Page 1 of 4

Getting first player bias?

Posted: 01 January 2025, 20:38
by pogiepika
Is getting first player truly random or do premium account holders get an advantage? I know anecdotal but I just played 7 straight games of 2 player Azul without getting first player to open the game. Should be a 1/128 occurrence. Could be chance but starting to feel off.

Re: Getting first player bias?

Posted: 01 January 2025, 20:41
by GTSchemer
I haven't done extensive testing on this game, but I do notice that over 17.5 million games of Azul have been played on BGA since roughly the end of 2021.

With over 16,000 games per day on average, I would barely blink at a 1/128 chance cropping up.

Re: Getting first player bias?

Posted: 02 January 2025, 15:39
by pogiepika
GTSchemer wrote: 01 January 2025, 20:41 I would barely blink at a 1/128 chance cropping up.
It did get up to 9 in a row, so 1/512

Re: Getting first player bias?

Posted: 02 January 2025, 17:07
by RobertBr
pogiepika wrote: 02 January 2025, 15:39 It did get up to 9 in a row, so 1/512
Its 1/256. The idea that the probabilities would be secretly weighted to favour one group of players (no doubt by the Illuminati to cement the power of the military industrial complex) is so absurd on the face of it that even suggesting it puts you in the category of conspiracy theorist.

Your brain wants to find patterns and because you don't understand how randomness works (or how your brain works) you jump to an absurd conclusion. You are not the first, and you won't be the last, person to push silly conspiracy theories on the forums about random elements in the games.

Re: Getting first player bias?

Posted: 02 January 2025, 17:46
by JohnnyCyclops
Now… losing nine games in a row… that would be something to focus on! 😂

Re: Getting first player bias?

Posted: 02 January 2025, 18:51
by pogiepika
RobertBr wrote: 02 January 2025, 17:07
pogiepika wrote: 02 January 2025, 15:39 It did get up to 9 in a row, so 1/512
It’s 1/256. The idea that the probabilities would be secretly weighted to favour one group of players
No, it’s 1 out of 512. It’s 1/256 to get first OR second player 9 times in a row. You get the first free, then 8 more 50% probabilities. If at start, you declare you want one or the other, it’s 9 50% probabilities, or 1/512. At least get your math correct before you start lecturing folks.

Re: Getting first player bias?

Posted: 02 January 2025, 19:28
by h_illes
A 1/512 occurence is not rare. With 16000 games played a day, it should happens around 30 times a day. One of them was to you. It's not fun, but it's nothing extraordinary.

Re: Getting first player bias?

Posted: 02 January 2025, 23:48
by pogiepika
h_illes wrote: 02 January 2025, 19:28 A 1/512 occurence is not rare. With 16000 games played a day, it should happens around 30 times a day. One of them was to you. It's not fun, but it's nothing extraordinary.
I’d say 1/5 of 1% is rare, lol. It may happen regularly, not 30 times, but about 3-4 times a day. Don’t use 16000 games but 1778 groups of 9 games. Regardless yes it may happen, but it feels off when it happens to you.

Re: Getting first player bias?

Posted: 03 January 2025, 03:00
by Cile
pogiepika wrote: 02 January 2025, 23:48
h_illes wrote: 02 January 2025, 19:28 A 1/512 occurence is not rare. With 16000 games played a day, it should happens around 30 times a day. One of them was to you. It's not fun, but it's nothing extraordinary.
I’d say 1/5 of 1% is rare, lol. It may happen regularly, not 30 times, but about 3-4 times a day. Don’t use 16000 games but 1778 groups of 9 games. Regardless yes it may happen, but it feels off when it happens to you.
Sure, it's weird when it happens to you, but it's going to happen to somebody at least once a day. If every day's somebody made a post about it these forums would be nothing but people asking about bias in random systems.

Congratulations. You were today's somebody.

Re: Getting first player bias?

Posted: 03 January 2025, 08:03
by Mathew5000
pogiepika wrote: 02 January 2025, 23:48 Don’t use 16000 games but 1778 groups of 9 games.
That's not the correct way to calculate it, because the string of 9 consecutive "second-player" might overlap two of your groups.

If you play, let's say 20 games a day, and for each game getting chosen to play first or second are equally likely, and stochastically independent, then the probability of a string of 9 or more "seconds" within one day is about 0.0126, i.e. 1 in 79.2. So you'd expect it to happen to you a few times a year, if you play 20 games every day, and that's not even including the strings of 9 that overlap between one day and the next.