Game ratings are tied to elo of 2 players. You earn points that are removed from the other player. It is an interaction with trade balance. Puzzle ratings are not tied to a point balance, so they can inflate. And they have inflated. The highest player ratings in blitz are 3200-3300. Highest puzzle ratings are 28.000-77.000 or something like that.
Difference between puzzle rating and live chess
Sort:
This is interesting. My rapid rating is currently around 1480, blitz around 900, and my puzzle rating is 2550.
JuhoNieminen wrote:
Game ratings are tied to elo of 2 players. You earn points that are removed from the other player. It is an interaction with trade balance. Puzzle ratings are not tied to a point balance, so they can inflate. And they have inflated. The highest player ratings in blitz are 3200-3300. Highest puzzle ratings are 28.000-77.000 or something like that.
I have chatted with a couple people on the top of the puzzle rankings. interestingly they tend not to play games at all and only do puzzles.
Puzzle rating is a bit inflated at this site and live ratings are actually lower than other sites.
Feb 20, 2023
0
#326
checkmate_kate wrote:
JuhoNieminen wrote:
Game ratings are tied to elo of 2 players. You earn points that are removed from the other player. It is an interaction with trade balance. Puzzle ratings are not tied to a point balance, so they can inflate. And they have inflated. The highest player ratings in blitz are 3200-3300. Highest puzzle ratings are 28.000-77.000 or something like that.
I have chatted with a couple people on the top of the puzzle rankings. interestingly they tend not to play games at all and only do puzzles.
There might be some puzzle specialists but I think there are also lots of cheaters at the top of the puzzle leaderboard, as it is impossible to know whether they are using engine as only the precise moves are required in puzzles.
checkmate_kate wrote:
JuhoNieminen wrote:
Game ratings are tied to elo of 2 players. You earn points that are removed from the other player. It is an interaction with trade balance. Puzzle ratings are not tied to a point balance, so they can inflate. And they have inflated. The highest player ratings in blitz are 3200-3300. Highest puzzle ratings are 28.000-77.000 or something like that.
I have chatted with a couple people on the top of the puzzle rankings. interestingly they tend not to play games at all and only do puzzles.
I understand that they don’t play - they are completely focused on the puzzles. It wouldn’t be interesting to chat with these guys to hear how they get such a huge rating.
marekihnat wrote:
checkmate_kate wrote:
JuhoNieminen wrote:
Game ratings are tied to elo of 2 players. You earn points that are removed from the other player. It is an interaction with trade balance. Puzzle ratings are not tied to a point balance, so they can inflate. And they have inflated. The highest player ratings in blitz are 3200-3300. Highest puzzle ratings are 28.000-77.000 or something like that.
I have chatted with a couple people on the top of the puzzle rankings. interestingly they tend not to play games at all and only do puzzles.
There might be some puzzle specialists but I think there are also lots of cheaters at the top of the puzzle leaderboard, as it is impossible to know whether they are using engine as only the precise moves are required in puzzles.
i guess it's possible, but i really don't understand why someone would spend hours of every day acting as a fleshy robotic arm, inputing one computer's output into another computer - why would that be enjoyable?
Feb 23, 2023
0
#329
checkmate_kate wrote:
marekihnat wrote:
checkmate_kate wrote:
JuhoNieminen wrote:
Game ratings are tied to elo of 2 players. You earn points that are removed from the other player. It is an interaction with trade balance. Puzzle ratings are not tied to a point balance, so they can inflate. And they have inflated. The highest player ratings in blitz are 3200-3300. Highest puzzle ratings are 28.000-77.000 or something like that.
I have chatted with a couple people on the top of the puzzle rankings. interestingly they tend not to play games at all and only do puzzles.
There might be some puzzle specialists but I think there are also lots of cheaters at the top of the puzzle leaderboard, as it is impossible to know whether they are using engine as only the precise moves are required in puzzles.
i guess it's possible, but i really don't understand why someone would spend hours of every day acting as a fleshy robotic arm, inputing one computer's output into another computer - why would that be enjoyable?
Hard to tell, hopefully there are not lot of those.
Still going?
lol
this is so old
I think i just need to do more theory studying. I have a bullet rating of 800 and a puzzle rating of 1700.
1000 blitx 1300 rapid, 2700 puzzles
Caesar49bc wrote:
Some people have puzzle ratings several hundred points higher. It mostly boils down to the fact that a puzzle is a static position, in which the solver knows there is a BEST MOVE.
Games are pretty organic, since after every move, the board position changes. (Obviously there are scenarios of dead draws in which is matters little if the board position changes, or draws by 3 fold repetition.)
So it boils down to the fact that humans, in a timed game, don't have the luxury of looking for the best most every turn. Even Stockfish reportedly had to spend 5 to 8 hours looking at a single position to find the move that Google's Alpha Zero Chess played against it, and most likely, it was a newer version of Stockfish that looked at the position.
This, and it seems that a lot of chess puzzles on chess.com involve taking advantage of your opponent's blunder/mistake. If you look at the analysis of the puzzles, the advantage bar is usually close to even, but when you play the opponent's move that creates your starting position, the advantage bar usually shoots over in your favor because of the tactic their move allows. For this reason, it seems to me that the greatest benefit of doing puzzles is that it helps you learn to spot your opponent's mistakes and the tactics they allow. But if your opponent doesn't make a major blunder, it can be difficult to find a "best move".
I have 2181 puzzle but only 1028 rapid. I don't play too much, but would my actual rapid rating be higher if I played enough to show a more accurate rating?
Jul 7, 2023
0
#335
My puzzle rating is 2200 and blitz is 800 
lost 160 points in blitz in a week....
Ultimate-trashtalker wrote:
My puzzle is 2500 and my live is 2100
my rapid is 1400 and my puzzle is 2800
My rapid is in the 1200's (which is what I like to play) but my puzzle rating is in the 3100's. Usually, I only get +5 each time I solve a puzzle because it takes me a long time to see the sequence of moves.
Wait I have puzzle rating 1980 and I a 940 elo in rapid. That's a 1k difference
2100+ in puzzles, 1000+ in Rabid (sic) .
I also usually fair better taking extra time and am happy if I get it the +5. I try to find the answer asap but my aging brain needs timeto prove the certainty of it.
I'm currently 1360 rapid and 2230 puzzles. I'm not sure which aspect of my games needs work in order to close this gap.