dice used in catan is biased

[Jellby]

It's not wrong, it's random.

If it was over 1000000 rolls, you might have a point, but over less than 50 rolls it says absolutely nothing.

[Romain672]

ya my bad, still, equal % of 8 9 10 vs same % of all the rest of numbers, its wrong either way

That's not though. You can try to understand my breakdown, or ask questions about it.
Firstly I will repeat, but that cluster of 9s is the same than the cluster of rolling lots of 2s and rolling few of 7s, so that doesn't count.
And that cluster of 8s (41.3%) is extremely expected. In nearly all of your game (41%) you should roll 8s at least 7 times.

And secondly, something as weird should happen in average one game out of 5:
Rolling 9 times (or more) the number 10 in 44 rolls happen in 0.94% of games.
But now if you account the fact you could talk about rolling too few that number 10, or you could talk about any number that makes that probability roughly x22 (that's more complicated but for low probability that will be close, that's why I find 18.7%), which give one game out of 5.

So no, that game isn't weird at all.

[LightKnight]

Has BGA ever considered adding an alternative random generator and allowing its use in the settings in specific games? Sure it would alter the game, and you would definitely need the agreement of the editor, but BGA is a big website anyway, you could have that sort of power (not necessarily of getting the authorization, but at least of being answered). Provided BGA supports the idea (I can get it if people really dislike its spirit).
The idea would be to use a random generation system that would reduce the probability of a given result (for instance a 2d6 result, or a single d6 result) after it's appeared once. I suppose there exist such systems.
(or alternatively, the result of the dice would be taken out of a pool of possible outcomes, with outcomes not reappearing in the pool once taken out; in case a pool of outcomes would be used up, there would be a refill)
Such a system should never become the norm, but it could be an option to have it allowed as an alternative setting on some games (with enough warnings and so on). That would allow players that prefer controlled randomness over pure randomness to play in a way they prefer, and would potentially reduce the complaints (I feel like half of the complaints in the forum are over randomness, especially here in the Catan subsection - I'm not sure why it affects Catan more than other games).
I think it could be an idea. What do you think?

[Jellby]

The game is designed with dice, and dice behave like dice.

If the designer wanted a different kind of "randomness" they would have used a different system. For instance, a 12-sided die if they wanted all numbers equally probable, or a deck of cards or a bag of tokens if they wanted some "memory" about previous results (I believe there's some variant/expansion of Catan that uses such cards).

[ufm]

I think it could be an idea. What do you think?

This idea is not even good on paper: this cannot persuade conspiracy theorists at all. They just want to play a blame game.
I heard the same thing in other platforms using cards, tiles, and etc.