1: I have some thoughts on Resonance that would remove the level bonus to it, too.
Change 2: I love the triangular numbers, 1, 3, 6, 10, 15, ... :-) The gaps between them add the possibility of unique proficiency bonuses for signature skills:
- Untrained: -2 (Signature -1)
- Trained: +0 (Signature +1)
- Expert: +1 (Signature +2)
- Master: +3 (Signature +4)
- Legendary: +6 (Signature +8)
Change 3: Sorry, but altered variance in 2d10 is a big difference in the odds. Suppose you need to roll 15 to hit a boss. On 1d20, that is 30% chance. On 2d10 that is 21%.
The untrained 20th level barbarian might have a +18 in Arcana. However, he cannot:
-Identify magic items
-Read magic
-Automatically detect magic auras just by looking
-Trick a magic item into working if not normally usable by himThe untrained 20th level Wizard might have a +18 in Athletics. However, he cannot:
-Use Athletics to disarm an opponent
-Make a jump without a running start
-Fight effectively underwater
-Climb a slope one-handed
-Parkour up a wall via Wall Jump
-Swim like a fish via Legendary Swimmer
That's the standard defence of this, that "Proficiency is a gate for certain activities which is more important than your +bonus" but looking at the actual rules, I don't see that limitation.
The skills section lists the activities for each skill, and some of them are listed as trained only. None of them, in the entire chapter, are listed as expert only, master only or legendary only. That means it's on the GM to say, "No, you can't do that," without being able to point to a good reason why not.
In case you haven't read them, both the Bestiary and Doomsday Dawn adventure list exactly these things for particular challenges.
| 1 person marked this as a favorite. |
If any change happens to proficiencies, it shouldn't be raw numbers, for all of the reasons that have been clearly exemplified in the many threads that deal with this issue.
Feat gating is still not perfect or even that great and looking for places where it could be done better and discussing how is a really good use of people's time.
I have some ideas:
Master proficiency feats that let you roll 2 20 and choose the highest when making certain kinds of checks (should be limited to non-combat applications). (this math works a lot better than massive bonuses for normalizing average rolls and mitigating critical failure, with out making your high end roll shoot so high that DCs must be artificially inflated).
Get rid of item bonus to skills (so that all DCs don't have to assume that you have spent the money on an item to have a decent shot at it) and instead have the items grant an extra D20 to roll for those checks. With the skill mastery feat this would give you three dice.
Let the the DCs for some skill tasks be adjusted by proficiency in related skills. This would have to be left to GM fiat to avoid excessive meta-gaming munchkin shenanigans, but for example, let a character with Master proficiency in society have trivial DC for gathering information in a city. Or Master Proficiency in Nature have an easier time climbing trees or vines in a forest if they can identify them.
The key to all of these suggestions is that they are limited in scope to specific situations and don't arbitrarily inflate the numbers of all skill checks because one character is capable of having an excessively high number all the time.
With the 2d10, proficiencies play an even larger role than they do with the d20. In addition, it helps mitigate Crit Success/Failure (in PF2 it is way too often, IMO). Tactically, it would benefit the game as well, because now those little bonuses that you help are even more important in order to hit or prevent being hit.
Getting Piazo to walk away from the D20 is a very tall order. It would be an even bigger deal than getting them to walk away from classes or attributes, as far as the legacy of the game. It might work fine as a house rule or later variant rule, but rolling 2d10 instead of a D20 is an incredibly unlikely step for the company to take. For better or worse,, it is probably necessary to focus on solutions that don't completely underwrite the single most sacred element of the game: Player declares their intended action, the GM determines the difficulty, and then the player rolls a D20 to determine the outcome.
Even though I think having so much riding on a single D20 roll is often a mistake, I am pretty sure nothing would push people away from the game faster than taking away their favorite polyhedron.
| 1 person marked this as a favorite. |
For the record, I’m fine with the +1 per level rule as is. But for a thought experiment, I would probably consider something along the following lines.
There are three pillars to skills. Raw ability or talent if you will, training, and practical experience. Using this as the basic structure I would try to allow each pillar to equally weigh in on the total number.
Your attribute, or raw talent generally goes to 20 in PF2 (there are exceptions of course), so having a good stat can give you up to +5.
Your training is your profiency I’d keep the same numbers as PF2 (minus the + per level of course). This gives a difference of untrained to legendary of 5. So this pillar matches the scale of raw talent.
Your experience comes from level. I think I would grant a +1 for character levels 1-3, +2 for character levels 4-7 and so on. This scale is pretty lose to the raw talent and training proficienciesemding with a bonus of 6
The end result is if you have all three pillars, you are the best there is. If you have two of the pillars you would likely be slightly superior to most people you meet. If you only had one (such as a high level), your overall skill total wouldn’t be better than a commoner who is both gifted and well trained which is a common argument I see.
To me this represents a true balance of key factors (excluding things such as magic items). This system still leaves a high variability of luck due to the d20 roll but it is somewhat mitigated by character who have two or three strong pillars.
Edit: one other change I would make is rename Legendary to something else. I think this single word sets a higher expectation of the benefits you should have for being in this category. IMO people are generally fine with slight bonuses from trained to expert to master, but the word legendary makes you feel like you should be absolutely awesome - and another +1 does not convey that. So maybe the proficiency levels should be untrained, trained, skilled, expert, and master. The skill bonuses don’t change - just the name does. I think this better matches the perception of the benefit you are getting.