Difference between puzzle rating and live chess

Sort:
Avatar of cenekp

Im 1300 rapid and 2200 puzzle

Avatar of brasileirosim
PillageAndBlunder wrote:
My puzzle rating is 1100 pts higher than my blitz and 900 more than Rapid.. and 400 more than Daily… so for me, in the longer time controls, I shrink the gap. Time pressure has everything to do with it for me. I find more tactical ideas in daily games because I can use the analysis board to play out positions.

In Rapid (with increment especially) I don’t often run into time trouble and can spend more time avoiding mistakes and taking advantage of tactical positions.

But in the shorter time controls, I spend too much time on mundane moves and get myself into time trouble. I miss a lot of opportunities this way and lose an enormous amount of games in winning positions because of it.

The more time I have, the closer I get to my puzzle rating (puzzles have no time limit)

This is the story of my life!

Avatar of wooksgambit69

Peak rapid rating was like 1280 and I'm 1800 at puzzles

Avatar of portgas_d_adam

hey guys, i'm currently 900 elo, 1900 peak problems, how can I claim my 1000 elo back pls?

Avatar of hucker233
haiaku wrote:
hucker233 wrote:

Why keep trying to make comparisons with their game ratings?   Chess.com puzzles are a game with a set of rules where you get points.  Because you have infinite time you can get ratings that hare VERY high, like 30,000, which obviously have no relationship to chess ratings.  Those people play the game to win the puzzle game, they (likely) play each puzzle on paper until they get it and then they make the moves and get their 5 points, even if it takes 30 minutes or more. 

 I'm not sure what it means to "beat" someone in puzzles.  Your "score" is not a rating on how you would do against someone else, it indicates how you play one version of tactical puzzles on chess.com.  If you want speed, play the puzzle battle.  Those players are amazing and makes me understand just how terrible at chess I am.

As I said, I agree with you, but these puzzle ratings are not casual numbers. They are not, for example, 1.37, or 10953, or 3567852359.78986; they are 1256, 1789, 2477 or 2862. In general they resemble game ratings and that's why it can be misleading. So if your puzzle rating is 200 lower than a FM, you could (wrongly) think that in a sharp tactical game you have 20-25% chances to beat them, while in fact that is usually far from true.

Actually this is not correct.  Check out the top puzzle ratings. The top ratings are 65,000, 50,000 and 50,000.  A puzzle rating of 2000 means you are terrible compared to people who are world class at them (assuming they aren't cheating).  They only resemble "real" ratings if you ignore most of the data.   Just don't compare them beyond higher scores are better.

 

Avatar of technical_knockout

i got my 3600 puzzles pb legitimately by scoring 67% correct on 11,000+ puzzles over hundreds of hours of solving time... i never once put 'pencil to paper' & my average game stat rating is probably about 1700 across the board.

Avatar of Hilmi45
brasileirosim wrote:
Hilmi45 wrote:

2900 puzzles 
1200 blitz

2900 is quite high. I would expect this for a 1500 rated player. Perhaps your knowledge on openings is underdeveloped?

thats exactly im lacking at, also im down 2 2700 again.. so 

Avatar of haiaku
hucker233 wrote:
haiaku wrote:

As I said, I agree with you, but these puzzle ratings are not casual numbers. They are not, for example, 1.37, or 10953, or 3567852359.78986; they are 1256, 1789, 2477 or 2862. In general they resemble game ratings and that's why it can be misleading. So if your puzzle rating is 200 lower than a FM, you could (wrongly) think that in a sharp tactical game you have 20-25% chances to beat them, while in fact that is usually far from true.

Actually this is not correct.  Check out the top puzzle ratings. The top ratings are 65,000, 50,000 and 50,000.  A puzzle rating of 2000 means you are terrible compared to people who are world class at them (assuming they aren't cheating).  They only resemble "real" ratings if you ignore most of the data.   Just don't compare them beyond higher scores are better.

 

Assuming they aren't cheating. Considering these leaders' solving time, it seems they don't spend hours on every puzzle, so I don't know how they actually do. FMs usually do not reach those levels, they are just above 3000 on average.

Avatar of Sai7179

Bullet peak 2184, puzzles around 3000.

Avatar of StatoKoule

Nah im like 1100 but Puzzles 2200

 

Avatar of KingKilgy7

Im about 300 bullet and 2200 puzzles...

Avatar of Mathieu9229
KingKilgy7 a écrit :

Im about 300 bullet and 2200 puzzles...

That is a huge gap... Now I just had à quick look at one if your loss (last rapid game at this point) and I think you should learn about opening principles (control, centre, develop knights and bishops first, king safety...) because you develop your queen too early when your king is not safe and you lose (may be you can play the Italian or something). But if you can have 2200 in problems you should reach 1000 in games  easily.

Avatar of molechess

Many people are also cheating in puzzles and it's no way to prevent it. Some also specialize on tactics or even are solving so much that they have many problems memorized. In general I think the difference is about 1000 points between puzzle rating and playing ratings.

Avatar of portgas_d_adam
marekihnat a écrit :

Many people are also cheating in puzzles and it's no way to prevent it. Some also specialize on tactics or even are solving so much that they have many problems memorized. In general I think the difference is about 1000 points between puzzle rating and playing ratings.

 

cheating is irrelevant here cause if we don't assume that nobody is cheating this topic serve no purpose, though I am convinced what you said about memorizing patterns is true, I take myself as a witness of that

Avatar of haiaku
portgas_d_adam wrote:

cheating is irrelevant here cause if we don't assume that nobody is cheating this topic serve no purpose, though I am convinced what you said about memorizing patterns is true, I take myself as a witness of that

We can simply consider common gaps of about 1000 points, as @marekihnat says, and ignore extremes, but that doesn't solve the main problem: you can stay on a puzzle a few seconds or an eternity and that makes a difference; one can have the same rating of another player, who is in fact much better at tactics, because he solves puzzles more quickly. Time constraints are necessary to give a measure of a player's tactical skill during a game; only then we could compare that tactical rating with the actual game rating. So these puzzles are useful to train our tactical skill, but not to measure it imho.

Avatar of brasileirosim
haiaku wrote:
hucker233 wrote:
haiaku wrote:

As I said, I agree with you, but these puzzle ratings are not casual numbers. They are not, for example, 1.37, or 10953, or 3567852359.78986; they are 1256, 1789, 2477 or 2862. In general they resemble game ratings and that's why it can be misleading. So if your puzzle rating is 200 lower than a FM, you could (wrongly) think that in a sharp tactical game you have 20-25% chances to beat them, while in fact that is usually far from true.

Actually this is not correct.  Check out the top puzzle ratings. The top ratings are 65,000, 50,000 and 50,000.  A puzzle rating of 2000 means you are terrible compared to people who are world class at them (assuming they aren't cheating).  They only resemble "real" ratings if you ignore most of the data.   Just don't compare them beyond higher scores are better.

 

Assuming they aren't cheating. Considering these leaders' solving time, it seems they don't spend hours on every puzzle, so I don't know how they actually do. FMs usually do not reach those levels, they are just above 3000 on average.

AFAIK memorization plays a major role in such cases. Doing puzzles so many times that they begin to see the puzzles again. People who take really seriously will even make courses in Chessable (private) to drill difficult puzzles. At least I know once guy doing this. I memorized something like 3000 puzzles in Chessable, but from Chessable courses. My goal is 5000, then I will stop learning new puzzles and focus on middlegame, opening and endgame.

Avatar of StatoKoule

Im now 1100 in blitz rapid and bullet

but 2400 in puzzles

Avatar of brasileirosim
StatoKoule wrote:

Im now 1100 in blitz rapid and bullet

but 2400 in puzzles

Strong in puzzles, in rapid you should get soon 1300. Are you reviewing your games?

Avatar of Dotty_GM

804 rapid and 1611 puzzle. Any tips on how I can improve my rating?

Avatar of brasileirosim
DOTTY12 wrote:

804 rapid and 1611 puzzle. Any tips on how I can improve my rating?

Go to Chessable and learn following books:

Common Chess Patterns (tactics, very good course)

Basic Endgames (for free, excellent book)

After you learned well these books (you have to practice every day until all positions are marked as “mature”), you can go through Smithy’s Opening Fundamentals. 
After you learned well these 3 books you will be at least 200 points stronger. Good luck!