What does Lichess 2000 puzzle rating signify? • page 2/2 • General Chess Discussion • lichess.org

It should also be mentioned that at a certain point the puzzles have a higher rating than they should have, in lichess there comes a point that the problems are quite complicated, but from my experience in chess dot com, the tactical problems have a rating much more inflated. I think lichess 2200 or less problems are actually quite simple, but on chess dot com you see easy puzzles even at 3200 rating despite having gone through many player attempts (obviously there will be many tricky ones). The difficulty starts to get intense at 3100 rating in chess dot com and in lichess at 2700, but the strange thing is that in chess dot com you see very difficult problems in the middle of the road (from 2400).

To tell the truth, the chess dot com puzzles have more of a psychological tinge than a specific difficulty (tempting variations that don't work, tricky details in some easy variants and positions with many atypical details) and those of lichess are specifically more demanding in understanding the position and specific calculation (if you make a mistake it is usually more due to the specific difficulty than to other factors).

In both lichess and chess dot com the impatience of high rated players in tactics causes it to inflate the rating of problems that are not really that complicated. An element to take into account to fail less problems is that it is necessary to count the total material of pieces and pawns, some combinations fail because one forgets to count the final material result of the combination or because a good variant that does not win pieces is underestimated. but it wins a relatively easy ending (if you don't take these things into account, you can fail puzzles that would be easily solved in a real game).

It should be mentioned that it is not necessary to solve the problems as fast as possible, you will always earn points no matter how long it takes, I think some people do not understand this. If you want to solve problems and have your slowness punish you by subtracting points, you can go to chess tempo and put it in blitz mode (the puzzle classifications are divided into three different ones and one of these is with unlimited time).

Each platform has its interesting things in puzzles and I don't prefer any one over the other (between lichess and chess dot com), I guess their differences are due to the tactical problem search engine that works differently. I haven't gone to chess tempo for a long time, but if I'm not mistaken it's much better for training tactics and I remember that the rating of the problems was more realistic.

@XaviCrow said It should also be mentioned that at a certain point the puzzles have a higher rating than they should have, in lichess there comes a point that the problems are quite complicated, but from my experience in chess dot com, the tactical problems have a rating much more inflated. I think lichess 2200 or less problems are actually quite simple, but on chess dot com you see easy puzzles even at 3200 rating despite having gone through many player attempts (obviously there will be many tricky ones). The difficulty starts to get intense at 3100 rating in chess dot com and in lichess at 2700, but the strange thing is that in chess dot com you see very difficult problems in the middle of the road (from 2400). To tell the truth, the chess dot com puzzles have more of a psychological tinge than a specific difficulty (tempting variations that don't work, tricky details in some easy variants and positions with many atypical details) and those of lichess are specifically more demanding in understanding the position and specific calculation (if you make a mistake it is usually more due to the specific difficulty than to other factors). In both lichess and chess dot com the impatience of high rated players in tactics causes it to inflate the rating of problems that are not really that complicated. An element to take into account to fail less problems is that it is necessary to count the total material of pieces and pawns, some combinations fail because one forgets to count the final material result of the combination or because a good variant that does not win pieces is underestimated. but it wins a relatively easy ending (if you don't take these things into account, you can fail puzzles that would be easily solved in a real game). It should be mentioned that it is not necessary to solve the problems as fast as possible, you will always earn points no matter how long it takes, I think some people do not understand this. If you want to solve problems and have your slowness punish you by subtracting points, you can go to chess tempo and put it in blitz mode (the puzzle classifications are divided into three different ones and one of these is with unlimited time). Each platform has its interesting things in puzzles and I don't prefer any one over the other (between lichess and chess dot com), I guess their differences are due to the tactical problem search engine that works differently. I haven't gone to chess tempo for a long time, but if I'm not mistaken it's much better for training tactics and I remember that the rating of the problems was more realistic.

Voy-aSer-ElMejor

#12

@check-sandwich_1-2-3

On chess dot com I got to 3250 just by solving the 3 free puzzles every 24 hours.

@check-sandwich_1-2-3 On chess dot com I got to 3250 just by solving the 3 free puzzles every 24 hours.

turtleleo77

#13

I'm completely awful at chess but I'm 2100 puzzles. Clearly doesnt mean much

v_u

#14

I am 1600 at puzzle but my rating is higher @turtleleo77 this whole puzzle thing is sus.

Bellendo

#15

There has to be a link between puzzle rating and other ratings, but I'm not sure what it is. Puzzles on this site are hard I think, I am a couple of hundred points higher on chess.com for puzzles, the opposite of other ratings which are lower there.

BlotterFan

#16

@Voy-aSer-ElMejor said in #2:

The tactic reflects your ability to generally calculate short variations and pattern recognition. Some puzzles do require calculating long variants, but sometimes those long variants go hand in hand with recognizing patterns that you practically don't compute in most sequences of each variant.

Variant -/- Variation are different. Variant is koth, 960 etc, variation depth is calculation depth

@Voy-aSer-ElMejor said in #2: > The tactic reflects your ability to generally calculate short variations and pattern recognition. Some puzzles do require calculating long variants, but sometimes those long variants go hand in hand with recognizing patterns that you practically don't compute in most sequences of each variant. > Variant -/- Variation are different. Variant is koth, 960 etc, variation depth is calculation depth

xXRegicideXx edited

#17

I would also say that if you glance at the puzzles dashboard, it will give you a performance rating that is more closely related to a Glicko-2 score. For instance, performing at a 2000+ puzzle rating will give you a performance rating in the 1800s or higher depending on your success-to-failure ratio in reaching that puzzle rating. The performance is a bit like a performance in arena tournaments: not a reflection of your actual Glicko-2 score, but definitely a hint at your peak tactical abilities