matching algorithm
Why would your accuracy be dependent on who you are matched against?
@albacored then what does it mean that literally everyone is having long loss/win streaks?
@albacored then what does it mean that literally everyone is having long loss/win streaks?
Not everyone does. However, when someone is having a bad day, they may lose more than they win and that can happen in streaks. It happens in OTB play as well.
the more you win, the higher your rating, the stronger your opponents, the more you lose, the more you tilt, the lower your rating...
@martin_stahl.. literally look at everyone on this thread... it's always " streak of green vs streak of read" , even yours..this whole thing doesn't make any sense...
@martin_stahl.. literally look at everyone on this thread... it's always " streak of green vs streak of read" , even yours..this whole thing doesn't make any sense...
Yeah, it does. Streaks are not uncommon and are part of playing chess. I can have days at the chess club I can't win games and some where I win most. And that's against the same people most of the time. 
Plus, almost all my games here are in events, so the matchmaking algorithm doesn't really come into play.
Bigtreeb you're literally just making stuff up. Your elo is in the dumpster I think it's you that is the problem. Quit trying to find conspiracy theories on why the algorithm treats you unfairly.
I have also noticed that I get paired several times in a row with people who play the same opening, but that can easily be statistical fluke (or maybe they made a pairing algorithm in such a way that players get to play people who play the same opening several times in a row, if it is more addictive that way, then maybe the algorithm is designed that way).
On lichess I once got paired against the same player 7 times in a row. No, it was not a rematch (and I can prove it, I can show you the matches and you will see that there were no rematches offered). I don't know what was going on there.
the more you win, the higher your rating, the stronger your opponents, the more you lose, the more you tilt, the lower your rating... - no, here it's not the case quite often.
Within a few years, I noticed a similar thing: a winning raw of 5-6 games, then it sends you to someone who would be +40 or 50 higher rated. Then you start to lose.
If you are unfortunate and would lose 3-4 times in a role; then the algorithm will send you to those who are 50-60 points lower rated (but somehow, it is just not the case that you are stronger, and when losing, you lose more than by gaining from the game).
Also, at the beginning of the month, when they reset servers rating drops to 8-10% , starting from mid-month and closer to the end, and you regain back that 8 - 10 % score.
Why?
I'm sorry to say, but the matching algorithm here sucks.
the more you win, the higher your rating, the stronger your opponents, the more you lose, the more you tilt, the lower your rating... - no, here it's not the case quite often.
Within a few years, I noticed a similar thing: a winning raw of 5-6 games, then it sends you to someone who would be +40 or 50 higher rated. Then you start to lose.
If you are unfortunate and would lose 3-4 times in a role; then the algorithm will send you to those who are 50-60 points lower rated (but somehow, it is just not the case that you are stronger, and when losing, you lose more than by gaining from the game).
Also, at the beginning of the month, when they reset servers rating drops to 8-10% , starting from mid-month and closer to the end, and you regain back that 8 - 10 % score.
Why?
I'm sorry to say, but the matching algorithm here sucks.
There's nothing going on other than normal fluctuations. The site also does not reset the servers at the beginning of the month; they do restart Live occasionally for code updates, like last night, but it's not doing anything crazy with the ratings (unless there ends up being a bug in the code).
