On smaller bonuses and the d20 mechanic

[Majoru Oakheart]

Small modifiers mean small target numbers (otherwise chances of success are too small), and, when using the d20, create a feeling that the character's effect on the outcome isn't adequately reflected in the game mechanic. Would you be happy with those implications?

I don't know that they create a feeling that a character's effect on the outcome isn't adequately reflected.

If I have a +2 to hit and I need a 10 to hit, I know that I have a 10% greater chance to hit than the Wizard beside me who only has +0. I'm better at fighting, and that is reflected in the rules. You don't get the same feeling as having +35 vs his +0 and being overwhelmingly more powerful, however.

Besides, your character can have effect in other ways. Sure, you only have a 10% greater chance to hit the enemy, but what if, after you hit you are doing double or triple the amount of damage that the Wizard does. Your overall effect on the combat is greater, even if your attack bonus isn't.

 

[Cadfan]

Technically there are a few other options. For example, if success at a task requires multiple die rolls, then even if a hero is only slightly better than a commoner on each individual die roll, that hero might still be significantly better than a commoner in terms of overall chance of success.

 

[Flatus Maximus]

I don't know that they create a feeling that a character's effect on the outcome isn't adequately reflected.

More generally, the mechanic can be expressed as (fate) + (character ability) = outcome. If the numerical representation of the character's ability is small relative to the range of values that come from the fate, then it can be said that the character's ability has relatively little to do with the outcome.

In this sense, 1d20+0 and 1d20+2 aren't very different -- the modifiers are relatively small compared to the range of possible values fate brings to the calculation. (This, by the way, is completely independent of target numbers, i.e., probability of success.) While it's true that one has a 10% better chance of success, that might be small consolation to folks that don't feel like their character is adequately represented in the determination of the outcome. (And indeed it is not.)

Going to the other extreme, the outcome of 1d20+35 is determined mostly by the modifier (character ability). Sure, the 1d20 is still there, but it's now relatively small. (Again, this is totally independent of probability of success.)

Somewhere in between is the middle ground. For example, I bet if one looks at typically target numbers and modifiers for the "sweet spot" of 3.x edition, we'll see that the modifiers have a good "heft" to them, while the target numbers are neither too small (so that commoners can do the job) nor too large (so that PCs, with their moderate modifiers, can do the job).

 

[Oldtimer]

More generally, the mechanic can be expressed as (fate) + (character ability) = outcome. If the numerical representation of the character's ability is small relative to the range of values that come from the fate, then it can be said that the character's ability has relatively little to do with the outcome.

I hear this a lot, but it's simply incorrect. Since it's a linear function, the magnitude of the total modifier and target number does not change the influence of character ability.

In this sense, 1d20+0 and 1d20+2 aren't very different -- the modifiers are relatively small compared to the range of possible values fate brings to the calculation. (This, by the way, is completely independent of target numbers, i.e., probability of success.) While it's true that one has a 10% better chance of success, that might be small consolation to folks that don't feel like their character is adequately represented in the determination of the outcome. (And indeed it is not.)

That's a different thing. Now you're comparing two modifiers and saying that a difference of +2 is small compared to the difference of +19 (of the die rolls). 10% better chance of success is still 10% better.

Going to the other extreme, the outcome of 1d20+35 is determined mostly by the modifier (character ability). Sure, the 1d20 is still there, but it's now relatively small. (Again, this is totally independent of probability of success.)

There is nothing relative about this. 1d20+35 vs 50 is exactly the same as 1d20+2 vs 17. The outcome is determined by the die equally much in both cases.

Somewhere in between is the middle ground. For example, I bet if one looks at typically target numbers and modifiers for the "sweet spot" of 3.x edition, we'll see that the modifiers have a good "heft" to them, while the target numbers are neither too small (so that commoners can do the job) nor too large (so that PCs, with their moderate modifiers, can do the job).

It seems that you are actually saying something else - that the difference in total modifier of a PC and a commoner should not be too small or too large compared to the spread of 1d20. This is the old scaling debate.

But you simply cannot say that the die roll is shadowed by the modifier in itself. We can build a system where modifiers start at +1000 and target numbers at 1010 and the influence of the die roll is still the same.

 

[Flatus Maximus]

I hear this a lot, but it's simply incorrect. Since it's a linear function, the magnitude of the total modifier and target number does not change the influence of character ability.

There are two considerations when looking at the d20 mechanic: the size of the modifier compared to the range of values the die can take, and the difference between the target number and the modifier. The first is both a mathematical and psychological consideration (how much a player feels as if there character is contributing to the outcome), and the other is a consideration of the odds of success. In the first part that you quote, I'm talking about the first consideration, while it appears you think I'm talking about the second.

That's a different thing. Now you're comparing two modifiers and saying that a difference of +2 is small compared to the difference of +19 (of the die rolls). 10% better chance of success is still 10% better.

There is nothing relative about this. 1d20+35 vs 50 is exactly the same as 1d20+2 vs 17. The outcome is determined by the die equally much in both cases.

It absolutely is relative, but let me be clear what "it" I'm talking about. While you are correct that the odds are exactly the same (second consideration), the result of 1d20+35 is determined more by the character (modifier) than fate (1d20), while the result of 1d20+2 is determined mostly by fate. That they have exactly the same chance of producing success (for target numbers 50 and 17, respectively) is irrelevant to the first consideration (2 is small compared to 20, while 35 is relatively large).

It seems that you are actually saying something else - that the difference in total modifier of a PC and a commoner should not be too small or too large compared to the spread of 1d20. This is the old scaling debate.

But you simply cannot say that the die roll is shadowed by the modifier in itself. We can build a system where modifiers start at +1000 and target numbers at 1010 and the influence of the die roll is still the same.

At this point, I'm bringing the two considerations together to suggest why the sweet spot is what it is. Regarding the first consideration, modifiers aren't too small (to distinguish PCs from commoners) but not too big (so that fate is still a significant part of the roll). Regarding the second consideration, the target numbers haven't grown so much that even with relatively small modifiers, most PCs still have some chance of success.

 

[Yora]

I think that compared to the d20, modifiers of +5 or +7 may not seem like much is true, but in reality, it really doesn't matter. As the modifiers rise, so do the target numbers.

1d20+28 against a DC of 41 seems really cool.
But the chances are just the same as 1d20+8 against a DC of 21. Dealing with large numbers only for the illusion of being totaly awesome and doing totaly incredible things is not worth it in my book.
It also has the downside that some things become meaningless very fast, and others are unreachable for a very long time. As you advance through the levels, monsters and equipment gets thrown away with new ones comming to replace them at a very fast pace. Which is good for some groups, but really annoying for more "heroic" style games in which monsters and items are supposed to be relevant pretty much all the time for the whole campaign.

So I am glad 5th Edition is planned to keep the power curve much flatter than before.

 

[JamesonCourage]

I think that compared to the d20, modifiers of +5 or +7 may not seem like much is true, but in reality, it really doesn't matter. As the modifiers rise, so do the target numbers.

Everything changes as soon as things become opposed, or when looking at the internal consistency of the world via game mechanics.

That is, look at a Str 10+0 Wizard and a Strength 16+3 Fighter. The Fighter is obviously stronger. They both decide to break down a door. Strength checks are called for. If the DC is 12, the Fighter will succeed 60% of the time, while the Wizard will succeed 45% of the time. If the Fighter rolls a little low, and the Wizard a little high, then the Wizard will kick in the door before the obviously stronger Fighter.

I think the current proposed solution with 5e is "in this case, make the Fighter succeed." This is pretty unappealing to me, personally, as it kind of just tries to sweep the problem under the table. Let's look at another example.

The Ranger gets +3 to Hiding from his Dexterity. The Rogue gets +3 to Hiding from his Dexterity, and +2 from his class (his total is +5). They both decide to hide, and enemies attempt to spot them. The Wizard and Fighter (+1 bonus to Hiding) also attempt to hide. When it comes to what the enemies need to roll on their Notice Hiding Enemies check, it's kind of a crap shoot on who will roll low.

A larger spread between modifiers means that the Rogue can hide much more reliably than the Wizard and Fighter, and somewhat more reliably than the Ranger. Instead, what we have is a Rogue who feels a little gimped because the Wizard and Fighter are only 20% worse off than he is, and luck plays such a large part of his success.

If the game does take up a system of opposed ability checks (I disarm him, a Strength check, and he opposes with Dexterity), then the Wizard can defend himself nearly as well as the Fighter can (from an in-game perspective), because the bonuses are so similar. The personal skill of the Wizard and Fighter are close martially, just as the ability to hide between a Rogue and Wizard are close.

Statistically, the odds of success can remain controlled and consistent, but this is due to luck, not skill. For skill to be meaningfully separated from an in-game perspective, the Fighter must be consistently significantly better than the Wizard at fighting, and the Rogue must be consistently significantly better than the Wizard at hiding.

This is not accomplished with such small modifiers. There's not enough spread. +1 to +5 is not enough to establish this on a significant level. Noticeable, yes. Appreciated to some degree, definitely. However, there's just too much of a chance that the Wizard will wildly succeed (natural 19) while my rogue bombs (natural 4).

If they wish to keep numbers small, maybe they'll implement another mechanic, like Good At Sneaking: Roll 2d20 and take the larger of the two rolls or Good At Bashing Doors: Roll 2d20 and take the larger of the two rolls.

I'm not talking about statistics of succeeding on rolls, I'm talking about comparing capabilities of different PCs (and, importantly, NPCs) from an in-game perspective. There's a few "fixes" (for those that find this problematic), including multiple rolls (taking the higher), rolling a lower die (I prefer a d6, and use it in my game for attribute checks), or raising the spread (+10 to +1 is more reliable).

I'm not sure if they'll change anything, but right now the prospect of 1d20+(0 to 6) is by far the thing that makes me feel the most uneasy about the announcements.

Of course, it's all speculation, and it might be different from what I think it is. I hope it is. But as it stands, there isn't anything that massively separates skilled warriors from skilled rogues from an in-game perspective, and that certainly affects my immersion, and my interest in the game.

Again, since it keeps getting brought up, this isn't about chances of succeeding consistently (you always have 55% chance of success). It's about how well one person can succeed in relation to all other PCs and NPCs, and whether or not his skill is properly reflected within the fiction. As always, play what you like

 

[Zireael]

I'm not talking about statistics of succeeding on rolls, I'm talking about comparing capabilities of different PCs (and, importantly, NPCs) from an in-game perspective. There's a few "fixes" (for those that find this problematic), including multiple rolls (taking the higher), rolling a lower die (I prefer a d6, and use it in my game for attribute checks), or raising the spread (+10 to +1 is more reliable).

Which one of those would be the best?

 

[Mokona]

More generally, the mechanic can be expressed as (fate) + (character ability) = outcome. If the numerical representation of the character's ability is small relative to the range of values that come from the fate, then it can be said that the character's ability has relatively little to do with the outcome.

In this sense, 1d20+0 and 1d20+2 aren't very different -- the modifiers are relatively small compared to the range of possible values fate brings to the calculation. (This, by the way, is completely independent of target numbers, i.e., probability of success.) While it's true that one has a 10% better chance of success,...

A gap of +2 does not equal a 10% better chance to hit in most situations. If the d20 needs to be 11 to succeed for +0 and 9 to succeed for +2 that is a 20% "better" chance to succeed. If the d20 needs to be 17 to hit for +0 and 15 to hit for +2 that is a 50% better chance to hit. Each '+1' increment is a +5 "percentage point" improvement but that is a relatively meaningless number by itself.

When we're talking about game design, and not some fixed universe, the Target Number (TN) for the outcome is not yet determined. Once designers decide that Character Ability (aka Modifier or M) is relatively large compared to the range of Fate value (1d20), then they just raise TN to negate the impact of M.

We've been talking about the impact of a large M (in the d20 >= TN-M equation) on the character who is expected to tackle the task and for whom the TN was set in order to create a "fun / exciting" probability of success.

The true impact of large M or a steep growth curve in character ability is for characters who are not normal. Some characters try to specialize for certain TN situations (skill focus is for these people) and other characters are less proficient than the average skilled person. These groups are heavily impacted by the scaling issue of how big M is relative to 1d20.

For example (4e), an archer ranger with a Dex 20 and bow expertise has a +3 better chance To Hit than a melee ranger with Str 16. Against an AC that requires a 12 on the die roll for the specialist, the weaker to hit roll character will be 33% less likely to succeed. That is, for every 3 times one character hits the other will only hit 2 times. Since those rangers are making two attack rolls per round by the time five rounds are over we'll have 10 rolls * 45% = 4.5 hits versus 10 rolls * 30% = 3 hits. Over the course of three fights thats 13.5 hits versus 9 hits. One of those players is going to feel a noticeably worse at hitting & having fun than the other. In this case the AC's I used were too high, really, but I wanted to be able to run the math in my head. Also, one of the players chose to have less fun in combat because he wanted to play a certain class + a certain race and didn't back down when the rules penalized him for his choices. Note: 4e melee rangers are just more MAD (multiple attribute dependent) than 4e archer rangers because the archer build can afford to have horrible attributes in everthing but Dex while melee needs Str + Dex + Con to survive.

The DM will have a hard time balancing encounters for characters where there are large gaps between the "best" to hit rolls in the party and the "worst" to hit rolls in the party. The big number syndrome where attack rolls grow quickly / on a steep curve makes this problem worse.

A more common problem is skill checks. If it takes a +8 to be stealthy then no one in the party (+0 to +1 typically) other than the rogue can ever be stealthy. These kind of large gaps close off lots of interesting innovation by parties outside of combat and lead to a "break down the front door" scenario over and over again. This is, in fact, why 4e gives people modifiers even on skills they "don't know". In order to fix the broken problem of large M the designers gave people default bonuses to skills across the board. This default capability ruined verisimilitude for some players and added a lot of needless bookkeeping to the game. The alternate solution would have been to slow down the progression of numbers for skilled characters (this solution appears to be the choice as of now for D&D Next).

 

[Hassassin]

I think that compared to the d20, modifiers of +5 or +7 may not seem like much is true, but in reality, it really doesn't matter. As the modifiers rise, so do the target numbers.

Only in opposed rolls. The DC of a certain task remains the same, whatever the number. The characters will, however, be able to succeed at different kind of tasks.

The absolute scale of bonuses does matter. If the lowest you can have to a skill is -4 (ability) while the highest is +9 (ability and 5 from skill training), it means DC 16 is the highest that everyone has a chance of succeeding at, while DC 11 is the lowest everyone has a chance of failing at.

The higher the range of bonuses, the more tasks become automatic successes to trained characters. Meanwhile the number of tasks where poorly suitable characters have no chance of succeeding increases.