Dice Alternatives in Unearthed Arcana

[Cthulhudrew]

(It isn't a rules question, per se, but a question about if people are using a certain rule. Apologies in advance if this thread is in the wrong forum.)

Looking through Unearthed Arcana last night again, and noticed the "Players rolling dice" alternatives they have. They sound intriguing, and I was just curious if anyone else has implemented them, and how it has affected your games? Good? Bad? No real difference?

 

[Greatwyrm]

Good? Bad? No real difference?

In a 2e game I ran, I used what basically amounts to the Defense Roll variant in that section. It did help out when running larger battles, like if the PCs were outnumbered 3 or 4 to one. You could basically just say, "okay, everyone make three defense rolls and tell me if any of them aren't at least X," instead of rolling 30 or so attacks and asking for everyone's AC.

The only time I would hesitate to use this is with a BBEG or similar boss creature. I think the DM should be rolling the dice there so that the PCs have a little less knowlege of the opponent's abilities. Also, it's tough to fudge rolls in a PCs favor when the player himself is rolling them right in the open. But to speed up low level combats with obviously lesser powered opponents, it was great.

 

[Aaron2]

Good? Bad? No real difference?

Good. I've played that way with Champions as well. One thing, you need to add 12 instead of 11. Since I roll in the open anyway, there is no problem with fudging. Plus, I found that it is just as easy for players to figure out an Attack Bonus and AC as with the normal rolling system.


Aaron

 

[Cthulhudrew]

The only time I would hesitate to use this is with a BBEG or similar boss creature. I think the DM should be rolling the dice there so that the PCs have a little less knowlege of the opponent's abilities. Also, it's tough to fudge rolls in a PCs favor when the player himself is rolling them right in the open. But to speed up low level combats with obviously lesser powered opponents, it was great.

That's an interesting idea- to do the Big Bad's rolls yourself, and let the players do the rest. I was wondering how that might work out. Sounds like a very cool variant of the variant.

Good. I've played that way with Champions as well. One thing, you need to add 12 instead of 11.

Why 12 instead of 11? Is it just a mathematical thing they messed up? Should it be in the errata?

Also, while we're on the topic, has anyone tried the "Bell Curve" dice variant in UA? I'm not sure if I like it or not. It sounds kind of cool, but at the same time, kind of unnecessary.

 

[WaterRabbit]

I like this variant myself and tried it way before UA was published. The real trick is breaking yourself of the habit of rolling the attack roll. The only problem is rolling for damage is slightly awkward since I usually roll both the attack and damage rolls together.

Why 12 instead of 11? Is it just a mathematical thing they messed up? Should it be in the errata?

It is 11. He is just in error on this point. I assume he meant 11 instead of 10. Using 12 will skew the probabilities.

Also, while we're on the topic, has anyone tried the "Bell Curve" dice variant in UA? I'm not sure if I like it or not. It sounds kind of cool, but at the same time, kind of unnecessary.

I actually prefer 2d10 instead of 3d6 or 1d20. The idea behind rolling multiple dice is that it makes the bonuses and penalties to the attack roll more meaningful. A +1 sword is actually worth more when rolling 2d10 over 1d20. It also makes critical hits much more rare. With a crit. range of 1, you get a possible critical hit 5% of the time with 1d20. However, with 2d10 you get a possible critical only 1% of the time and only .5% of the time with 3d6. Conversely if you use fumbles it reduce those as well.

The reason I prefer 2d10 over 3d6 is that the probabilities are easier to calculate. The reason they use 3d6 is that the average is 10.5, the same as 1d20, as opposed to 11 with 2d10.

 

[Aaron2]

Why 12 instead of 11? Is it just a mathematical thing they messed up? Should it be in the errata?

Let's take the simplest example. An attacker with +0 will hit a defender with an AC of 10 on a 10-20. That is 11 out of the possible 20 rolls or 11/20 or 55%. When we reverse the rolls, the defender now wins the ties, but he should still only succeed 45% (100%-55%) of the time. 45% is 9/20 or 12-20. So, the defender needs a 12 or better to avoid being hit.

I've been doing this for several years and I don't have UA yet. They might have changed it so the defender doesn't win ties, which is the opposite of every other die roll in the entire system. If so, then you add 11 but that seems like adding an extra step to me.

Sorry to add to the confusion.

http://jtjaguar.home.texas.net/playersroll.html


Aaron

 

[fuindordm]

I actually prefer 2d10 instead of 3d6 or 1d20. The idea behind rolling multiple dice is that it makes the bonuses and penalties to the attack roll more meaningful. A +1 sword is actually worth more when rolling 2d10 over 1d20. It also makes critical hits much more rare. With a crit. range of 1, you get a possible critical hit 5% of the time with 1d20. However, with 2d10 you get a possible critical only 1% of the time and only .5% of the time with 3d6. Conversely if you use fumbles it reduce those as well.

The reason I prefer 2d10 over 3d6 is that the probabilities are easier to calculate. The reason they use 3d6 is that the average is 10.5, the same as 1d20, as opposed to 11 with 2d10.

I tried this too in a HERO campaign I ran once, and it worked quite well. I was finding that 3d6 didn't give enough variation in the dice.

The problem I was having in the campaign was this: when there was a difference of
several skill/BAB/AC ranks between the PCs and the opponents, then the PCs would
hit 99% of the time and the opponents would miss 99% of the time on 3d6. The result was that I couldn't inflict much attrition damage on them from low-level opponents, and all the encounters had to be at the challenge level of a Big Boss to be any fun.

Switching to 3d6 rolls in D&D might lead to the same behaviour. You have to remember that almost all your rolls will be between 6 and 15, so you've actually cut the range of possible results on your dice in half. Your lowly orc (BAB +3) cannot
reasonably expect to hit ACs higher than 18, whereas with a d20 he can hit AC 20 15% of the time.

d20 gives you lots of variation and makes long-shot rolls more likely (and the game a bit more exciting for that). The downside is that small differences between characters in ranks/BAB don't have much effect.

3d6 gives you less variation in rolls, and makes long shots very unlikely, very memorable, and something rarely attempted by the PCs. The upside is that skill ranks have a huge impact on the probabilities; a difference of only one rank between PCs becomes really palpable.

2d10 gives you something of a happy medium between the two. Skill ranks
increase in importance, you lose a little dynamic range in the results, but
long shots are still common enough that they can have an impact on the game.

Threat ranges aren't difficult to deal with either.
20 (5%) --> 19-20 (3%) or 18-20 (6%)
19-20 (10%) --> 17-20 (10%)
18-20 (15%) --> 16-20 (15%)
17-20 (20%) --> 15-20 (21%)
15-20 (30%) --> 14-20 (28%)
You can always get a threat range within 2% of the original probability,
and two of the most common are actually exact. You can even keep the
rule "Keen doubles the threat range", which you can't in 3d6.

You could even add a little extra spice at the top, such as :
natural 19: roll twice for damage, take best result.
natural 20: crit automatically confirmed.

--Ben

 

[Aaron2]

I've been doing this for several years and I don't have UA yet. They might have changed it so the defender doesn't win ties, which is the opposite of every other die roll in the entire system. If so, then you add 11 but that seems like adding an extra step to me.

I finally broke down and got AU today and there was not exception, the defender still won ties. So it should be +12 after all. Look at the Ogre example they use. He has +8 so will hit an AC 10 on a 2 or better; he only misses on a one. If you use there example of an attack DC of 19, the attack will miss if the defender rolls a 19 or 20. The DC should be 20.


Aaron

 

[woodelf]

That's an interesting idea- to do the Big Bad's rolls yourself, and let the players do the rest. I was wondering how that might work out. Sounds like a very cool variant of the variant.

Why 12 instead of 11? Is it just a mathematical thing they messed up? Should it be in the errata?

Or change who wins ties. One point of assymetricality comes from the fact that the average for d20 is 10.5, not 10, while the other point comes from the tie rules. (IIRC, make it "die-roller loses ties" rather than "defender wins ties", and it's all fixed. And, IMHO, it's a more symmetrical/consistent solution than the +12 solution.)

In any case, i think it's kinda silly to worry about one point, even a consistent point, when you're rolling a d20 (wide, flat randomizer). I've used essentially this variant for years, 'cept i just add 10 all-round. Fine, so the players get a point or two of advantage (usually) or disadvantage. So what? It's less than the modifier for any miscellaneous minor circumstance, or the difference between a strong and really strong person, or one who has put lots of points into a skill and one who has only put some points in, or any number of other variables.

Also, while we're on the topic, has anyone tried the "Bell Curve" dice variant in UA? I'm not sure if I like it or not. It sounds kind of cool, but at the same time, kind of unnecessary.

If you want a bell curve, there's a much better way to do it. It's called mid20: roll three d20s, toss the highest and lowest, and use the remaining one for your result. No addition, you get the full range from 1 to 20, and the curve is smoother than 2d10 and less extremely center-weighted than 3d6. And you can pretty much just toss it into the system without any further changes--though keep in mind that crits will of course get rarer (but no where near as much rarer as 3d6, which is why you can probably get away with not changing the threat ranges). You'll probably want to use a regular d20 for confirming threats, or crits will get *really* rare. We've also explored some interesting variations you can do with the dice to roll threat-confirmation into the basic die roll, and even give multiple degrees of success right in the die roll, but then it actually starts to make the probabilities very much different, and you would have to take them into consideration a bit more.

Oh, some probabilities, for those who're wondering:

value	d20	3d6	3d6+2dF	2d10	mid20	mid20 ≥
1	0.05	--	0.001	--	0.007	1
2	0.05	--	0.003	0.01	0.021	0.993
3	0.05	0.005	0.008	0.02	0.033	0.972
4	0.05	0.014	0.017	0.03	0.043	0.939
5	0.05	0.028	0.031	0.04	0.052	0.896
6	0.05	0.046	0.049	0.05	0.06	0.844
7	0.05	0.069	0.071	0.06	0.066	0.784
8	0.05	0.097	0.093	0.07	0.07	0.718
9	0.05	0.116	0.11	0.08	0.073	0.648
10	0.05	0.125	0.119	0.09	0.075	0.575
11	0.05	0.125	0.119	0.1	0.075	0.5
12	0.05	0.116	0.11	0.09	0.073	0.425
13	0.05	0.097	0.093	0.08	0.07	0.352
14	0.05	0.069	0.071	0.07	0.066	0.282
15	0.05	0.046	0.049	0.06	0.06	0.216
16	0.05	0.028	0.031	0.05	0.052	0.156
17	0.05	0.014	0.017	0.04	0.043	0.104
18	0.05	0.005	0.008	0.03	0.033	0.061
19	0.05	--	0.003	0.02	0.021	0.028
20	0.05	--	0.001	0.01	0.007	0.007

 

[WaterRabbit]

Let's take the simplest example. An attacker with +0 will hit a defender with an AC of 10 on a 10-20. That is 11 out of the possible 20 rolls or 11/20 or 55%. When we reverse the rolls, the defender now wins the ties, but he should still only succeed 45% (100%-55%) of the time. 45% is 9/20 or 12-20. So, the defender needs a 12 or better to avoid being hit.

I've been doing this for several years and I don't have UA yet. They might have changed it so the defender doesn't win ties, which is the opposite of every other die roll in the entire system. If so, then you add 11 but that seems like adding an extra step to me.

Sorry to add to the confusion.

http://jtjaguar.home.texas.net/playersroll.html


Aaron

Ok, let’s take your example.

Normal:
d20 + 0 >= 10 yields a 55% chance to hit. (10-20 or 11/20)

As a Defense Roll:

Condition: 1 d20 + 0 > 11 yields a miss or the attacker hits on a 1-11 or 11/20.

Condition: 2 d20 + 0 >= 12 yields the same result.

The book result is incorrect in stating “If that defense check equals or exceeds the attack score of the enemy, the attack misses.” It should state “If that defense check exceeds the attack score of the enemy, the attack misses.”

I have always used condition 1, because it seems more natural to tell the play to “Beat an 11” rather than “Make a Defense Roll DC 12”.

Six one way, half a dozen the other.