Evaluating weaknesses and customized training with ChessVis and Chesstempo

Reading time: 3 min

Yesterday it was time again to search fpr weaknesses in my understanding of tactics. Here is what I did.

I choose a puzzle level according to my rating which is about 1.800 in rapid on lichess. I use the app ChessVis, which uses the lichess puzzles as source and You can customize two important factors. Time and level.

I can adjust the level between 1.700 and 1.799 and the solving time with a maximum of one minute. This feature is for subscribers but can set your own timer in the free version.

Then I do 100 puzzles to evaluate my strengths and weaknesses in every category. 100 is the minimun of puzzles to have good tendency and a solid sample size. It still consumes much time, so that I need 2-3 Blocks of 25 minutes to solve all hundred puzzles. Exact time is measured for track my overall progress.

I used to solve 109 puzzles in 4.913 seconds with a succes rate of 60% about 2 months ago. Unfortunately I did not write down my weaknesses on this test but I did today. Learing and improving everywhere.

For 104 puzzles I needed 3.407 seconds and an unchanged success rate of 60%. So, I am 30% faster for the same result. Approximately 34 seconds for one puzzle with 60% success means that I would theoretically need 58 seconds with perfect solving rate and that means that I could save all puzzles in the frame of one minutes and in 2 1/3 Blocks of 25' minutes.

But more important are the weaknesses from which I create my customized training via chesstempo. Yes chesstempo is a bit clunky but very powerful when we know how to use it.

As we can see, forks, endgame and capitalization of an advantage are my momentary weaknesses. For the fork we only have a little sample size but I know from further evaluations that this is my weakness. I still surprises me as I consider a fork to be an easy tactic. Arrogance does not pay well! Forks in combinations and different kinds of forks are the key. Unfortunately the kind of forks can not be analyzed in ChessVis or on lichess but it does not matter really much. What will do is to set the type of puzzle to fork, play hundres of these puzzles and therefore split up the weaknesses as well. But for now this overview is enough.

Also, when setting up a custom set of puzzles on chesstempo I am very aware of my strenghts and that I leave them out in my custom set. I do not want to train mates, puzzles with more than 10 pieces (=endgame) and puzzles under three moves.

With this settings I have a set with 80 puzzles which I can loop. I do the second loop with descending order from at least solved to best solved so that my worst solved puzzles are directly focused. This loop is repeated until I have at least 85% success rate as mentioned on my previous post.

The good thing that chesstempo tracks all important data automatically so that this saves time.

Thank You for reading.