Dice generator - the non-conspiracy take

Forum rules
Please DO NOT POST BUGS on this forum. Please report (and vote) bugs on : https://boardgamearena.com/bugs
User avatar
Coleus1
Posts: 15
Joined: 22 April 2025, 14:51

Re: Dice generator - the non-conspiracy take

Post by Coleus1 »

Only on GBA can a number either get rolled way more often than others, or the other way around. Look at different championsships etc, play the game in real life. It only happens here. I played a game where only 11 was rolled, and one knew from the start that blue player would win. The other game, only two 9s were rolled, because orange (the player the game tries to favor and you knew would win from the start of them game) did not possess 9. If it happened once in a while that wouldn't be wierd, but it happens all the time on here.
User avatar
Jontia
Posts: 236
Joined: 06 March 2014, 22:59

Re: Dice generator - the non-conspiracy take

Post by Jontia »

Coleus1 wrote: 26 June 2025, 09:47 If it happened once in a while that wouldn't be wierd, but it happens all the time on here.
It doesn't happen all the time. You just remember the times it happens to you and forget all about the times it doesn't happen and ignore the huge variation you've lumped in together to count as weird.

Saying that game was weird there were no sixes after it happens is meaningless. Because you would have said that if there were no 5s, 6s, 8s or 9s, or 10s... And the more numbers you add in the more likely a 'weird' event is to show up. And given that it counts as weird if there's 0, 1 or 2 of these numbers you end up with weird games being perfectly normal.
User avatar
Jontia
Posts: 236
Joined: 06 March 2014, 22:59

Re: Dice generator - the non-conspiracy take

Post by Jontia »

You don't have to take my word for it though. Take a prompt like this off to Gemini, or the AI generator of your choice.
Take the code it spits out to an online python runner like replit or online-python and see what it spits out.
Just make sure you're honest with yourself beforehand about what you think is "weird" and what percentage is "unlikely".

Feel free to tweak the number or rolls in a game or the number of games in the simulation. And definitely work out what you think is weird.
But this is relatively simple stuff and an AI will spit out code that does this without breaking a sweat.
I would like some python code to;

Roll two six sided dice 40 time recording the total and storing the results - this is one game.
Repeat this 10,000 times. - play 10,000 games.

Print out the number of times games in which each total was seen 0, 1 or 2 times, please include the percentage of games this was in the output.

Finally I'd like to define a weird game as one where any one or more of the numbers 5,6,7,8 or 9 is rolled less than the number 12. I'd like a line that prints the number of weird games as an absolute and a percentage of the data set.

Also write out the flip of the weird games output to say 'one in X games will be weird.' Include 0 decimal places in X
User avatar
Theana
Posts: 62
Joined: 01 July 2016, 21:35

Re: Dice generator - the non-conspiracy take

Post by Theana »

Adding an example about the only glitch I can notice and I see happening but never tracked.

I am not saying there is something wrong, just sharing the table and the move

Table #695777898
Move 31
https://boardgamearena.com/archive/repl ... s=84261660;
User avatar
euklid314
Posts: 679
Joined: 06 April 2020, 22:56

Re: Dice generator - the non-conspiracy take

Post by euklid314 »

Theana wrote: 03 July 2025, 11:13 Adding an example about the only glitch I can notice and I see happening but never tracked.

I am not saying there is something wrong, just sharing the table and the move

Table #695777898
Move 31
https://boardgamearena.com/archive/repl ... s=84261660;
There is no glitch in move 31, the dice are rerolled because of the variant chosen in the game.
User avatar
Theana
Posts: 62
Joined: 01 July 2016, 21:35

Re: Dice generator - the non-conspiracy take

Post by Theana »

euklid314 wrote: 03 July 2025, 12:43

There is no glitch in move 31, the dice are rerolled because of the variant chosen in the game.
sorry, the variant is to roll three 7s in a row? This is weird.

Or you just assumed that I am stupid and I don't know to read and play?
User avatar
Jontia
Posts: 236
Joined: 06 March 2014, 22:59

Re: Dice generator - the non-conspiracy take

Post by Jontia »

Theana wrote: 03 July 2025, 13:50
euklid314 wrote: 03 July 2025, 12:43

There is no glitch in move 31, the dice are rerolled because of the variant chosen in the game.
sorry, the variant is to roll three 7s in a row? This is weird.

Or you just assumed that I am stupid and I don't know to read and play?
Three sevens in a row is not that usual. Even just rolling two dice three times this would happen once in every 200 attempts. A 30 round games gives 10 goes at that even totally ignoring overlapping possibility. So you'd see this in one in every twenty games.

Quick python mod later...
The linked game seems to be 70 dice rolls, including the re-rolls.
Chance of three 7s in a row in a 70 roll game? 23.5% Every 4 or 5 games has this streak.

It is not weird.
User avatar
Theana
Posts: 62
Joined: 01 July 2016, 21:35

Re: Dice generator - the non-conspiracy take

Post by Theana »

Jontia wrote: 03 July 2025, 13:56 Three sevens in a row is not that usual. Even just rolling two dice three times this would happen once in every 200 attempts. A 30 round games gives 10 goes at that even totally ignoring overlapping possibility. So you'd see this in one in every twenty games.

Quick python mod later...
The linked game seems to be 70 dice rolls, including the re-rolls.
Chance of three 7s in a row in a 70 roll game? 23.5% Every 4 or 5 games has this streak.

It is not weird.
Thanks Jontia, this is the only type of "weirdeness" that I see sometimes, the same number rolled multiple time in a row and as I said this is the only time I took mental note of it and tried to understand if it is worth of keeping track of it.

To be clear I don't believe in any conspiracy theory!
User avatar
euklid314
Posts: 679
Joined: 06 April 2020, 22:56

Re: Dice generator - the non-conspiracy take

Post by euklid314 »

Theana wrote: 03 July 2025, 13:50
euklid314 wrote: 03 July 2025, 12:43

There is no glitch in move 31, the dice are rerolled because of the variant chosen in the game.
sorry, the variant is to roll three 7s in a row? This is weird.

Or you just assumed that I am stupid and I don't know to read and play?
I did not consider that you might find three 7s weird. Sorry.

After each 7, there is a 2.8% chance that two more 7s will follow. That is not rare and if you think it happens more often on BGA you should elaborate your reasoning.

Please note that I did talk about the probability of 2.8% for two 7s and not of 0.5% of having three 7s in a row. If you are interested in this probability than you have to consider all rolls. If the game starts with a 5, you should mentally make a note that just no 3 7s have happened...
User avatar
Suoivax
Posts: 345
Joined: 13 April 2022, 00:27

Re: Dice generator - the non-conspiracy take

Post by Suoivax »

Coleus1 wrote: 26 June 2025, 09:47 Only on GBA can a number either get rolled way more often than others, or the other way around. Look at different championsships etc, play the game in real life. It only happens here. I played a game where only 11 was rolled, and one knew from the start that blue player would win. The other game, only two 9s were rolled, because orange (the player the game tries to favor and you knew would win from the start of them game) did not possess 9. If it happened once in a while that wouldn't be wierd, but it happens all the time on here.
Prove it.
List the table numbers so we can see.
Post Reply

Return to “Catan”