What are the practical limits of d20+mod vs DC?

[Desdichado]

I'm a fan of the notion that somewhere around +10 or so is the practical limit. One of the reasons I like E6 is that even with crazy specialization, you're not likely to get much beyond that number. You can't get a BAB bonus higher than +6--factor in some STR bonus and maybe a masterwork or magic bonus on your weapon, and you're in the neighborhood of +10. Your good base save is going to be +5. Again, with some ability mods and a cloak of resistance or something, you'll be in the range of +10. For skills, you can get a little higher, since you can get have a rank bonus of +10 already at 6th level. Add to that your ability modifier, any bonuses from any magic, the Skill focus feat, any racial bonuses, etc. you can easily get up to +15 or even a bit higher. But, in general, I think that for skil bonuses the practical limit is a bit softer, and skill checks in general tend to have a slightly less dramatic impact on play than attack rolls and saving throws anyway. Which I think makes a difference in what the practical limit should be, too.

Come to think of it, I think external factors like that make a big difference on practical limits. Another problem with the practical limit on attack rolls is AC. D&D doesn't handle this as well as other games in the d20 family, because it doesn't have a scaling AC or Defense score, like d20 Modern, d20 Star Wars, d20 Wheel of Time, etc. all do. Therefore, attack rolls start to become kind of silly without patches to make the AC scale upwards; either natural armor for monsters, or actual armor for characters. If all the ACs you're going to be attacking were unarmored humanoids, for example, then you'd rarely face an AC higher than 14 or 15 at the top limit, even for higher level characters. In that case, there's really no reason to have an attack bonus of even +10; you're almost in autohit mode when your bonus is that high.

 

[Wiseblood]

I am not sure what the practical limit of a d20+mod is but I am certain 3e and 4e break the limit.

IMO skill DC's should be 20. A dm can modify based on circumstance (not skill of the character that has already been represented) if so desired. This modifier should not exceed +/- 2.

If you are looking at basic, consider this Rogue skills used to be handled with d100. Climb Walls had a base chance of 60% modified by armor, race, and ability score. D20 translation with DC 20 as target your modifier is +11.

Heres the kicker, The DM always knows the DC and it will always be fair. Skilled characters are more likely to succeed than the unskilled. Futher, as long as the total modifier is capped at +18 there usually remains a Choke Factor for the masters and a Lucky Shot for the beginners.

 

[Crazy Jerome]

Bell curve systems have an unfortunate side effect of making it really difficult to estimate probabilities in play and making the value of bonuses unpredictable. The only bell curve system I've seen that avoids this issue is Fudge, since the average value of the dice is always zero-- you have a 1-in-2 chance of succeeding (or failing) when your bonus equals the difficulty, with the probability of success (or failure) scaling exponentially per point of difference.

Then, if you're using smaller numbers, you run into far fewer cases where difficulties fall outside of the range of possible results.

I've experimented with 2d10 as a good compromise. As with any bell curve, you can't completely escape the "mutating modifier" effect. But at least the probabilities can be farily easily explained, as they reduce down to even percentages. And the curve is shallow enough that you get a decent distribution of common results.

2, 20 - 1% each
3, 19 - 2% each
....
10, 12 - 9% each
11 - 10%

If you keep a d20+mod system within the parameters that GSHamster outlined (most results running 5 to 15, with mods ranging +1 to +10), and you are careful about what the baseline is (either no scaling or very careful scaling), then you can even substitute the d20 for 2d10 with very little other changes. For people that like the auto fail and auto success at the extreme ends to be more rare, this is a handy way to do it.

It doesn't bother anyone else at our table, because they don't care about that aspect. But any system that sets up the mods, such that I can't reasonably easily switch out the dice, tends to irk me a bit. It's a minor irritant, under the skin.

 

[innerdude]

I think Wizards partially answered this question for us:

30 character levels in 4e = 15 natural bonus to all numbers, across all classes.


If you start with that as a baseline, then everything else only becomes character specific--feats, weapon enhancements, skill focus, and so on.

In fact, if 3.x had taken the same tactic as 4e--lopping off levels 16-20, but then thinning the range of power for the remaining 15 levels between 25 or 30--I wonder if the cries of the wild "imbalance" of high-level 3e would have been as bad.

This would have necessitated also limiting some of the more "world breaking" 8th and 9th level spells as well.

Actually, I think I would have really liked a D&D 3.x that did just that......

 

[Viktyr Gehrig]

It isn't numbers that break high-level D&D, it's spells.

 

[innerdude]

It isn't numbers that break high-level D&D, it's spells.

You're right, to a point, though Hobo's comment about scaling AC is a big factor as well.

When AC doesn't scale, meaning your character has very little real way to improve their ability to not get hit, the thing that HAS to scale is hit points.....and then so too does the ability to "heal" hit points, through magic, surges, whatever. And as the "I don't get the dislike of healing surges" thread proves, this is a pretty big bone of contention for a lot of gamers.

When the numbers start to get ridiculous in D20, particularly the attack and defense numbers, the whole system begins to feel less like "heroic fantasy," and more like a Player vs. GM Arms Race to see who can "stay on top" of it all.

 

[Rogue Agent]

So easy DC of 10, when you got +10 is an auto-success (ignoring 1 as auto-fail).

the problem gets worse as the + goes up. Because the + dominates the value of the total, as it contributes to the die roll.

To sum up, when the modifer is worth more that the die roll, the glitch in the mechanic is revealed.

That's a feature, not a bug: The system creates a 19 point range of DCs for which success is doubtful-but-possible. If the DC is below that range, success is automatic. If the DC is above that range, success is impossible.

If your total modifier is +0, that range goes from DC 2 to DC 20. If your total modifier is +10, that range is DC 12 to DC 30. If your modifier is +50, that range is 52 to 70.

If you feel that it's a "glitch" that 1st level characters don't generally find DC -5 checks to be interesting (because they'll auto-succeed on them), then you have a point. If you don't think auto-succeeding on DC -5 checks is a problem, then you need to apply the same logic to the mechanic as a whole.

With that being said, the system becomes problematic when the "interesting range" for one character no longer overlaps with the interesting range for another character. However, this is only really problematic in the narrow range of cases where all PCs are expected to be able to hit the same DC. IME, this is generally limited to combat, saving throws, and stealth.

Once Character A has a modifier 10-15 points higher than Character B, it becomes difficult to set a DC which is interesting to both characters (it will either be too easy or too hard). The actual total modifier is irrelevant (and I, personally, prefer characters who actually advance in their abilities).

So if you want to make sure the "interesting range" for PCs overlap each other (so that there's a range of DCs you can meaningfully assign), then you need to:

(1) Figure out (and cap) the total bonus characters can get from outside sources (spells, equipment, etc.). You can exclude any bonuses which you're assuming all characters will get (for example, the fact that pretty much everybody gets a +X weapon).

(2) Subtract that number from 15. This determines the point at which you need to cap any divergence between "skilled" and "unskilled" characters. (If you've just generated a negative number, you've got a problem.)

3E allows #1 to get out of control.

Take attack bonus, for example. Give somebody a +5 sword. 15 - 5 = 10. At 20th level, the difference between "good BAB" and "low BAB" is 10. Which means you're right at the limit. But, of course, there are lots of other bonuses to your attack rolls. (You could probably expect to see another +10 from stat bonuses at those levels, for example, completely shattering the system.)

 

[howandwhy99]

What are the practical limits of d20+mod vs DC?

20 is an easy number and in a decimal system it works pretty well. Add 20, subtract 20, or anything smaller down to 1 in either case.

Other cases work too like multiply is double x10, divide is half x 0.1. move the decimal place or something like that. Smaller numbers, but over 10 and we get into some trouble for some players doing quick math in their heads. So we should try and avoid the last two functions.

When does the compare this number to another number get impractical? Well, I guess when we get off the natural number line or the numbers get so large they are actually a chore to read. But then we get shorthand notation like 8.3848 x10³⁴³ for example. Really, it's hard to impractical here.

But there are other elements we might be missing. Why not roll a d10 instead? It's decimal-based, smaller than 2 digits for multiplication and subtraction, but the drawback is in granularity. The 10% chunks are a little big. Still, this could work.

D100? D1000? D10,000? Not we get impractical, but some games use it still. Just order the d10s. But additions get way out of whack. Not to mention subractions or anything else. d100 percentiles have been in viable systems, but the bump: +1, is often too small for some people.

So we do d20. What can't it do well? Large modifiers make it irrelevant. d20+8,382 vs DC...... what? Who cares! The high roll is almost 42 times the modifier. So is the low roll. Why roll at all? You win! Just compare notes.

Do some games work like this? Yep, diceless games are more common now. Many games remove randomizing mechanics altogether. But do you want them? What are they to the game designer, the DM, the player? For many it's the perceived risk with the built in anticipation. "I need a 12 or better" *rattle* *rattle* *rattle* roll....

Die roll results are understood to be on average an even distribution. Make that distribution irrelevant to the game and the d20+mod vs. DC is no longer practical either. Not just because the functions are too hard or the numbers too large. The die roll is unnecessary and wasting players time.

My question to you is: What does the d20 distribution represent? What is a 1 or a 20 for the type of roll you are making? 10 might be a hit, 15 a difficult hit, 20 near impossible. Add your modifier and these get easier. Add a modifier over 18 and the roll effectively makes all DCs 20 and below irrelevant. Why face them? They are impossible to fail without a penalty.

Maybe that's okay though. Maybe we want to reach another tier of play where we reach another level? But the old stuff is like slapping mosquitoes. They end up as wake in our passage through the higher DCs. Another break happens at +39 and soon we're advancing without any regard for what's made irrelevant. It happens every level.

My suggestion is d20+mod vs. DC has an upper mod score of 18 and that it makes several other elements impractical or irrelevant with 2-3 times that score.

 

[pemerton]

So we do d20. What can't it do well? Large modifiers make it irrelevant. d20+8,382 vs DC...... what? Who cares! The high roll is almost 42 times the modifier. So is the low roll. Why roll at all? You win! Just compare notes.

But this isn't true. If the DC is (for example) 8392, then you need to roll 10+ on your d20 to succeed, in which case the die roll is not irrelevant, but crucial.

Big modifiers aren't a problem provided that the scaling is systematic on both the bonus side and the DC side. 4e aims for this, but (arguably) doesn't quite hit the mark.

30 character levels in 4e = 15 natural bonus to all numbers, across all classes.


If you start with that as a baseline, then everything else only becomes character specific--feats, weapon enhancements, skill focus, and so on.

But the real issue for 4e is, what exactly are the limits on, and variation across, these character specific bonuses? The fact that they've had to have 3 goes at setting suitable Easy, Medium and Hard DCs shows that WotC haven't found this an easy question to answer!

 

[prosfilaes]

But this isn't true. If the DC is (for example) 8392, then you need to roll 10+ on your d20 to succeed, in which case the die roll is not irrelevant, but crucial.

But only in an artificial environment will you have a 8382 + d20, and be asked to beat 8392. If that's your high jump skill, and you can jump 8382 + d20 feet, you can jump to the top of a cliff 8382 feet tall with no problem and can't jump to the top of a cliff 8402 feet tall with those in middle giving you problems.

Big modifiers aren't a problem provided that the scaling is systematic on both the bonus side and the DC side. 4e aims for this, but (arguably) doesn't quite hit the mark.

Any way you scale it, 8382 + d20 is going to be a problem. If your DC is always in the 8382-8402 range, what's the point in having a skill (or whatever) that high? It doesn't mean anything; your character is no better at doing things related to that skill in practice then when they had a 0 and DCs were 0-20.