Change SW code to have more uniform distribution of numbers roled

[clasone]

I know that the game is supposed to represent dice rolled with some variability and luck involved, but many times I see 3s and 11s in excess and 6s and 8s not rolled enough. If you rolled the 2 dice a thousand times you would get a normal distribution or bell curve. I am interested in the possibility of a software code change to the game, as on option, that would take most of the luck out of the game and self correct on the fly to follow the bell curve or normal distribution of numbers. In other words for 36 rolls you would get only one 2 and 12 rolled, two 3s and 11s rolled, three 4s and 10s rolled, four 5s and 9s rolled, five 6s and 8s rolled and six 7s rolled. This would self correct on the fly to ensure a normal distribution is maintained in each 36 roll sets. Again this would only be an option, the normal game could still be played. Thoughts please jump in.

[vamoscosmos719]

If the dice rolls is a normal distribution, then say for the first 16 rolls, no 8's rolls. Then for the next 20 rolls, you can expect the 8 to roll 25% of the time (5/20).

Knowing that information makes the game more predictive and can lead to situations where a number gets blocked just because for sure, its more likely to roll.

[euklid314]

I know that the game is supposed to represent dice rolled with some variability and luck involved, but many times I see 3s and 11s in excess and 6s and 8s not rolled enough. If you rolled the 2 dice a thousand times you would get a normal distribution or bell curve. I am interested in the possibility of a software code change to the game, as on option, that would take most of the luck out of the game and self correct on the fly to follow the bell curve or normal distribution of numbers. In other words for 36 rolls you would get only one 2 and 12 rolled, two 3s and 11s rolled, three 4s and 10s rolled, four 5s and 9s rolled, five 6s and 8s rolled and six 7s rolled. This would self correct on the fly to ensure a normal distribution is maintained in each 36 roll sets. Again this would only be an option, the normal game could still be played. Thoughts please jump in.

Some thoughts:

*) If the game author Klaus Teuber had wanted to implement your variant he would have added 36 cards in the game with the values 2,3,3,4,4,4,...11,11,12 and asked us to shuffle these 36 cards. "Rolling the dice" would have been replaced by drawing one of these cards and discarding it afterwards. If all 36 cards are played they are shuffled again.

*) Your variant would lead to a completely different game. It would be much more strategic and very much a memory game. If a player manages to keep track of the numbers that have already appeared, say, in the last 30 "rolls" he knows which numbers will turn up in the next 6 "rolls". He therefore can plan his moves (robber, ressource income, trades...) accordingly. The game would be divided into phases where (almost) nothing is known (1st move, 37th move, 2nd move, 38th move,...) to phases were there is (almost) no luck involved (36th move, 72th move, 35th move, 71th move,...)

*) In my opinion the game of CATAN would never have gotten that famous and big in such a variant - because it is not suited as a family game any more! The game got popular BECAUSE the dice are unfair (i.e., each roll is independent of all the previous rolls - thus for the small number of rolls during a single game some numbers will appear much more often and others much less often than they will appear long-term). With the help of dice every young kid can sometimes beat its parents even if it performs several subotimal game choices. Only in the long run optimal moves will be rewarded by more wins. Optimal moves do not necessarily win you a single game, though.

Since BGA does *never* implement house rules, your suggestion is interesting to discuss theoretically - but it will not be implemented on BGA.

[edit: It is not important in our discussion, but the dice distribution is not binomial, i.e. bell shaped, but it is triangular shaped - as you wrote yourself. 1/36 are 2s, 2/36 are 3s, 3/36 are 4s,...]

[Kestererer]

There is actually an existing expansion to Catan that does exactly this by replacing the dice rolls with a deck of cards - with the wrinkle that there are events associated with some of the cards, and five cards are not played from the deck each time through in order to retain some uncertainty. See https://boardgamegeek.com/boardgameexpa ... vent-cards

So this isn't just a house rule and shouldn't be excluded on that basis - though presumably implementation of that expansion is unlikely to be top of the priority list.