change the dice generator.

[Roskill1]

"I don't know who created the dice generator (or who made the decision to use it), but it is truly a mess. It only leads to discrimination against some players who theoretically (statistically) have a chance to win. However, it can generate more extreme numbers like 11 and 3 than 6 and 8 in some games. So please kindly change this crappy dice generator."

[Antonio_I]

I agree.

[Cos-]

Please please please cut it out with the endless series of dice threads on this forum. The dice have a very reasonably random distribution (even though it's probably pseudorandom). There is nothing whatsoever wrong with the dice. Every single person who complained and has been challenged to show evidence did not show any evidence, they only showed that they don't understand random dice will ALWAYS show some sequences that look like patterns, and will cause some less likely numbers to show up more often than some more likely numbers in a large majority of games. That's how randomness works. It's not supposed to be an evenly probabilistic distribution, it's supposed to be RANDOM.

We have plenty of HUGE long threads about this already on this forum. We don't need more. Read the threads that are already here, and don't add to them unless you have real actual evidence. And don't claim one game you recently played is evidence until you actually read and understand the numerous long threads that are already on this forum.

[CaptainDex]

I agree with the opening comment, the dice basically roll in such a way at the beginning that there is a very high chance that each player will receive the knight to protect their city, this is very noticeable across all games. Later in the game, this becomes noticeably relativized.


It is also very strange when the dice roll which result, so for around 40 rounds there is no "6", but 5 times in 10 rolls there is a "9" and then the even worse player has passed everyone. The highlight are games in which, in the early stages, a player comes up with numbers whose probability is low to very low (example 3 or 4) 6-7 times in a row. Either we are being manipulated externally or the programming is bad or to the advantage of individuals.


Proofs? No problem!

https://boardgamearena.com/table?table=623814376
https://boardgamearena.com/table?table=623135865

Please don't explain to me that I may have left the game earlier, that doesn't change my statement, it just confirms the frequency with which numbers with a bad probability are rolled at short intervals and yes, this is not a feeling, but shows exactly what I say.

If a "4", 6x - 7x comes out of 15 throws, you could easily call it a coincidence (if it were only in one game), but this phenomenon can be observed more often.


This is the best...

https://boardgamearena.com/table?table=622575649

[Suoivax]

I agree with the opening comment, the dice basically roll in such a way at the beginning that there is a very high chance that each player will receive the knight to protect their city, this is very noticeable across all games. Later in the game, this becomes noticeably relativized.


It is also very strange when the dice roll which result, so for around 40 rounds there is no "6", but 5 times in 10 rolls there is a "9" and then the even worse player has passed everyone. The highlight are games in which, in the early stages, a player comes up with numbers whose probability is low to very low (example 3 or 4) 6-7 times in a row. Either we are being manipulated externally or the programming is bad or to the advantage of individuals.


Proofs? No problem!

https://boardgamearena.com/table?table=623814376
https://boardgamearena.com/table?table=623135865

Please don't explain to me that I may have left the game earlier, that doesn't change my statement, it just confirms the frequency with which numbers with a bad probability are rolled at short intervals and yes, this is not a feeling, but shows exactly what I say.

If a "4", 6x - 7x comes out of 15 throws, you could easily call it a coincidence (if it were only in one game), but this phenomenon can be observed more often.


This is the best...

https://boardgamearena.com/table?table=622575649

3 games? that's your statistical proof? 3 games out of the tens of thousands that get played here?

I'm pretty sure that if i dug deep enough i could find a full game of Catan in which ONLY 5's were rolled for the whole game. does that mean something is broken? No, it just means i found a really wierd game.

[Jellby]

I have "proof" that it's right: https://boardgamearena.com/table?table=624376864
The largest deviation from the ideal distribution is 4%, which is very small.

[gmrsmrt]

....

Proofs? No problem!

https://boardgamearena.com/table?table=623814376
https://boardgamearena.com/table?table=623135865

Please don't explain to me that I may have left the game earlier, that doesn't change my statement, it just confirms the frequency with which numbers with a bad probability are rolled at short intervals and yes, this is not a feeling, but shows exactly what I say.

If a "4", 6x - 7x comes out of 15 throws, you could easily call it a coincidence (if it were only in one game), but this phenomenon can be observed more often.


This is the best...

https://boardgamearena.com/table?table=622575649

Proof mean show with some statistical test that the dice don't respect the expected distribution WITHIN STATISTICAL FLUCTUATIONS. So you
should establish some observable compute the expected value and check that this is not respected from the games you quote.

The three games you quote have a probability of
0.011
0.11
0.000076
to occur. (What I mean is that those are the probabilities that you find an outlier that is at least as extreme as the observed one)
The first two have nothing noticeable, similar sequences should occurr many times a day. The last one is indeed quite weird, however 76 per million
should have occurred many times on BGA due to the number of played games. You picked up a very rare event, but that's not enough to claim a problem.

[Romain672]

0.000076
to occur. (What I mean is that those are the probabilities that you find an outlier that is at least as extreme as the observed one)

Can you explain more what you mean? I disagree with your sentence.
Rolls:
2 3 4 5 6 7 8 9 10 11 12
0 1 8 0 3 5 0 2 0 0 0

I find a probability of 0,0000757 to roll eight 4 or more in 19 rolls.
Which give 1/13 204 to roll eight 4s or more.

But here we are still taking a random game and looking for rolling too much 4s.

If now you look at all numbers and try to find a number too much or too few rolled, I like to multiplay that probability per 22 (any number rolled too much or too few):
1-((1-0,0000757)^22) which now gives 0.00166, so 0.17%.

Which mean that you should see any number rolled too much or too few more extreme than on that game 1 times out of 588.

The last thing I do with that probability when it's low, is to look at the profile, to find how many games they have played. Here we are at 458 games, so yeah, if it's really the most extreme game, we are around what we expect. It's just slightly under the average.
But it's rare on the forum to have an example pretty low like that one. But we don't have anything weird. If something is weird, you should be able to point something specific we could check, and it's likely that we would find probas of 1/1M. We are still at 1/588 here.

Note that here, since you leave the game, you add a biais in the calculations. Because if you stop when you see that something is weird, you don't let the dices be 'more close to what you expect' over time.
If you compare one player who always play a game with 100 rolls.
And another one which as he wish play a game with 10/20/42/100 rolls and stop when he feel that something is wrong, then you will find on their games much more weird results.

[Steve725]

I agree.

Yeah I just had a game where there were 27 8s and 2 5s, I based my strategy on those 5s and was sitting without any resources. So 5 has probability of 11 percent or whatever and rolled 2 out of 60-70 rolls, it is broken

[CaptainDex]

According to my probability calculation, the chance of getting such a game is ~1:500,000, and in only 11% of these games the player would have a chance of getting 5 of the 8 fours rolled, so the value is far outside the norm at ~1:5,000,000 in around 4,600,000 games on BGA, a true phenomenon.
Add to that the luck of occupying both four fields at the beginning.