After 20 times playing, I am done with the BGA version

[Meeplelowda]

For a game with 100 dice rolls, the average number of sevens should be 17, but the standard deviation is 4, meaning that one in three games will include either 22 or more sevens, or 12 or less sevens. A game of 9 or less sevens rolled would occur once every 40 games - its actually pretty common.

In my experience (both on BGA and every other platform involving a game that involves any random outcome as part of the game), people who complain about these things are never, ever seeking to be or willing to be educated about probability and statistics.

[Meeplelowda]

Repost my suggestion (below) - PHP randomness script is not great - (https://boardgamearena.com/faq?anchor=f ... ing_random) Even the explainer they link in the FAQ explains how PHP is not random https://www.random.org/analysis/

I just played yahttzee and we had two first rolls yahtzee with 5 - 1 die rolls in one game and at the beginning of the next game the player that didn't get it in the first game got first time roll yatzhee same number. the probability of rolling yahtzee on the first roll is .077% (https://www.yahtzeemanifesto.com/yahtzee-odds.php) Two days ago I got 4 yahtzee's in one game - that's out of the 13 rolls you take in the game. Odds of getting 3 yahtzees is 1.7%+ of ALL games. You are likely to get a yahtzee 1 time in every 22 rolls.

The psuedo randomness of php is prone to issues
https://stackoverflow.com/questions/235 ... lly-random
When these errors occur they seems to cascade and true randomness suffers

I am suggesting they partner with roll 20 (https://app.roll20.net/home/quantum) or use a system like the random.org folks they cite

I had made this suggestion back in August 2020.
https://boardgamearena.com/forum/viewto ... l20#p65866

Given how they have grown since 2 years ago they should invest in a true randomness generator of some sort

Of course no pseudo random generator is "truly" random. But that's not the point. A well designed one isn't so bad that you can distinguish it from a truly random source just by a small sample size of observations. If I gave you 10 sequences of 100 random numbers, 9 of the sequences which were generated by a source of truly random noise, and one of which used PHP, you'd do no better than chance picking out which was pseudo random.

Another example: ask a lecture hall of 100 students to write down 20 digits between 0 and 9. A vast majority of them will produce a series with no repeats, whereas a truly random process will only produce about 13 sequences with no repeats. This is actually done in some Probability and Statistics courses to drive home how rubbish the human brain is at interpreting randomness.

We don't need cryptographically secure randomness. We're just rolling dice. I 100% guarantee you that people would still complain about things being rigged if they switched to random number generation by truly random physical processes.

[silentProtest]

Repost my suggestion (below) - PHP randomness script is not great - (https://boardgamearena.com/faq?anchor=f ... ing_random) Even the explainer they link in the FAQ explains how PHP is not random https://www.random.org/analysis/

I just played yahttzee and we had two first rolls yahtzee with 5 - 1 die rolls in one game and at the beginning of the next game the player that didn't get it in the first game got first time roll yatzhee same number. the probability of rolling yahtzee on the first roll is .077% (https://www.yahtzeemanifesto.com/yahtzee-odds.php) Two days ago I got 4 yahtzee's in one game - that's out of the 13 rolls you take in the game. Odds of getting 3 yahtzees is 1.7%+ of ALL games. You are likely to get a yahtzee 1 time in every 22 rolls.

The psuedo randomness of php is prone to issues
https://stackoverflow.com/questions/235 ... lly-random
When these errors occur they seems to cascade and true randomness suffers

I am suggesting they partner with roll 20 (https://app.roll20.net/home/quantum) or use a system like the random.org folks they cite

I had made this suggestion back in August 2020.
https://boardgamearena.com/forum/viewto ... l20#p65866

Given how they have grown since 2 years ago they should invest in a true randomness generator of some sort

second this !!

[silentProtest]

the guy is bad not the dice. he has low elo at every single game hes tried. main character syndrome / dunning kruger

Me too ??


I try to argue here: https://boardgamearena.com/bug?id=66082 that something seems off (not based on occurrence of some numbers [2-12], but based on clusters happening, of specific number, again and again per individual game)

[silentProtest]

Repost my suggestion (below) - PHP randomness script is not great - (https://boardgamearena.com/faq?anchor=f ... ing_random) Even the explainer they link in the FAQ explains how PHP is not random https://www.random.org/analysis/

I just played yahttzee and we had two first rolls yahtzee with 5 - 1 die rolls in one game and at the beginning of the next game the player that didn't get it in the first game got first time roll yatzhee same number. the probability of rolling yahtzee on the first roll is .077% (https://www.yahtzeemanifesto.com/yahtzee-odds.php) Two days ago I got 4 yahtzee's in one game - that's out of the 13 rolls you take in the game. Odds of getting 3 yahtzees is 1.7%+ of ALL games. You are likely to get a yahtzee 1 time in every 22 rolls.

The psuedo randomness of php is prone to issues
https://stackoverflow.com/questions/235 ... lly-random
When these errors occur they seems to cascade and true randomness suffers

I am suggesting they partner with roll 20 (https://app.roll20.net/home/quantum) or use a system like the random.org folks they cite

I had made this suggestion back in August 2020.
https://boardgamearena.com/forum/viewto ... l20#p65866

Given how they have grown since 2 years ago they should invest in a true randomness generator of some sort

Of course no pseudo random generator is "truly" random. But that's not the point. A well designed one isn't so bad that you can distinguish it from a truly random source just by a small sample size of observations. If I gave you 10 sequences of 100 random numbers, 9 of the sequences which were generated by a source of truly random noise, and one of which used PHP, you'd do no better than chance picking out which was pseudo random.

Another example: ask a lecture hall of 100 students to write down 20 digits between 0 and 9. A vast majority of them will produce a series with no repeats, whereas a truly random process will only produce about 13 sequences with no repeats. This is actually done in some Probability and Statistics courses to drive home how rubbish the human brain is at interpreting randomness.

We don't need cryptographically secure randomness. We're just rolling dice. I 100% guarantee you that people would still complain about things being rigged if they switched to random number generation by truly random physical processes.

you cant compare what humans would write down as random numbers to real random numbers^^

my specific point would be, that clusters of same couple numbers [2-12] seem to occurre, within one game {https://boardgamearena.com/bug?id=66082} - which , if true, would affect catan massivly (not much experience with other dice games here) ... but to proof that, or say that with enough certainty, would require a ton of game statistics, which BGA has not implemented to give in an simple format [in a CSV or something],.. short of scrapping tons of game histories, which is forbidden ....

[Jegpeg]

The PRNG used on sites like this will be far less biased than the dice that come with the boxed version of the game.
http://mathtourist.blogspot.com/2021/02/dice-bias.html
Clearly we should be complaining that on BGA rolling a 1 is just as common as rolling a 6, and therefore a 12 is as likely as a 2 but in physical game you are better off having a town next to a 12 resource than a 2 resource.

[Meeplelowda]

you cant compare what humans would write down as random numbers to real random numbers^^

Of course what humans would write down are not real random numbers. THAT'S THE WHOLE POINT! What people *perceive* to be random, or more importantly what they think is not random, is precisely what causes threads like this. In my example the vast majority of people perceive that repeating digits are not random, so they do not write down any repeating digits even when it is very likely to happen at least once. The human brain is a pattern seeking machine, and people come to believe one way or another that there can't be patterns in randomness, so if you find a pattern in something it can't be random. "Random means all mixed up" they think.

But I'm all for using a "better" source of randomness for BGA, just because it will prove my point that these threads will continue regardless.

[Stroom]

What Meeplelowda said. People who complain about randomness do not understand randomness.

[kjrevphoto]

Ya im super frustrated that the rolls are rigged. we can not have more than 7 or it WILL come up 7. kinda disgusting to think its not random

[Jellby]

Of course it's not random, otherwise people would complain that it's too random.