too many black pieces

@nadjarostowa said ^

This just happens from time to time. It is hard to avoid. As a simple example, consider that both players had Black in their two last games; one of them will have three blacks in a row now.

Same holds for White, but nobody complains...

thats kinda racist honestly dawg dont be racist or give ideas to it

There are 10 combinations of 3 or 4 of the same.

HHHH, HHHT, HHTH, HTHH, THHH, TTTT, TTTH, TTHT, THTT, HTTT

and there are 2^4 = 16 total cases which means the probability is 10/16 = 5/8 = 62.5%

P(atleast 3W or 3L or 4W or 4L) = 10/16
S = {WWWW, WWWL, WWLW, WLWW, LWWW, WWLL, WLLW, LLWW, WLWL, LWLW, LWWL, WLLL, LWLL, LLWL, LLLW, LLLL}

I assume that you ignored WWWW and LLLL, that does yield 50%...

I think there is a small difference between the two probabilities being discussed here.

For 4 games:

• Getting Black in at least 3 games: \(5/16 = 31.25\%\).
• Getting either Black or White at least 3 times: \(10/16 = 62.5\%\).

So, 62.5% is correct when we consider either colour appearing at least 3 times. If we specifically mean getting Black in 3 or 4 games, the probability is 31.25%, assuming each game's colour is independent and equally likely.

Getting the same colour repeatedly is normal in random matchmaking. However, over a large number of games, it is worth checking the overall colour distribution before concluding that the system is unfair. ️

@thibault said ^

3 in a row

ngl I misread the above comments at first so I got confused. I assumed "3 in a row" but the comments were mostly "3 out of 4".

For 3 in a row it's only 3/8 chance. But for at least 3 out of 4, it is indeed 5/8 chance (or 5/16 for just black). Finally for 3/4 games with black in a row, 3/16 chance.

HHHH 1a
HHHT 2b
HHTH c
HTHH d
HHTT
HTHT
HTTH
HTTT 3e

I get it that everyone loves flipping coins, and being very smart about counting color streaks...

Are you sure that the probabilities transfer 1:1 to the "random" pairings? Because I think you would need not only random players but also random color selection for that - and I always thought that this is not what happens.

Lichess random pairings are not actually random. They keep count of how many white/black games everyone plays, and tries to keep everyone at 50%.
Every time you play with one color, you increase your chances of getting the other color on next game.

This prevents the streaks that true random would naturally produce. People don't accept streaks can be random. So to make things feel more random, I had to make them less random. By counting and breaking streaks.

Spotify does the same trick in "random" playlists to avoid having 2 songs of the same album in a row, something people deemed to not be random enough.

Nobody wouldn't complain if they got White pieces 3 times in 4 games. This is just a psychological feeling because you don't like the outcomes.
Examples would be:
"This guy plays like a bot"
"Why am I BLACK AGAIN!!?!?!"
"I swear this player is cheating..."
"These puzzles are too hard"

Nothing is wrong with having Black multiple times, no need to complain and whine about it in the Feedback forum.

True actually I kinda like playing black

@Kitsune9Tails said ^

Nobody wouldn't complain if they got White pieces 3 times in 4 games.

I might complain, because Lichess' rating formula accounts for first move advantage and my openings are bad. I'm more likely to gain rating points moving second and lose rating points moving first.