2006 About.com Game Design Competition Friday, February 24, 2006

GameTableOnline and About.com have just announced the theme for their 2006 Game Design Competition: dice games. The winner's game will be added to GTO's online suite of games and gets a trophy of some kind. I've actually been looking for something like this, so I think I may give it a try. My Amber game doesn't use dice, so it would be nice to break out and think about something new for a little while. There are only a few limitations:

• Games must be designed for play on an easy-to-replicate game board, using checkers, Go stones, Chess pieces, Poker chips, dice, a standard deck of cards, or other items likely to be found in the average gamer's collection.
• Games must be designed for two players.
• Games must include dice as a central element. Acceptable dice are limited to: 4-sided, 6-sided, 8-sided, 10-sided, 12-sided and 20-sided.
• Prizes will be awarded to the top games, as chosen by a panel of judges. Among the prizes, the winning game will be put into the GTO development queue and the runner-up will receive a free six-month subscription to GameTable Online.
• All designers who submit a valid entry will receive a free three-month subscription to GameTable Online.
• The deadline for entries is March 31, 2006.

I think the key to winning the competition will be to make the dice run the game, and engage the player in a fun, new way. After all, the theme is all about the dice.

Luckily, I have a very interesting dice mechanic that I designed when I was about 13 years old that is simple, expressive and can involve lots and lots of dice. I liked the idea that you can use dice of different kinds together to generate interesting probability curves. And you got to roll lots of cool dice, and you got to roll them together.

The basic mechanic boils down to rolling your dice and taking the highest result. The "highest result" rule makes your curve always favor the highest result you can achieve with the most dice in your pool. Remember, on a single d6, the chance of rolling a 1 is the same as rolling a 6. It's always 1-in-6. But if you roll 2d6 and take the highest result, you have only a 1-in-36 chance of rolling a 1, and an 11-in-36 chance of rolling a 6. If you roll a d6 and a d8, your chance of rolling a 1 is 1-in-48, your chance of rolling a 6 is 11-in-48, and your chance of rolling an 8 is 6-in-48. d4/d6/d8 is even more interesting: your chance of rolling a 4 is 37-in-192 (about 19%), your chance of rolling a 6 is 44-in-192 (about 23%), and your chance of rolling an 8 is 24-in-192 (about 13%). Here's what I found:

	Probability of Rolling
Die Pool	4	6	8	10
d6	1/6 (17%)	1/6 (17%)	0	0
d6/d8	7/48 (15%)	11/48 (23%)	6/48 (13%)	0
d4/d6/d8	37/192 (19%)	44/192 (23%)	24/192 (13%)	0
d4/d6/d8/d10	175/1920 (9%)	364/1920 (19%)	360/1920 (19%)	192/1920 (10%)

From this, we can derive a few general rules. First, that the chance of rolling a 1 is relatively high with a single die, but decreases very quickly with multiple dice. Second, that the most likely outcome of a given throw depends on the number and kinds of dice, and there will be peaks and valleys in the curve if the number of sides are close to one another. Third, if you do not have overlaps in the number of dice that can roll a particular number, then the probability curve is flat.

In any event, the probabilities aren't much help without a game that gives some kind of meaning to the numbers, but maybe it's possible to back into a design based on manipulating probability curves. Maybe.

I'll give it some more thought, but at least the juices are flowing. I have an idea where each player will be allocating dice from a pool in order to generate particular numbers. It will probably use d4/d6/d8/d10/d12. This will ensure a wide range while still keeping the number of overlapping sides in the pool relatively large. More to come when I have a game idea as opposed to an unnatural fascination with probability curves.