[Donjon] Dice Probabilities

[epweissengruber]

One of the members of IndieRPG Toronto has done a good job of breaking down the dice probabilities of Donjon.  I, Erik Weissengruber, Lord Grand Pooh Bah of Indie RPG Toronto, am posting it here.

Note, this is done by the

c) One of my concerns is that I suspect the dice mechanics in this game have a very significant 'whiff' factor; mathematically, having a lot of extra dice doesn't necessarily give you as much extra chance of winning a roll as you might think. However, it DOES make it so that if you do succeed, you will succeed with more successes. For example, if the GM is rolling between 5 and dice, here is the table of probabilities to succeed, and the probabilities to get at least X successes.

5 6 7 8 9 10 11 12
Any Success 51% 58% 59% 62% 65% 66% 68% 72%
> 1 32% 37% 38% 41% 47% 50% 53% 56%
> 2 14% 21% 25% 29% 33% 36% 39% 41%
> 3 4% 7% 10% 16% 21% 23% 27% 30%
> 4 0% 1% 2% 7% 10% 14% 17% 20%

I bring this up because one of the very important tactics in the game is "linking" tests, where you roll the successes from test over as dice into another. This is worth doing not so much because you will succeed much more often, but because it greatly increases your chances, if you do succeed, of succeeding BIG. That seems to me an important difference. Also, if you feel like you are whiffing a lot, just sit back and enjoy; its a design feature, not a flaw, and pretty much the best you can hope for, regardless of the number of dice you roll, is a 3 out of 4 chance of success.

[Hans]

For example, if the GM is rolling between 5 and dice, here is the table of probabilities to succeed, and the probabilities to get at least X successes.

5 6 7 8 9 10 11 12
Any Success 51% 58% 59% 62% 65% 66% 68% 72%
> 1 32% 37% 38% 41% 47% 50% 53% 56%
> 2 14% 21% 25% 29% 33% 36% 39% 41%
> 3 4% 7% 10% 16% 21% 23% 27% 30%
> 4 0% 1% 2% 7% 10% 14% 17% 20%

I did it through the brute force "make Excel roll A LOT of dice for me" method, not any kind of actual figuring out formulas or iterating through the entire probability space (which, for, say, 10 dice versus 10 dice, has ~2E+18 entries), and I have a more complete table if anyone is interested.

[Sonja]

See also the post from a couple of weeks ago http://www.indie-rpgs.com/forum/index.php?topic=20064.0 where I got mathematicians to answer the same questions for me.

Sonja

[Hans]

Thats not Donjon related math...THIS is Donjon related math!