Strong Fighter, Wimpy Mage

[KarinsDad]

Right, that's what we're saying, a single d20 is a huge variance, especially for low modifier or DC rolls. All you are doing by requiring multiple rolls is normalizing the data. In that vein, and to the OP's cause, I would suggest rolling 3 d20 for every check and using the mean average...it generates that bell curve and reduces extreme results.

2D10 does this as well, only faster with no division. The difference is that the curve is not as steep.

And, 2D10 + Ability Score instead of D20 + Ability Modifier works even better. A 12 Strength character does have a slightly better chance of success than an 11 Strength character.

 

[Nebulous]

I'm just trying to help the OP. Personally, I don't have problems with statistical anomalies.

And it's much appreciated. There's been some good and useful feedback here.

Right, that's what we're saying, a single d20 is a huge variance, especially for low modifier or DC rolls. All you are doing by requiring multiple rolls is normalizing the data. In that vein, and to the OP's cause, I would suggest rolling 3 d20 for every check and using the mean average...it generates that bell curve and reduces extreme results.

And what about removing the roll altogether in some circumstances? If you have at least X ability score or modifer, you succeed. And is the 3d6 solution just as good as using the 3d20?

 

[ThirdWizard]

I don't have my PHB handy, but isn't there an example in the skills section of an arm wrestling match being decided by checking the two opponents' strength scores and the higher one wins, with no dice rolls or random chance thrown in?

The RAW, IMO, is on the side of comparing the scores of the troll vs. the figthters and deciding with no rolls.

 

[prosfilaes]

Ah, but this is adding to the current rules in order to change the outcome based on the problem with the current rules.

I don't see why. A tug-of-war is a combat over an extended period of time. I don't see why it has to be resolved in the current rules with one single roll.

 

[werk]

And what about removing the roll altogether in some circumstances? If you have at least X ability score or modifer, you succeed. And is the 3d6 solution just as good as using the 3d20?

3d6 reduces the range to 3-18, so each point isn't 5% it's 6.25%, and creates the curve, so it would definitely make everything more stable/predictable. But changes like this will ripple through the whole d20 system as everything will need to be slightly adjusted...critical ranges, DCs, AC, all kinds of things. You wouldn't be able to just use anything out of a book without first considering if it needs to be fudged up or down a point or three...then that is completely arbitrary, and you're back to unpredicability.

 

[Tiberius]

White Wolf systems do this. One problem is, it's statistically fragile; in several of their games, you're more likely to crit fail at times the more dice you have.

I don't think this is a problem of the math so much as a problem with the Rule of One in the old versions of the ruleset. When "1"s cancel out successes, you do run into problems where more dice = more 1s = more botches/failures, but I seem to recall seeing some graphics recently in regards to Exalted indicating that this problem is a thing of the past.

 

[KarinsDad]

3d6 reduces the range to 3-18, so each point isn't 5% it's 6.25%, and creates the curve, so it would definitely make everything more stable/predictable. But changes like this will ripple through the whole d20 system as everything will need to be slightly adjusted...critical ranges, DCs, AC, all kinds of things. You wouldn't be able to just use anything out of a book without first considering if it needs to be fudged up or down a point or three...then that is completely arbitrary, and you're back to unpredicability.

You would have to do this with 3D20 as well since although the range does not change, the results do. The chances of rolling a 20 to threaten with a Club would be 1 in 8000.

The concept of using a curve for skills and for ability checks is ok since there are no special boundary rules such as treatening on a 20 or always failing a save on a 1. Some DCs might, however, have to be lowered slightly for skills, but nothing would have to be done for opposed ability checks.

The concept of using a curve for to hits, saves, etc., does not make sense due to all of the special boundary rules for them.

However, 2D10 is generally preferable (IMO) to 3D6 anyway since 3D6 has fewer "frequent" results and has too sharp of a curve.

The basic results (i.e. 4% chance or better) for 2D10 are: 2-4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18-20. 16 different basic results.

The basic results for 3D6 are: 3-5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16-18. 12 different basic results. This might be good for opposed ability checks, but would not be that good for skills. I actually prefer 2D10 for opposed ability checks using the ability score instead of the modifier.

 

[KarinsDad]

I don't see why. A tug-of-war is a combat over an extended period of time. I don't see why it has to be resolved in the current rules with one single roll.

Who says that a Tug of War is over an extended period of time? In the Baby vs. the Troll case, the Troll should be able to yank the Baby over with a single pull, every single time. Less than one round. One roll.

 

[KarinsDad]

I don't have my PHB handy, but isn't there an example in the skills section of an arm wrestling match being decided by checking the two opponents' strength scores and the higher one wins, with no dice rolls or random chance thrown in?

The RAW, IMO, is on the side of comparing the scores of the troll vs. the figthters and deciding with no rolls.

Yes. This rule was quoted earlier in this thread.

The issue becomes one of plausibility (and heroic fun). Is it fun and plausible that a 23 Strength Troll will win every single time over a 22 Strength Human and that the Troll will also win in a single round? If so, why do we have dice rolls for Grapples, Bullrushes, etc.?

 

[Lamoni]

So in conclusion, the rules work great for opposed checks when
a: There should be a lot of randomness to the result (roll a d20)
b: There is no randomness (no rolling)

You need a house rule if you want to fill in the gap. It would be nice if they addressed this as an optional rule or something. One simple house rule that you might want to consider is to reduce the die as you reduce the randomness of the event.
Therefore: Opposed rolls like Bluff and sense motive can remain at d20
Holding/pushing a door might be reduced to a d10
Arm wrestling or tug of war would probably be best with a d4 (there is a little technique involved).

This keeps it simple in that you only have to roll one die and you don't need to add or average anything. However, it adds complexity to the number of different dice you have to roll. I like it though since it gives some of the less used dice a chance to shine