How can I improve my calculation? • page 2/2 • General Chess Discussion • lichess.org

Play over COMPLETE GAMES of yours & GMs ... Work with the Engine & Visualize it's moves ... STUDY PLAY PLAY STUDY Plan To Attack by Gary lane I looked at a bit & it helped with putting Pieces " In The Vicinity" Then yeah LOOK ... Allways pay attention to THE SHOTS Computers see >> So Lets say you click Watch Lichess TV after the game QUICKLY SEE THE SHOTS MISSED BY BOTH with the engine in Analysis . Also Problem Solving with THEMES can really help Cyrus Lakdawala has some new book out on that Cy

hendrixmaine

#12

Doing (hard) puzzles will make your calculation more accurate and a bit quicker. You can improve in both areas. Of course there is some tradeoff (and Aagard recommends NOT trying to calculate super fast for example so you don't run in circles) but it's possible to improve in both areas.

I think #6 kind of missed the point even if what he's saying is sort of correct.

Doing (hard) puzzles will make your calculation more accurate and a bit quicker. You can improve in both areas. Of course there is some tradeoff (and Aagard recommends NOT trying to calculate super fast for example so you don't run in circles) but it's possible to improve in both areas. I think #6 kind of missed the point even if what he's saying is sort of correct.

dnowmects edited

#13

For fast calculations, make sure you know all the rules of thumb; bishops are better than knights, any two knights/bishops are better than a rook, connected pawns are much better than isolated pawns, a defended passed pawn is worth close to a full piece. Great for vague positions with too many possibilities to calculate everything, though they'll bite you in sharper positions.

Start at the end and work backward. Find yourself a position you'd like to have, then figure out where your pieces need to be to get there, and how your opponent can stop them getting there, and how you can stop your opponent from stopping you, and so on back to the current position. Do it a whole lot and you'll start doing it automatically, which will make you really fast at finding good attacks and rejecting bad ones; get used to deflecting pieces and you'll be able to rip up "protected" positions. Puzzles are pretty much designed around this. Probably use the engine to start out with that (after the game obviously); doesn't help your calculations if you overlook opponent resources.

Don't worry about finding the "best" move, just find any winning move. Mate in Ten or Mate in Thirty doesn't matter much, unless you're low on time, and you're only low on time because you tried to find a Mate in Ten instead of immediately playing the Mate in Thirty.

Slumps are unavoidable, wait them out and don't worry. Stress makes you dumber. Or at least less happy.

For fast calculations, make sure you know all the rules of thumb; bishops are better than knights, any two knights/bishops are better than a rook, connected pawns are much better than isolated pawns, a defended passed pawn is worth close to a full piece. Great for vague positions with too many possibilities to calculate everything, though they'll bite you in sharper positions. Start at the end and work backward. Find yourself a position you'd like to have, then figure out where your pieces need to be to get there, and how your opponent can stop them getting there, and how you can stop your opponent from stopping you, and so on back to the current position. Do it a whole lot and you'll start doing it automatically, which will make you really fast at finding good attacks and rejecting bad ones; get used to deflecting pieces and you'll be able to rip up "protected" positions. Puzzles are pretty much designed around this. Probably use the engine to start out with that (after the game obviously); doesn't help your calculations if you overlook opponent resources. Don't worry about finding the "best" move, just find any winning move. Mate in Ten or Mate in Thirty doesn't matter much, unless you're low on time, and you're only low on time because you tried to find a Mate in Ten instead of immediately playing the Mate in Thirty. Slumps are unavoidable, wait them out and don't worry. Stress makes you dumber. Or at least less happy.

PawnCreeper

#14

I’ll teach you. Just need a buddy to work with. Chat me.

dvg11

#15

To improve calculations first do tough puzzles. Later try to practice some games , remember not to go to complicated just keep it simple. Also think of your opponents move as that's how GM play. In any position don't go far than 2 or 3 ideas it will make you lose very much time and you will become confuse. So just make it simple .

StingerPuzzles

#16

A chessable course just came out. It appears to be a good calculation course but it's not free.

Basically it teaches a method of evaluating positions with the idea that you're steered in the direction of calculating the right candidate moves based on features of the position that you're given a method to identify.

It's quite natural to make a list of the type of calculation errors you tend to repeat, but that knowledge doesn't seem to make calculation more accurate. That's why the course tries to teach a structured method.

I don't know why @Alientcp has the completely unrelated concepts of accuracy and length tied together in their mind. Further I don't know why their mistake has caused them to be unable to answer your question.

A chessable course just came out. It appears to be a good calculation course but it's not free. Basically it teaches a method of evaluating positions with the idea that you're steered in the direction of calculating the right candidate moves based on features of the position that you're given a method to identify. It's quite natural to make a list of the type of calculation errors you tend to repeat, but that knowledge doesn't seem to make calculation more accurate. That's why the course tries to teach a structured method. I don't know why @Alientcp has the completely unrelated concepts of accuracy and length tied together in their mind. Further I don't know why their mistake has caused them to be unable to answer your question.

Usia-Nous

#17

The most important and most successful thing (it took me, almost certainly, from 1200 to 1700 in blitz and from 1500 to 1900 in rapid) is what @hendrixmaine said.

I haven't done practically anything else in several months, and I practice tactics in books that I have at home and play against some neighbors who are 1800 and 1900 on chess.com, and it has eventually improved my calculation skills and my results.

A MISTAKE SOME PEOPLE MAKE:

They solve puzzles very quickly and using the green arrows (as masters do on YouTube). Using arrows is for didactic purposes (that the audience understands the thought process) but to improve, you must calculate without arrows, only by looking at the position.

All the best.

The most important and most successful thing (it took me, almost certainly, from 1200 to 1700 in blitz and from 1500 to 1900 in rapid) is what @hendrixmaine said. I haven't done practically anything else in several months, and I practice tactics in books that I have at home and play against some neighbors who are 1800 and 1900 on chess.com, and it has eventually improved my calculation skills and my results. A MISTAKE SOME PEOPLE MAKE: They solve puzzles very quickly and using the green arrows (as masters do on YouTube). Using arrows is for didactic purposes (that the audience understands the thought process) but to improve, you must calculate without arrows, only by looking at the position. All the best.

Alientcp

edited

#18

@RJblue said in #9:

@Alientcp I don't mean 2 seconds per calculation. Quick is a very general statement. I was using GM bullet games as an example of quick calculation, but quick could also pass as 5-10 seconds per move. Besides, I asked for advise, not whatever you're trying to say. And sorry for the late response, had to go.

I know what you mean, quick is relative, but regardless of that, i am giving you legitimate information, which is true, you cant have quick and accurate, because if you want to be accurate, it is self explanatory where the pieces may end up, so at quick glance you think you have all that covered, but every time a piece moves, the geometry of the landscape changes, so new moves that were not possible before, are possible now and you may miss them.

And also depth is important, because the deeper you search, the tree of choices start to get bigger, and thus, more variations you have to calculate, which by definition, will make the calculation slower. And even if you calculate them accurately until there, the next move, which you did not calculate might render a variation not viable, but you dont know, but because you didnt calculate that deep (check the last armaggedon between Carlsen and Mamedyarov https://www.youtube.com/watch?v=tZTDSlXEEcw&t=185s) .

So again, if you want quick, you cant calculate deep, which might give you an innacurate moves, since you cant really know the viability of the ending position.

If you want accuracy, you cant be quick, as you have to double or triple check every single variation on every single candidate move for missing moves, intermezzos, discoveries, possible new moves, etc. So 3 moves deep with 3 candidate moves might result calculating well over 20 different variations. The tree really gets big very fast as deeper as you go.

You can increase accuracy by doing puzzles with long sequences, like mate in 7. but you cant really have both.

If you want quick and as accurate as it gets, you cant calculate deep, 1-2 moves only and in 1 or 2 variations max, which is what GMS do in bullet or fast time controls, but as i said, once they are out of theory, the inaccuracies go to the roof, either because they didnt calculate deep enough, chose an non viable variation, out right miscalculated or all of the above.

Train to increase accuracy, but you really cant cut time.

@RJblue said in #9: > @Alientcp I don't mean 2 seconds per calculation. Quick is a very general statement. I was using GM bullet games as an example of quick calculation, but quick could also pass as 5-10 seconds per move. Besides, I asked for advise, not whatever you're trying to say. And sorry for the late response, had to go. I know what you mean, quick is relative, but regardless of that, i am giving you legitimate information, which is true, you cant have quick and accurate, because if you want to be accurate, it is self explanatory where the pieces may end up, so at quick glance you think you have all that covered, but every time a piece moves, the geometry of the landscape changes, so new moves that were not possible before, are possible now and you may miss them. And also depth is important, because the deeper you search, the tree of choices start to get bigger, and thus, more variations you have to calculate, which by definition, will make the calculation slower. And even if you calculate them accurately until there, the next move, which you did not calculate might render a variation not viable, but you dont know, but because you didnt calculate that deep (check the last armaggedon between Carlsen and Mamedyarov https://www.youtube.com/watch?v=tZTDSlXEEcw&t=185s) . So again, if you want quick, you cant calculate deep, which might give you an innacurate moves, since you cant really know the viability of the ending position. If you want accuracy, you cant be quick, as you have to double or triple check every single variation on every single candidate move for missing moves, intermezzos, discoveries, possible new moves, etc. So 3 moves deep with 3 candidate moves might result calculating well over 20 different variations. The tree really gets big very fast as deeper as you go. You can increase accuracy by doing puzzles with long sequences, like mate in 7. but you cant really have both. If you want quick and as accurate as it gets, you cant calculate deep, 1-2 moves only and in 1 or 2 variations max, which is what GMS do in bullet or fast time controls, but as i said, once they are out of theory, the inaccuracies go to the roof, either because they didnt calculate deep enough, chose an non viable variation, out right miscalculated or all of the above. Train to increase accuracy, but you really cant cut time.