Year 2020 I was 800-900 blitz. If I now make a Game Analysis on those old games, the analysis confirms that today's ratings for those games are 550-750 for both players. The games at 800 level are way tougher today.
Chess.com ratings have massively deflated since 2020
Sort:
JuhoNieminen wrote:
Year 2020 I was 800-900 blitz. If I now make a Game Analysis on those old games, the analysis confirms that today's ratings for those games are 550-750 for both players.
It "confirms" nothing. Those estimates are wildly inaccurate and are usually all over the place. You can call that feature "experimental", but I prefer to call it "trash".
Apr 28, 2025
0
#104
It does not matter if the analysis is trash or not. That trash analysis shows that games of 2021 were less accurate that games of 2025, when the rating of the player pool is the same. The trash analysis is just a measuring stick.
May 12, 2025
0
#105
Blitz Rating Deflation:
May 3, 2025: 98.6 percentile = 1836 rating.
May 12, 2025: 98.6 percentile= 1722 rating.
May 12, 2025
0
#106
This is just simply wrong. The number of players at a given rating increase as the player population increases but the portion of players at different ratings (should) stay around the same.
May 12, 2025
0
#107
98.6 Percentile was the percentile for the following ratings on these dates:
1850: 8/11/2024
1843: 2/4/2025
1836: 5/3/2025
1722: 5/12/2025
It was drifting downward very slowly, but on May 9th there was an abrupt change. On my first game of that day, I noticed that my rating went down from 1771 to 1764, but my percentile went up from 98.5 to 99.1, and also, my global rank went from 142,043 to 80,794. (the implied total population remained steady at about 8.9 million.)
JimLake1 wrote:
This is just simply wrong. The number of players at a given rating increase as the player population increases but the portion of players at different ratings (should) stay around the same.
No statistically it has decreased overtime the average used to be 800 its now like 500 or 600
May 12, 2025
0
#109
GrunfeldSlam wrote:
98.6 Percentile was the percentile for the following ratings on these dates:
1850: 8/11/2024
1843: 2/4/2025
1836: 5/3/2025
1722: 5/12/2025
It was drifting downward very slowly, but on May 9th there was an abrupt change. On my first game of that day, I noticed that my rating went down from 1771 to 1764, but my percentile went up from 98.5 to 99.1, and also, my global rank went from 142,043 to 80,794. (the implied total population remained steady at about 8.9 million.)
This makes sense but why am I 98.6% percentile at 1601 rapid rating? Thats way lower than the 1722 you are saying. Is it because a lot of cheaters got banned or what? I gained like +500 rating in the last month so I have doubts there is rating deflation ongoing. Might just be stats error or something else entirely.
May 13, 2025
0
#110
@ Techwiz911: That's a great question. The Blitz and Rapid ratings mostly don't line up with each other. This is due to the Blitz rating having a greater standard deviation (fluctuation range) than Rapid. Blitz are shorter games with more mistakes, due to the shorter time of the game. Thus, it is a greater probability for a blitz game to not truly reflect your skill level (especially if you're on "tilt"!). Notice the high end of the Blitz range (Nakamura/Carlsen) is hundreds of points higher than the high end of the Rapid Range.
If Chess.com wanted to attempt to make Blitz and Rapid ratings to line up with each other, (which would be difficult at this point), they would need to reduce the number of points lost/gained per game in Blitz compared to Rapid. (right now it's the same). I can't say I know exactly how much it would need to be reduced. It's difficult to say, since within Blitz there is still a significant range of time controls, with and without increments.
On your last point - it is true that this last sharp fluctuation of May 9th is due to some kind of bug in the system (from what I read in a different forum). But it is true that gradual deflation is happening over time, from the data I have examined prior to May 9th.
May 13, 2025
0
#111
KmBoor wrote:
GrunfeldSlam wrote:
@ Techwiz911: That's a great question. The Blitz and Rapid ratings mostly don't line up with each other. This is due to the Blitz rating having a greater standard deviation (fluctuation range) than Rapid. Blitz are shorter games with more mistakes, due to the shorter time of the game. Thus, it is a greater probability for a blitz game to not truly reflect your skill level (especially if you're on "tilt"!). Notice the high end of the Blitz range (Nakamura/Carlsen) is hundreds of points higher than the high end of the Rapid Range.
If Chess.com wanted to attempt to make Blitz and Rapid ratings to line up with each other, (which would be difficult at this point), they would need to reduce the number of points lost/gained per game in Blitz compared to Rapid. (right now it's the same). I can't say I know exactly how much it would need to be reduced. It's difficult to say, since within Blitz there is still a significant range of time controls, with and without increments.
On your last point - it is true that this last sharp fluctuation of May 9th is due to some kind of bug in the system (from what I read in a different forum). But it is true that gradual deflation is happening over time, from the data I have examined prior to May 9th.
People at higher levels don’t play rapid nearly as often as blitz, that’s generally why the ratings are lower for rapid. Especially titled players, look at how many rapid verse blitz they played in the last 90 days, tons don’t even play rapid at all.
Or maybe they simply have an alt for playing rapid, and don't tell that to everybody.
KmBoor wrote:
ChessAce1111 wrote:
KmBoor wrote:
GrunfeldSlam wrote:
@ Techwiz911: That's a great question. The Blitz and Rapid ratings mostly don't line up with each other. This is due to the Blitz rating having a greater standard deviation (fluctuation range) than Rapid. Blitz are shorter games with more mistakes, due to the shorter time of the game. Thus, it is a greater probability for a blitz game to not truly reflect your skill level (especially if you're on "tilt"!). Notice the high end of the Blitz range (Nakamura/Carlsen) is hundreds of points higher than the high end of the Rapid Range.
If Chess.com wanted to attempt to make Blitz and Rapid ratings to line up with each other, (which would be difficult at this point), they would need to reduce the number of points lost/gained per game in Blitz compared to Rapid. (right now it's the same). I can't say I know exactly how much it would need to be reduced. It's difficult to say, since within Blitz there is still a significant range of time controls, with and without increments.
On your last point - it is true that this last sharp fluctuation of May 9th is due to some kind of bug in the system (from what I read in a different forum). But it is true that gradual deflation is happening over time, from the data I have examined prior to May 9th.
People at higher levels don’t play rapid nearly as often as blitz, that’s generally why the ratings are lower for rapid. Especially titled players, look at how many rapid verse blitz they played in the last 90 days, tons don’t even play rapid at all.
Or maybe they simply have an alt for playing rapid, and don't tell that to everybody.
I guess that could be an option, but I think the majority of people after a certain level prefer blitz/bullet for the fast dopamine hit and less chance of cheating, along with having a larger pool of higher ratings and titled players in the queue.
It's probably the cheating plus the stalling annoys me too much to play rapid frequently
@KmBoor do you have an old alt your accounts new (not judging just curious )
KmBoor wrote:
BigChessplayer665 wrote:
KmBoor wrote:
ChessAce1111 wrote:
KmBoor wrote:
GrunfeldSlam wrote:
@ Techwiz911: That's a great question. The Blitz and Rapid ratings mostly don't line up with each other. This is due to the Blitz rating having a greater standard deviation (fluctuation range) than Rapid. Blitz are shorter games with more mistakes, due to the shorter time of the game. Thus, it is a greater probability for a blitz game to not truly reflect your skill level (especially if you're on "tilt"!). Notice the high end of the Blitz range (Nakamura/Carlsen) is hundreds of points higher than the high end of the Rapid Range.
If Chess.com wanted to attempt to make Blitz and Rapid ratings to line up with each other, (which would be difficult at this point), they would need to reduce the number of points lost/gained per game in Blitz compared to Rapid. (right now it's the same). I can't say I know exactly how much it would need to be reduced. It's difficult to say, since within Blitz there is still a significant range of time controls, with and without increments.
On your last point - it is true that this last sharp fluctuation of May 9th is due to some kind of bug in the system (from what I read in a different forum). But it is true that gradual deflation is happening over time, from the data I have examined prior to May 9th.
People at higher levels don’t play rapid nearly as often as blitz, that’s generally why the ratings are lower for rapid. Especially titled players, look at how many rapid verse blitz they played in the last 90 days, tons don’t even play rapid at all.
Or maybe they simply have an alt for playing rapid, and don't tell that to everybody.
I guess that could be an option, but I think the majority of people after a certain level prefer blitz/bullet for the fast dopamine hit and less chance of cheating, along with having a larger pool of higher ratings and titled players in the queue.
It's probably the cheating plus the stalling annoys me too much to play rapid frequently
It’s def a time thing for me, I’m not on here to improve my game, speed chess is terrible for that, I’m on here for fun. I use other time to study and improve, so rapid takes too much time away from being able to do both.
Actually speedchess is good for endgames and openings and endgames tend to be what people struggle with the most but speed chess is really only good for improving certain things middle game strategies it's not really that great for
May 13, 2025
0
#115
GrunfeldSlam wrote:
@ Techwiz911: That's a great question. The Blitz and Rapid ratings mostly don't line up with each other. This is due to the Blitz rating having a greater standard deviation (fluctuation range) than Rapid. Blitz are shorter games with more mistakes, due to the shorter time of the game. Thus, it is a greater probability for a blitz game to not truly reflect your skill level (especially if you're on "tilt"!). Notice the high end of the Blitz range (Nakamura/Carlsen) is hundreds of points higher than the high end of the Rapid Range.
If Chess.com wanted to attempt to make Blitz and Rapid ratings to line up with each other, (which would be difficult at this point), they would need to reduce the number of points lost/gained per game in Blitz compared to Rapid. (right now it's the same). I can't say I know exactly how much it would need to be reduced. It's difficult to say, since within Blitz there is still a significant range of time controls, with and without increments.
On your last point - it is true that this last sharp fluctuation of May 9th is due to some kind of bug in the system (from what I read in a different forum). But it is true that gradual deflation is happening over time, from the data I have examined prior to May 9th.
Thanks, that explains it. I am a slow and tactical thinker so my rapid rating is like +600 from my blitz rating lol. Its unfortunate a lot of strong players don't play rapid, for me its the number 1 way to get better at chess as beginner.
techwiz911 wrote:
GrunfeldSlam wrote:
@ Techwiz911: That's a great question. The Blitz and Rapid ratings mostly don't line up with each other. This is due to the Blitz rating having a greater standard deviation (fluctuation range) than Rapid. Blitz are shorter games with more mistakes, due to the shorter time of the game. Thus, it is a greater probability for a blitz game to not truly reflect your skill level (especially if you're on "tilt"!). Notice the high end of the Blitz range (Nakamura/Carlsen) is hundreds of points higher than the high end of the Rapid Range.
If Chess.com wanted to attempt to make Blitz and Rapid ratings to line up with each other, (which would be difficult at this point), they would need to reduce the number of points lost/gained per game in Blitz compared to Rapid. (right now it's the same). I can't say I know exactly how much it would need to be reduced. It's difficult to say, since within Blitz there is still a significant range of time controls, with and without increments.
On your last point - it is true that this last sharp fluctuation of May 9th is due to some kind of bug in the system (from what I read in a different forum). But it is true that gradual deflation is happening over time, from the data I have examined prior to May 9th.
Thanks, that explains it. I am a slow and tactical thinker so my rapid rating is like +600 from my blitz rating lol. Its unfortunate a lot of strong players don't play rapid, for me its the number 1 way to get better at chess as beginner.
Rapid is typically better for beginners but once your better it's whatever stronger opponents you can find so it can switch between blitz and rapid
Dec 20, 2025
0
#117
I'm not exactly sure about any of the points that everyone is arguing, but I think I may have something that could be considered proof for somebody's argument.
I joined chess.com about 7 years ago and never really played more than a couple games every few days or so and I am not very good. I stopped playing after a few months. I joined again under a new account in 2025 and this time around I got into it a lot more and have been playing quite a bit so I believe I am a better player and my rating is at about 700 right now. So then just for fun I looked up my old account to see what my rating was back then and to my surprise I was rated above 1000. Under this account I played some players in that rating range and am about 50/50 for wins and losses. I have beat multiple players over 1000. But I am having trouble getting past 700 on my other account. Any thoughts???
People are playing better chess now than pre-pandemic days (2019 and earlier)
I think there was an inflation. I don't study chess since 2017 and just casually playing my top rating went from 18xx to 20xx. Also the top rated players from the server weren't so high as 32xx some years ago.
Actually more players joining a pool tends to cause deflation at lower levels the average now is around 600 something but it used to be higher (at least deflation accures mostly at lower levels )