Dice games with other than d6 dice?

May 30, 2009 3:12 PM

I just picked up a rare d5 and d7 to augment my rpg dice collection. I want to use them, but for more than just table-top rpgs. Are there any dice games that are played with more than just d6 dice?
posted by hpliferaft to sports, hobbies, & recreation (13 comments total)

Ooh.... adapt Yahtzee to be played with one each d2, d3, d4, d5, d6, and d7 (most people use a coin for a d2). That would be fun.

Play any board game that uses a pair of d6s and just sub in the d5 and d7. The min and max rolls are different, but the average roll isn't and it would be fun to try.
posted by Night_owl at 3:23 PM on May 30

The min and max rolls are different, but the average roll isn't and it would be fun to try.

I like that idea, but the min and max don't change (they're still 2 and 12). I think the odds of any total-number remain the same too (I count the same combos), but I'm no math savant so I might be wrong there.
posted by rokusan at 3:42 PM on May 30

Gack. Rokusan, you're right. I meant to say, the min and max rolls are the same, but I think the average roll and the odds for any particular number combo is different.
posted by Night_owl at 3:56 PM on May 30

I think the average rolls are the same. I did up an excel spreadsheet comparing the average of a d5 added to a d7 with an added pair of random d6's and the averages came up the same. I did it for several hundred rolls and I think I did it right.
posted by codswallop at 4:07 PM on May 30

I'm pretty sure the probabilities and means are most certainly different (though slightly) as follows...

For 2d6: 2 (1/36), 3 (2/36), 4(3/36), 5(4/36), 6 (5/36), 7 (6/36), 8 (5/36), 9 (4/36), 10 (3/36), 11 (2/36), 12 (1/36). Average roll = 6.66667.

For d5+d7: 2 (1/35), 3 (2/35), 4(3/35), 5(4/35), 6 (5/35), 7 (5/35), 8 (5/35), 9 (4/35), 10 (3/35), 11 (2/35), 12 (1/35). Average roll = 6.65714.
posted by drpynchon at 4:11 PM on May 30

Wow, um, who the hell posted such idiotic drivel under my login?!
posted by codswallop at 4:15 PM on May 30 [2 favorites]

For future reference, monte carlo simulations usually only provide estimates of equivalency for probability. They're never rigorous proofs.
posted by Precision at 4:24 PM on May 30

For future reference, monte carlo simulations usually only provide estimates of equivalency for probability. They're never rigorous proofs.

The corollary being: don't ever post results of Monte Carlo on the internet, because then I'll hate you - use them to check a hypothesis before you set about proving it.

I think the average rolls are the same. I did up an excel spreadsheet comparing the average of a d5 added to a d7 with an added pair of random d6's and the averages came up the same. I did it for several hundred rolls and I think I did it right.

The average roll on a d5 is 3. The average roll on a d7 is 4, and therefore the average roll of d5+d7 is 7. Similarly, the average roll on a d6 is 3.5, and so the average on 2d6 is 7.

So yes, the averages are the same. The only "problem" is that the probability distribution has a little more variance. That is, you're more likely (though not by much) to see the extreme values, rather than the values near the mean. All in all, if you're in the unlikely situation of having a d5 and a d7 sitting around, but no d6's, then you'll probably be okay with the substitution.

Of course, this sort of substitution doesn't always work: d1 + d11 has a very, very different distribution from 2d6, despite having identical means and ranges.
posted by TypographicalError at 4:33 PM on May 30

This thread on boardgamegeek is about games (including dice games) that use odd dice. It looks like Senet can be modified to use a 5-sided die, but it's not a dice game. An medieval predecessor to backgammon may have used 7-sided dice.
posted by jedicus at 4:36 PM on May 30

I guess I'll do something novel in this thread and answer the poster's question.

Try Button Men or The Big Cheese or the various games here. Or search boardgamegeek for "games with polyhedral dice".
posted by inkyz at 4:38 PM on May 30 [2 favorites]

Not to derail but drivel on my end too. I made an arithmetic error (ignored the rolls of 7) and the averages actually are the same (7). The actual probability distributions are not. All of that said, to actually answer the question, they are close enough to use d5+d7 as an alternative to 2d6.
posted by drpynchon at 5:28 PM on May 30

What a difficult post to moderate.
posted by hpliferaft at 11:09 PM on May 30 [1 favorite]

[Folks, I love probabilistic noodling too but enough with the probabilistic noodling. The question is about games to play, please try to focus on that.]
posted by cortex at 6:46 AM on May 31