Are Martials really that screwed?

[Xervous]

If failing a check to produce a magnitude 6 effect is significant for the L1 character rolling for his 1/20 what changes the stakes when he’s L20 rolling for a 6/20?

Does it cost health, DM fiat, healing surges I mean hit dice, do you have to buy the group pizza?

[Asisreo1]

If failing a check to produce a magnitude 6 effect is significant for the L1 character rolling for his 1/20 what changes the stakes when he’s L20 rolling for a 6/20?

Does it cost health, DM fiat, healing surges I mean hit dice, do you have to buy the group pizza?

It costs whatever the DM thinks it should cost, but the punishment should remain consistent. If the lvl1 fighter takes 10d10 damage on a failure, the lvl 20 fighter should also take 10d10. The difference is that the lvl 20 fighter can take the punishment more.

If failing the check gives disadvantage or puts the party in a situation where, like, 20 guards show up to attack. Well, a lvl 1 fighter can't quite risk fighting 20 knights even with a party but a lvl 20 fighter will probably win.

[Xervous]

For the sake of brevity I will invoke the discussion of the dc for climbing a tree to refute the notion that DM arbitrated “skills can do uncapped amazing things with DM chosen consequences” makes for a sound system that players can expect to consistently engage with. This skill driven system we are considering shouldn’t be a per table basis, that’s house rules as these are very much new rules being discussed.

[OldTrees1]

Ability checks don't need to play by the same rules as spellcasting and what they allow. It appears that you want what they can do to be a mirror to the spellcasting system.

I want quantitative and qualitative growth. I want PCs to be able to grow into and then master new checks.

I only use spell levels as an example because it is an uncontroversial claim to say a 1st level wizard can't cast a 2nd level spell but a 20th level wizard can unerringly cast a 2nd level spell. I am drawing no more comparison than the growth qualitative growth described by growing into and then mastering new abilities. There is no mirroring involved.

You've tripled the quantity of success, I see nothing wrong with that.

I see something wrong with just +10% over 20 levels.

But if you notice, I keep referencing ability checks and not skill checks and only barbarians and fighters because it's where the smallest amount of these changes occur. With proficiency, it changes from +5 to +11. With a barbarian, it goes from +5 to +13 for athletics. If we talk about rogues with their expertise, the most trained in the skill system. It goes from +7 to +17. With reliable talent, you'll never have lower than a 27. A rogue can just decide that they want a 37. I don't see how this range isn't significant enough. It isn't available to all martials, but that isn't what you wanted, is it? It's a martial that has the biggest range of using their skill checks. Barbarians and fighters are superior in damage but doesn't quite interact as well as a rogue.

And if you noticed I have split my posts in half. The left side is for ability checks and the right side is for skill checks. Because I am critiquing both the +10% for ability checks and the +30% for skill checks.

You might have noticed I have ignored the rare ability that breaks bounded accuracy. I do it because it is rare and does not cover what I am critiquing.

A) 1/3 of the time is much less unlikely than 1/10 of the time. So it is, indeed, a large improvement.

3/10th is still very unlikely and it is not a big improvement considering it spans 20 levels.
Yes it is tripling the relative chance but the objective chance is so low that it does not matter.
You might be willing to pay $1 to risk a 5% chance to get $100. But you would not pay $1000 to risk a 15% chance to get $100.

B) A 1st-level fighter does not have a 60-70% chance to succeed a level 6-7 spell. They have a 5% chance to succeed.

I was being generous about where I placed that teeny +10%. If you want it as 5%->15% then it is worthless. The 20th level fighter would not attempt and the 1st level fighter would attempt.
5% chance at a 11th level ability is worthwhile at 1st.
15% chance at a 20th level ability is not worth it at 20th, and it is merely an 11th level ability.

Proficiency handles that well, as well as ability scores. No skill ability checks max out early but it isn't until level 17 that skills max out. You increase both roughly once or twice every tier upgrade. Overall, it's a +6 difference for all proficient skills except for expertise where it's a +10 difference.

No, proficiency does not handle it well. Skills increase by +30% which is better than the +10% for Raw checks but is worthless for showing progression.

[Asisreo1]

I want quantitative and qualitative growth. I want PCs to be able to grow into and then master new checks.

Have you considered that not everyone wants that? Or cares for it either way? Your quantitative growth comes in new features as well as better ability scores. You get more HD, more ASI's, more feats, and improvements to your base chassis at every level.

I only use spell levels as an example because it is an uncontroversial claim to say a 1st level wizard can't cast a 2nd level spell but a 20th level wizard can unerringly cast a 2nd level spell. I am drawing no more comparison than the growth qualitative growth described by growing into and then mastering new abilities. There is no mirroring involved.

A 1st-level wizard can cast a 2nd level spell. It's a DC 12 ability check with their spellcasting mod. They just need a spellscroll. It's not universal. Curious, outside of spellcasting, is there any other system in 5e that actually does this? If you want spellcasting progression, you should interact with spellcasting, whether you want to reflavor or not is up to you but you keep comparing the two systems as if they aren't supposed to be different.

I see something wrong with just +10% over 20 levels.

Then you're looking at it from the wrong perspective.

And if you noticed I have split my posts in half. The left side is for ability checks and the right side is for skill checks. Because I am critiquing both the +10% for ability checks and the +30% for skill checks.

You might have noticed I have ignored the rare ability that breaks bounded accuracy. I do it because it is rare and does not cover what I am critiquing.

Do you want 1 martial to engage with it, or all martials to engage? And out of all the relevant martials (those that cannot cast spells) half of them have expertise. Even if you were to include the monk and totem barb, it's still 1/4th of all relevant martials. I fail to see how it's rare or anything less than uncommon at best.

3/10th is still very unlikely and it is not a big improvement considering it spans 20 levels.
Yes it is tripling the relative chance but the objective chance is so low that it does not matter.
You might be willing to pay $1 to risk a 5% chance to get $100. But you would not pay $1000 to risk a 15% chance to get $100.

It's $1 to risk a 5% chance to get $100 or $1 to risk a 15% chance to get $100. The risk and reward are the same, only the success rate changes. And I'd take the second one.

I was being generous about where I placed that teeny +10%. If you want it as 5%->15% then it is worthless. The 20th level fighter would not attempt and the 1st level fighter would attempt.
5% chance at a 11th level ability is worthwhile at 1st.
15% chance at a 20th level ability is not worth it at 20th, and it is merely an 11th level ability.

Calling it teeny doesn't change the fact that it isn't teeny. A lvl 20 character can take whatever punishment would occur for the missed 15% chance much better than a 5% chance character.

No, proficiency does not handle it well. Skills increase by +30% which is better than the +10% for Raw checks but is worthless for showing progression.

Are you afraid of 26+ DC rolls for things that are beyond ridiculous? It gives a feeling of progression when that door that was DC 27 is doable by level 4 but the DC 29 wall still needs to wait until level 8-12. And remember, given no punishment, they needn't roll. They could just "take20" and see if they succeed given x10 the normal time it would've taken to do it. So a lvl 12 fighter can consistently do something a lvl 1 fighter cannot if neither are under any pressure or negative consequence.

[Eldariel]

Are you afraid of 26+ DC rolls for things that are beyond ridiculous? It gives a feeling of progression when that door that was DC 27 is doable by level 4 but the DC 29 wall still needs to wait until level 8-12. And remember, given no punishment, they needn't roll. They could just "take20" and see if they succeed given x10 the normal time it would've taken to do it. So a lvl 12 fighter can consistently do something a lvl 1 fighter cannot if neither are under any pressure or negative consequence.

Aren't we just buffing casters this way? Guidance, Bardic Inspiration, Enhance Ability and Peerless Skill just get better - which is nice but doesn't feel like it addresses the actual issue.

[Xervous]

And how do you take 20 when there’s arbitrary consequences for failure?

[KorvinStarmast]

Was it this post: https://forums.giantitp.com/showsing...ostcount=1094?

Very likely. I only get to visit now and again, so to catch up in a thread I often check posts that interest me using multi quote, and then reply to fewer than all of them. So in my clean up, I appear to have mixed up the 'quoted by' blocks ... not the first time this has happened.

[Asisreo1]

Aren't we just buffing casters this way? Guidance, Bardic Inspiration, Enhance Ability and Peerless Skill just get better - which is nice but doesn't feel like it addresses the actual issue.

You're buffing the assistance a caster can provide, encouraging team play.

[Asisreo1]

And how do you take 20 when there’s arbitrary consequences for failure?

It's not actually taking 20, but it has exactly the same effect.

DMG, pg 237 under the "multiple ability checks" section states that if there is only a time cost associated with a challenge, the DM assumes the character takes ten times as long and successfully does the task, provided it wasn't impossible.

If there is a cost that is more than time, then the character should roll repeatedly and take the punishment each time they fail or roll once and be gated from trying it again.

[Xervous]

It's not actually taking 20, but it has exactly the same effect.

DMG, pg 237 under the "multiple ability checks" section states that if there is only a time cost associated with a challenge, the DM assumes the character takes ten times as long and successfully does the task, provided it wasn't impossible.

If there is a cost that is more than time, then the character should roll repeatedly and take the punishment each time they fail or roll once and be gated from trying it again.

I’m not asking for a deflection to a rules quote. I’m asking how your proposed system handles the supposed taking twenty on skill/ability rolls that have arbitrary failure states as per its usage of gm whimsy in place of hard set rules.

[Asisreo1]

I’m not asking for a deflection to a rules quote. I’m asking how your proposed system handles the supposed taking twenty on skill/ability rolls that have arbitrary failure states as per its usage of gm whimsy in place of hard set rules.

The same way that success is arbitrary. Of course, assuming that the attempted action is not something that is part of the examples in the book. Even the examples in the book don't always give clear failure states. Failing a tracking check means the party must spend time re-finding the tracks. What's the price for failing to pick a lock? The punishment for failing a search? The consequence of failing to break an object? These are up to DM interpretation.

The DM has the power to control the world and what a fail state looks like. It usually comes down to common sense, though. If you fail to pick a lock, perhaps one of your lockpicks break on your thieves' tools. If you fail a search check, maybe you move the item unknowingly to a location where you're even less likely to find it. If you fail to break an object, perhaps you succeed in breaking the object but hurt yourself in the process (success at a cost). But if there are no such dangers present, they can just opt to take ten times the amount of time needed for the successful outcome, not needing to worry about failure. Perhaps they don't break their lockpick because they are now more careful and taking their time. Perhaps they search more thoroughly and there's less throwing objects across the room. Perhaps they figure out a way to break the object in such a way to avoid hurting themselves.

It's always been up to the DM for this stuff and the game has been better for it since the DM rarely has to worry about referencing a table to see the DC, the success clause, and the failure clause for every single type of challenge.

[Xervous]

The same way that success is arbitrary. Of course, assuming that the attempted action is not something that is part of the examples in the book. Even the examples in the book don't always give clear failure states. Failing a tracking check means the party must spend time re-finding the tracks. What's the price for failing to pick a lock? The punishment for failing a search? The consequence of failing to break an object? These are up to DM interpretation.

The DM has the power to control the world and what a fail state looks like. It usually comes down to common sense, though. If you fail to pick a lock, perhaps one of your lockpicks break on your thieves' tools. If you fail a search check, maybe you move the item unknowingly to a location where you're even less likely to find it. If you fail to break an object, perhaps you succeed in breaking the object but hurt yourself in the process (success at a cost). But if there are no such dangers present, they can just opt to take ten times the amount of time needed for the successful outcome, not needing to worry about failure. Perhaps they don't break their lockpick because they are now more careful and taking their time. Perhaps they search more thoroughly and there's less throwing objects across the room. Perhaps they figure out a way to break the object in such a way to avoid hurting themselves.

It's always been up to the DM for this stuff and the game has been better for it since the DM rarely has to worry about referencing a table to see the DC, the success clause, and the failure clause for every single type of challenge.

As before, I was not asking for another deflection. I’ll accept your proposal as calvinball and leave it at that.

So we find ourselves back at concrete rules for combat, concrete rules for spells, and the skill system lacking the substance required for it to serve as a reliable foundation for Martials to interact with the world.

Do some GMs only want Martials to have nice things when the GM says so, is that it?

[OldTrees1]

Calling it teeny doesn't change the fact that it isn't teeny. A lvl 20 character can take whatever punishment would occur for the missed 15% chance much better than a 5% chance character.

Calling it not teeny does not change the fact it is teeny.


Prove to me that going from 5% to 15% or 40% to 60% or 85% to 95% is not a tiny change over 20 levels. All of those are the +10% we see in the current system.
1) Both 5% and 15% are unlikely. I would not risk them unless the outcome was better than the opportunity cost. AND opportunity costs scale with level.
2) A d20 creates a range of 21 between auto fail and auto pass. Demonstrate 2/21 over 20 levels is not teeny tiny.

It is rather clearly a small change. You consider my perspective wrong, so prove it. Show me how 2 is a large fraction of 21.


Oh and then compare it to going from 43.75% to 87.75% (the smallest change under the adv L/3 L/3 system). Tell me why 44% is way too much and we should restrict ourselves to 10%.

[Pex]

For the sake of brevity I will invoke the discussion of the dc for climbing a tree to refute the notion that DM arbitrated “skills can do uncapped amazing things with DM chosen consequences” makes for a sound system that players can expect to consistently engage with. This skill driven system we are considering shouldn’t be a per table basis, that’s house rules as these are very much new rules being discussed.

Indeed. Many times when the DMs I play with have to think up a DC I remind them there doesn't always have to be a DC. A player can do a thing just because he wants to. I only say this when in my opinion an autosuccess appropriate of course. In any case they never listen, and the DC is 15 or 20 depending on how hard the DM thinks it is but always harder than I think it is by default since I think it should be automatic. If it was DC 10 or less the DM wouldn't be taking the time to think of it for me to bring it up even if I still think it should be auto. It gets annoying fast. I'd be perfectly fine if I was wrong in that the Thing deserved a roll of a particular difficulty and could not be autosuccess if there was some guidance to say so, but since everything is made up it's too dependent on DM whim on what you can do. We can hypothesize all we want here on how it should work, including absolutely loving it it's DM whim just discussing the differences of opinion on how to handle a particular instance, but it won't make any difference. As a DM you can run the game as you want to your satisfaction and be happy, just as I can when I DM, but I'm still going to be miserable as a player suffering DM whim of when I get to be Tarzan and when I'm George of the Jungle.

[Asisreo1]

As before, I was not asking for another deflection. I’ll accept your proposal as calvinball and leave it at that.

So we find ourselves back at concrete rules for combat, concrete rules for spells, and the skill system lacking the substance required for it to serve as a reliable foundation for Martials to interact with the world.

Do some GMs only want Martials to have nice things when the GM says so, is that it?

The rules for combat and spells are far from "concrete." There's rules where there absolutely needs to be, but where are the rules for exactly what somatic component you must do? Is it a subtle thing? Can you do them under your cloak? What about verbal? How loud must they be?

When you cast your spell, does another character recognize the spell? You can go by xanathar's rules, but that's a variant. What about the consequences of wishing for a non-listed effect? Up to the DM. The location to find a material component? Up to the DM since most components don't even show up on random loot tables. What about the acquisition of plane shift's tuning fork? Up to the DM. There's plenty other instances like "harmless sensory effects" and "Logical courses of actions" that are also dependent on the DM.

The only thing not determined by a DM in combat is what's on your character sheet and what's on an average creature's statblock. The terrain, lighting, weather, cover, and quantity of monsters are all up to the DM.

Even moreso, the combat system has the "improvised" option where it's literally saying that there is no clear defined rules for the majority of actions in combat, they're up to the DM. Surprise, one of the most game-changing aspects of how a combat could turn out, is decided by the DM. The game is rules-light everywhere, not just in the skills section.

[Asisreo1]

Calling it not teeny does not change the fact it is teeny.


Prove to me that going from 5% to 15% is not a tiny change over 20 levels.
1) Both chances are unlikely. I would not risk them unless the outcome was better than the opportunity cost. AND opportunity costs scale with level.
2) A d20 creates a range of 21 between auto fail and auto pass. Demonstrate 2/21 over 20 levels is not teeny tiny.

It is rather clearly a small change. You consider my perspective wrong, so prove it. Show me how 2 is a large fraction of 21.

I've proven it mathematically, you either fail to comprehend the math or refuse based on a pre-determined notion in your mind that tripling a percentage is non-significant.

If you had two banks that offered a loan and one had a 5% interest rate and the other had a 15% interest rate. You wouldn't owe 10% more from the second bank, you'd owe 200% more, triple the interest (assuming principal interest and not compound).

At this point, you'll probably just have to be ignorant. You don't want to see the math for how it is, only for how you want to interpret it. The difference between 1 & 2 is +1 but also doubled. The difference between 19 & 20 is +1 but only roughly 5%. But if you still won't see that, then this whole discussion is moot because you'll fail to see that the math actually supports the significance of small numbers becoming large.

I proved it with number of successful trials, I proved it with calculus, I proved it with rudimentary statistics but you just refuse. So there's just nothing I can do to have you see math as it is.

[OldTrees1]

I've proven it mathematically, you either fail to comprehend the math or refuse based on a pre-determined notion in your mind that tripling a percentage is non-significant.

You did not prove it. You showed that 3 times a tiny amount is a tiny amount.

PS: Be careful not to violate the forum rules.

If you had two banks that offered a loan and one had a 5% interest rate and the other had a 15% interest rate. You wouldn't owe 10% more from the second bank, you'd owe 200% more, triple the interest (assuming principal interest and not compound).

If an bank offered me a loan at 5%, then I would pay back 100% + interest. You would also pay back 100% + interest. Interest compounds and thus is a completely different math equation than skill checks. However assuming non compound interest. If we both paid at the same time it could be 115% vs 105%. Nowhere near 3x. And the fact 1.15/1.05 ~= 1.095 is irrelevant to the topic at hand because that is like comparing +5/+3 and saying it is 1.66x bigger. While it is true +5 is 5/3rds of +3, it is also irrelevant.

At this point, you'll probably just have to be ignorant. You don't want to see the math for how it is, only for how you want to interpret it. The difference between 1 & 2 is +1 but also doubled. The difference between 19 & 20 is +1 but only roughly 5%. But if you still won't see that, then this whole discussion is moot because you'll fail to see that the math actually supports the significance of small numbers becoming large.

I proved it with number of successful trials, I proved it with calculus, I proved it with rudimentary statistics but you just refuse. So there's just nothing I can do to have you see math as it is.

1) Remember the forum rules.
2) I understand the relative difference AND I understand the objective difference. However I also understand that progression is relative to the d20's influence. That is why it keeps coming back to 2/20 is 10% or why there are 21 positions in 5% increments from 0% to 100%.
You are refusing to address the questions of progression.

We can have it pinned as known that 15% is 3 times 5%. That does not prove that +10% is a significant fraction of 100%.

In other words, rather than address the issue, you have proved an unrelated factoid and mistaken it as being the issue.

[Tanarii]

For the sake of brevity I will invoke the discussion of the dc for climbing a tree to refute the notion that DM arbitrated “skills can do uncapped amazing things with DM chosen consequences” makes for a sound system that players can expect to consistently engage with. This skill driven system we are considering shouldn’t be a per table basis, that’s house rules as these are very much new rules being discussed.

It refutes the notion that DMs read and understood the system.

Complaints that DMs are setting the DCs too high, using DC 15 as an average, are justified. Because that means they read the system and are using it.

Complaints that DMs are setting DCs for things that are explicitly not covered and don't require checks are just complaints they didn't read the system.

The fact it keeps coming up as an example is more a tale of DM horror stories not knowing the rules. Not tales of system horror stories due to it encouraging DM arbitration. Or system horror stories of how the system math is broken.

[MoiMagnus]

I've proven it mathematically, you either fail to comprehend the math or refuse based on a pre-determined notion in your mind that tripling a percentage is non-significant.

If you had two banks that offered a loan and one had a 5% interest rate and the other had a 15% interest rate. You wouldn't owe 10% more from the second bank, you'd owe 200% more, triple the interest (assuming principal interest and not compound).

At this point, you'll probably just have to be ignorant. You don't want to see the math for how it is, only for how you want to interpret it. The difference between 1 & 2 is +1 but also doubled. The difference between 19 & 20 is +1 but only roughly 5%. But if you still won't see that, then this whole discussion is moot because you'll fail to see that the math actually supports the significance of small numbers becoming large.

I proved it with number of successful trials, I proved it with calculus, I proved it with rudimentary statistics but you just refuse. So there's just nothing I can do to have you see math as it is.

Relative % has its limit.
A change from 0% to 0.0001% might be a multiplication by infinity, it will most likely have less impact on the game that a change from 0.0001% to 100%.
A change from 5% to 15% is tripling the probability of success, but could be rebranded as a change from 95% of failure to 80% of failure, which is already much less significant (-14% relatively)

An additional factor much more relevant to the discussion is the implied context: I'd probably take a change from 60% -> 65% over a change from 5% -> 10%, because the first once actually affect an ability I will use an rely on (reasonably chance of success) while the second one affect an ability I will possibly never use (low chance of success). Since abilities don't exist in a vacuum, increasing a bad option, even if that's tripling its probability of success, is not worth any peanuts if the option is still too bad to be used at the end.

[Asisreo1]

If an bank offered me a loan at 5%, then I would pay back 100% + interest. You would also pay back 100% + interest. Interest compounds and thus is a completely different math equation than skill checks. However assuming non compound interest. If we both paid at the same time it could be 115% vs 105%. Nowhere near 3x. And the fact 1.15/1.05 ~= 1.095 is irrelevant to the topic at hand because that is like comparing +5/+3 and saying it is 1.66x bigger. While it is true +5 is 5/3rds of +3, it is also irrelevant.

You were going to pay off the principal loan regardless. It doesn't matter what that loan was. The interest could be 0% and you'd still have to pay it off. The part that matters, and the one everyone that pulls a loan cares about, is the interest rate.

1) Remember the forum rules.
2) I understand the relative difference AND I understand the objective difference. However I also understand that progression is relative to the d20's influence. That is why it keeps coming back to 2/20 is 10% or why there are 21 positions in 5% increments from 0% to 100%.
You are refusing to address the questions of progression.
We can have it pinned as known that 15% is 3 times 5%. That does not prove that +10% is a significant fraction of 100%.

It isn't about being a significant fraction of 100% because how it changes is what's important, right? We're comparing 2 states, one where you have a 5% chance of success, the other where you have a 15% chance of success. The d20 doesn't matter when comparing percent chances because it's uniform throughout. The percent chance changes are solely due to the mod bonuses.

In other words, rather than address the issue, you have proved an unrelated factoid and mistaken it as being the issue.

I don't know how politely I can tell you that you just aren't understanding the math of probability despite me showing you how significant a difference of 5% and 15% is. The most relevant fact is how frequently you can pull off the event successfully. And the 15% case does so with x3 more frequency. An absolute viewpoint doesn't matter if your issue is with the progression. You're arguing that it's a small change absolutely but you also argue that you want the change to be meaningful relatively.

I mean, on an absolute scale, giving you 10 million dollars doesn't amount to much. There's trillions of dollars constantly flowing everywhere. From your relative viewpoint, though, 10 million dollars would be significant and (probably) more than 100% than your current funds. Absolutely, it would probably be .001% of the total amount of money available, so you go from .00000001% to .00100001%. Absolutely, very minor change. But you only tell the difference from an relative scale.

[MoiMagnus]

I don't know how politely I can tell you that you just aren't understanding the math of probability despite me showing you how significant a difference of 5% and 15% is. The most relevant fact is how frequently you can pull off the event successfully. And the 15% case does so with x3 more frequency. An absolute viewpoint doesn't matter if your issue is with the progression. You're arguing that it's a small change absolutely but you also argue that you want the change to be meaningful relatively.

I can build a counterexample to that in the general case.

You are a fighter with two swords, and each turn you strike with both of them successively.
Sword 1: 5% chance of success for 20dmg.
Sword 2: 60% chance of success for 20dmg.

You have the choice between two buffs: the first is to increase Sword 1 from 5% to 15%. The second is to increase Sword 2 from 60% to 80%. Which one is better in average?

The mathematical answer is Sword 2, because it gives you 4 more damages, while increasing Sword 1 gives you only 2 more damages.
This is an example where the relative % approach is fully wrong, and the absolute % approach is correct.

EDIT: In general, relative % works better in situations where you can assume that "you have nothing other than it". Because doubling the % is effectively doubling your total value in average. At the contrary, absolute % works better in situations where "you already have other stuffs", because doubling a low % of gaining something will not double your value, and will in fact change almost nothing to it.

[Moreb Benhk]

In actual play, you care about the quantity of successes, not the success rate. It doesn't matter if you had a 40% success rate if you only succeed 2/10th if the time. You can't call it a 5% difference because quantity of successes are not a percentage system and you can't convert them to percent unless your being multiplicative. The contribution of the d20 just gives a baseline for the success rate for comparisons, it doesn't contribute much because without the bonuses, the difference of the 2 rolls would always be 0%. (I.E. the difference between success rates for 1d20 vs 1d20 on a DC 10 check = 0%. 0% increase). The d20 has been included in my calculations, though, because success rates differences and % increases need a solid reference for the success rates which needs a base system, which is the d20.

Why are you comparing the differences in two checks? Most checks are not opposed. We are talking about reaching a certain DC. Most of the contribution towards that is from the d20. The other contributions can be important (ability score, proficiency, etc), but struggle to even match let alone exceed this. The fact is a character with a +0 can still succeed at a check DC 10, 15 or even 20. Sure the nature of chance means that not every person's experience will match the probabilities, but that's the standpoint to evaluate how the system will work in the big picture.

I'm not sure why you're arguing with the math. It is, indeed, a higher quantity of success. Bounded accuracy makes +1's and +2's mean all the difference in the world. It seems you still haven't grasped the statistics of it all, which is normal since percentages often don't mean what people think they mean.

Because the math determines how things play out. A +1 or +2 simply can't match the influence of a d20. Bounded accuracy just means that the +1 or +2 is capable of contributing (most of the time - a DC 25/30 might still be off the range entirely), but it still represents 5% or 10% of the random variance. That is not 'all the difference in the world'. Against a DC 10 - someone with +0 succeeds 55% of the time, with a +1 they succeed 60% and a +2 65%. That is a difference, and a noticeable one. But what it looks like in practice is that the +2 will succeed roughly 1 time in 10 more than the +0.

We are comparing high tier martials, since low tier martials are agreed to not be bad at lower levels. Hard is indeed a fantastical DC. Dodging an expert wizard's fireball at full force is easier than a DC20 check. Bursting chains is fantastical. You are just as likely to resist the magical effects of a lich, a master of evil arcane energy and something that has studied magic for years. A DC 20 is completely fantastical, nothing here seems even remotely possible for me or you. The very basis of the skill system is fantastical.

But you've also said that this isn't level gated - so you are comparing low level martials too - because they can achieve these results (albeit less often). I think things get even less clear when you try to equivocate saves and ability checks (because they aren't the same thing and don't operate on the same scale). There is no ability check for dodging a fireball, so its hard to compare it to bursting chains. While I'd agree that bursting chains is beyond the strength of a regular human, its not on par with the Level 6-7 effects you claim to be enabling here.

You get to do these unlikely scenarios repeatedly, more than once, every day, though. A martial can lift castles, stop cannonballs midflight, punch through adamantine, take the deadliest poisons without skipping a beat, jump for miles, break the chains of carceri, sneak past the best sentinels in the multiverse, etc. Constant attempts, most likely fails but they can keep trying. Their high-level, damage isn't too much of a threat and disadvantage on something is whatever. They can keep attempting these things, though.

And it seems you forget some spells have saving throws and checks associated with them that could have them have 0 effect despite spending the spell slot.

They MIGHT be able to do many of these things, if the GM is permissive of it. But I'd also bring you back to your own assertion that failed attempts should have negative consequences (and the ones mentioned weren't simply 'you can't attempt that particular feat again right now') - because in a system set up for a high failure rate, the consequences of failure are hugely determinative of how the system functions.

I don't forget spells and saves/checks at all. There isn't much point in trying work out the balance unless the fundamental skill system is functioning first.

EDITED TO ADD

I've proven it mathematically, you either fail to comprehend the math or refuse based on a pre-determined notion in your mind that tripling a percentage is non-significant.
...

I proved it with number of successful trials, I proved it with calculus, I proved it with rudimentary statistics but you just refuse. So there's just nothing I can do to have you see math as it is.

It seems pretty clear that you have a certain belief about how this works and that our attempts to interact with that have been rebuffed. I'm not certain that we are making any headway in discussing the issue (of martial capacity) when things are bogged down in how fundamental statistics works. Can we agree to disagree?

For those of us who consider the current system problematic, the key issues I can see with massaging in more space for the fantastic in skills are:
1. casters get equal access to skills (admittedly, like magic items, same objective benefit means relatively more benefit to martials)
2. the easy option is to emulate spell-abilities, but this is also a boon to casters (admittedly, like magic items, same objective benefit means relatively more benefit to martials)
3. the fundimental system numbers don't shift enough to allow pure DCs to be a good metric (if it caps out at 30 and potentially 30 is reachable at level 1 (guidance+20 stat + expertise)... that's an issue). There isn't progress in capacity over the levels, only some improvement in reliability - which still remains low enough for the higher DCs that you'd be discouraged from attempting them in a pressure situation.
4. issues of repeat checks - spells discourage repeat checks by making each check deplete a resource (spell slots). if failure is not handled well skills easily devolve into re-attempting over and over and over until success (time being the only factor). - which makes in non-time critical situations (which will occur) any result that is potentially achievable without depleting resources (eg - bardic inspiration) is guarenteed achievable. This means anything 'unlocked' by those DC checks is doable repeatedly and as desired, starting from level 1 (for anything up to 25 at a minimum). This either means abilities that are 'higher level appropriate' need to all cluster in DCs going over 25 or not be part of the system. If the former, that means that many martials will not be able to use these abilities (especially outside of the strength/athletics zone).
5. Reliable talent becomes ridiculous - you now can attempt truly fantastical things without ever having a chance to fail - in areas of expertise especially.

Am I missing any other issues/overstated them?

It refutes the notion that DMs read and understood the system.

Complaints that DMs are setting the DCs too high, using DC 15 as an average, are justified. Because that means they read the system and are using it.

Complaints that DMs are setting DCs for things that are explicitly not covered and don't require checks are just complaints they didn't read the system.

The fact it keeps coming up as an example is more a tale of DM horror stories not knowing the rules. Not tales of system horror stories due to it encouraging DM arbitration. Or system horror stories of how the system math is broken.

Except that there is very little guidance about what is 'explicitly not covered and not requiring checks' (I can think of a few - some lifting and jumping numbers) most of what is not covered is implicitly left to the DM to arbitrate as they see fit, which is reading and using the system (such as it is, if you can call 'make it up as you like' and 'here are some DC numbers [that don't work well in practice]' a system), and also interacts with the maths problem of setting DC 15 as average.

[OldTrees1]

You were going to pay off the principal loan regardless. It doesn't matter what that loan was. The interest could be 0% and you'd still have to pay it off. The part that matters, and the one everyone that pulls a loan cares about, is the interest rate.

I don't suggest assuming the loan is free money. You pay back the principal + the interest. For the majority of loans (credit cards paid off the same month) the principal is the vast majority of the amount paid. For the majority of the money paid to loans (interest on long term compound interest loans) the compound interest is the biggest factor (and that is exponential). But neither example is analogous to the topic at hand.

It isn't about being a significant fraction of 100% because how it changes is what's important, right? We're comparing 2 states, one where you have a 5% chance of success, the other where you have a 15% chance of success. The d20 doesn't matter when comparing percent chances because it's uniform throughout. The percent chance changes are solely due to the mod bonuses.

Did you notice, this is where our perspectives differ drastically? We both see ourselves as caring about what change is important. The problem arises when you attempt to say the relation 15%/5% matters to me when you know I am talking about progression 2/20. Especially since I am concerned about more than just your 5%->15%, I also recognize the 60%->70% and the 85%->95%.

The d20 does matter when comparing progression. If it was 1d6+3 vs 1d6+5 there would be much more progression.

I don't know how politely I can tell you that you just aren't understanding the math of probability despite me showing you how significant a difference of 5% and 15% is. The most relevant fact is how frequently you can pull off the event successfully. And the 15% case does so with x3 more frequency. An absolute viewpoint doesn't matter if your issue is with the progression. You're arguing that it's a small change absolutely but you also argue that you want the change to be meaningful relatively.

I mean, on an absolute scale, giving you 10 million dollars doesn't amount to much. There's trillions of dollars constantly flowing everywhere. From your relative viewpoint, though, 10 million dollars would be significant and (probably) more than 100% than your current funds. Absolutely, it would probably be .001% of the total amount of money available, so you go from .00000001% to .00100001%. Absolutely, very minor change. But you only tell the difference from an relative scale.

Right or wrong, my understanding of the forum rules is to not say that. So I haven't.

Can you agree that you will probably fail a 5% chance?
Can you agree that you will probably fail a 15% chance?
Sounds like it is not worth the attempt.
Can you agree that you will have relatively even odds on a 45% chance?
Can you agree that you will have relatively even odds on a 55% chance?
Sounds like even odds.
Can you agree that you will probably pass an 85% chance? Edit: Fixed typo
Can you agree that you will probably pass a 95% chance? Edit: Fixed typo
Sounds like reliable regardless.
Can you notice how 1st and 20th could feel the same in that context?
The progression is so slow and so little that it feels like you go nowhere at all?

[Asisreo1]

I can build a counterexample to that in the general case.

You are a fighter with two swords, and each turn you strike with both of them successively.
Sword 1: 5% chance of success for 20dmg.
Sword 2: 60% chance of success for 20dmg.

You have the choice between two buffs: the first is to increase Sword 1 from 5% to 15%. The second is to increase Sword 2 from 60% to 80%. Which one is better in average?

The mathematical answer is Sword 2, because it gives you 4 more damages, while increasing Sword 1 gives you only 2 more damages.
This is an example where the relative % approach is fully wrong, and the absolute % approach is correct.

Right, but that's because you're comparing 2 percent changes over an absolute scale while we're being asked to determine how effective the percent change is to it's initial state.

The problem you present gives the person a choice on which upgrade they want with objective determinations: do you want to go from 5 to 15 or from 60 to 80? Of course, the absolute scale is worth using because you'll be more likely succeed more on the second point.

I'm not arguing that. Sure, if you'd rather put your ASI's into something to make it go from 60% success to 80% success, you should do so rather than 5% to 15%. But it's the fact that the act of going from 5% to 15% is a x3 change is what I'm arguing. There isn't an alternative to go from 60%-80%. If there was, an absolute argument could be made. But it's all relative.

[Asisreo1]

I don't suggest assuming the loan is free money. You pay back the principal + the interest. For the majority of loans (credit cards paid off the same month) the principal is the vast majority of the amount paid. For the majority of the money paid to loans (interest on long term compound interest loans) the compound interest is the biggest factor (and that is exponential). But neither example is analogous to the topic at hand.

Unless you start slacking on your payments and can only afford to pay off the interest. The 5% case gives you 20 months before you start paying more in interest. The 15% case gives you only half a year before you paid more for interest than principal.

Did you notice, this is where our perspectives differ drastically? We both see ourselves as caring about what change is important. The problem arises when you attempt to say the relation 15%/5% matters to me when you know I am talking about progression 2/20. Especially since I am concerned about more than just your 5%->15%, I also recognize the 60%->70% and the 85%->95%.

The d20 does matter when comparing progression. If it was 1d6+3 vs 1d6+5 there would be much more progression.

You have to look from a statistical angle. Your d6 example (assuming a DC of 9) goes from a 16.6% chance of success to a 83% chance of success. That's 5 times the success rate, but only because the actual percentages were not equivalent to the d20 example. Use percentile dice and have the bonus be from 1d100+20 to 1d100+35 with a DC of 120. It's the same success rates and therefore the same percent change. It also has that big shiny +15 across 20 levels. The difference is only that it's bloated.

Right or wrong, my understanding of the forum rules is to not say that. So I haven't.

Can you agree that you will probably fail a 5% chance?
Can you agree that you will probably fail a 15% chance?
Sounds like it is not worth the attempt.
Can you agree that you will have relatively even odds on a 45% chance?
Can you agree that you will have relatively even odds on a 55% chance?
Sounds like even odds.
Can you agree that you will probably fail an 85% chance?
Can you agree that you will probably fail a 95% chance?
Sounds like reliable regardless.
Can you notice how 1st and 20th could feel the same in that context?
The progression is so slow and so little that it feels like you go nowhere at all?

Yes.
Yes.
It might be, it might not be. Sometimes, it might feel like you might as well try.
No.
No.
Because they aren't even odds.
No.
No.
Yeah, though it isn't all that reliable.
Only in a certain mindset. Not objectively.
No. But if you feel that way, you're free to convert the d20 system to a d100 system and it will suddenly feel like you're increasing your modifiers by massive amounts. Increase of 5 every 4 levels!

[OldTrees1]

Unless you start slacking on your payments and can only afford to pay off the interest. The 5% case gives you 20 months before you start paying more in interest. The 15% case gives you only half a year before you paid more for interest than principal.

Serious talk, pay off loans sooner rather than later. That principal is not free money.

You have to look from a statistical angle. Your d6 example (assuming a DC of 9) goes from a 16.6% chance of success to a 83% chance of success. That's 5 times the success rate, but only because the actual percentages were not equivalent to the d20 example. Use percentile dice and have the bonus be from 1d100+20 to 1d100+35 with a DC of 120. It's the same success rates and therefore the same percent change. It also has that big shiny +15 across 20 levels. The difference is only that it's bloated.

I see. So when I said decreasing the die size increases the progression of the +2 across the die, you did the following:
1) You started by calculating the math of the d6 case. You noticed 16.6% to 83% was a 5x increase (the part you cared about). You also had to have noticed it went from a rather unlikely attempt to a rather likely attempt (the part I cared about).
2) Then you decided clearly I did not mean what I meant, so rather than adjust the ratio of bonus change to die size, why not keep it constant.
3) Then you doubled down on the misrepresentation by suggesting increasing the die side.
4) Then you topped off the misrepresentation by pretending I would want +15/100 across 20 levels with DCs increments similarly scaled up because "+15 is more progression than +2".While technically 3/20 is more than 2/20, your reference to "bloated" clearly shows you intended to highlight the large numbers rather than the progression.

No. But if you feel that way, you're free to convert the d20 system to a d100 system and it will suddenly feel like you're increasing your modifiers by massive amounts. Increase of 5 every 4 levels!

I see you made a math error here just like I had a typo, and/or you are intentionally misrepresenting my position.


A change of +2 on a d20
A change of +3 on a d20
A change of +15 on a d100
A change of +20 on a d100
These are not large enough progressions for 20 levels. The closest one is a mere 20% which is less than the 30% I dislike about 5E skills.
I don't care about the size of the bonus. I care about the change in the bonus per level relative to the die size. Aka progression. I care about the statistic that tells me will a PC stay the same, or will they grow into, through, and master new thresholds.


PS: You have argued 15%/5%=3. I agreed. I have argued 45% is relatively even odds. 45%/(1-45%)=9/11~=1. You disagreed, claiming 11=/=9, which while true, did not refute 0.81818181 ~=1 aka relatively even odds. You can't get closer on a d20 without it being literally even odds. So 45% is relatively even odds of pass/fail. Was the wording your objection? Would you have accepted "close to even odds"?

[Pex]

It refutes the notion that DMs read and understood the system.

Complaints that DMs are setting the DCs too high, using DC 15 as an average, are justified. Because that means they read the system and are using it.

Complaints that DMs are setting DCs for things that are explicitly not covered and don't require checks are just complaints they didn't read the system.

The fact it keeps coming up as an example is more a tale of DM horror stories not knowing the rules. Not tales of system horror stories due to it encouraging DM arbitration. Or system horror stories of how the system math is broken.

They've read the rules. They know the rules. They disagree on the difficulty of tasks. That's what causes the problems.

[Man_Over_Game]

Jesus, this thread got weird, and it didn't start all that sane to begin with.

[Asisreo1]

I see. So when I said decreasing the die size increases the progression of the +2 across the die, you did the following:
1) You started by calculating the math of the d6 case. You noticed 16.6% to 83% was a 5x increase. You also had to have noticed it went from an unlikely attempt to a likely attempt.
2) Then you decided clearly I did not mean what I meant, so rather than adjust the ratio of bonus change to die size, why not keep it constant.
3) Then you doubled down on the misrepresentation by suggesting increasing the die side.
4) Then you topped off the misrepresentation by pretending I would want +15/100 across 20 levels with DCs increments similarly scaled up because "+15 is more progression than +2".While technically 3/20 is more than 2/20, your reference to "bloated" clearly shows you intended to highlight the large numbers rather than the progression.

I saw what you meant, it was just a false equivalence. The +2 bonus and dice change the percent change. But that wasn't what you said bothered you. You said that going from a 5% to 15% wasn't a significant enough boost. I'm saying that independent from the dice system, regardless of what system you're using, going from 5% to 15% is very significant. Something with a mortality rate of 15% is much more lethal and concerning than something with a 5% mortality rate.

The dice and whatnot mean nothing so long as the percent change is consistent. But you're saying that changing the dice makes the bonus better, but that's only because the bonus is meant to scale the success chance. There is no equivalent in a d6 system but if there was and it could be in a bounded accuracy system like the current system, I'd use their equivalent. Just as I can have an equivalent 1d100 system.

I see you made a math error here just like I had a typo, and/or you are intentionally misrepresenting my position.


A change of +2 on a d20
A change of +3 on a d20
A change of +15 on a d100
A change of +20 on a d100
These are not large enough progressions for 20 levels. The closest one is a mere 20% which is less than the 30% I dislike about 5E skills.
I don't care about the size of the bonus. I care about the change in the bonus per level relative to the die size. Aka progression. I care about the statistic that tells me will a PC stay the same, or will they grow into, through, and master new thresholds.

If you like, you can give a +1 to the d100 system each and every level with +2 on the ones where they'd have ASI's.

If you want a larger magnitude but don't know how these percent changes in the current system is already large, you'll run into bloat problems quickly. If you understand it, but would rather have it by any other name, fine. But the underlying math most likely doesn't change.

Whether you roll a d20 +5 against an AC 15 character or you roll a d20 +25 against an AC 35 character, the percent change, the actual underlying math, remains the same.