Visualizing OTB combinations in the middle game

I have a hard time visualizing combinations while playing OTB chess past a few moves. Generally I can see 1-2 moves ahead reasonably well in my opinion, 3-4 gets hazy, and 5 moves are generally beyond what I can see. This causes me to miss many combinations (at times forced). This is worst in the middle game. During the endgame, it's generally a lot easier to see 4+ moves in. My middle game generally starts 4-5 moves in (since I don't know much more theory), and at this point it is the hardest. Does anyone know any strategies, or mental training that could help you achieve this.

Tactics training has not helped me much as I have been 1450-1500 on chess.emrald.net for over a year now. Any help would be appreciated. Thanks.

Originally posted by amolv06
I have a hard time visualizing combinations while playing OTB chess past a few moves. Generally I can see 1-2 moves ahead reasonably well in my opinion, 3-4 gets hazy, and 5 moves are generally beyond what I can see. This causes me to miss many combinations (at times forced). This is worst in the middle game. During the endgame, it's generally a lot easi en 1450-1500 on chess.emrald.net for over a year now. Any help would be appreciated. Thanks.

try chesstempo.com, in the standard mode (the rated solving mode without the clock), accurate calculation is called for and it's excellent for such exercise.

It's blitz mode is not superman-fast like CTS either.

I had that problem you're describing. Playing blindfold chess has helped me a lot with that. I can visualize and hold positions in my head much better than I used to. Also, positional knowledge and having a plan will reduce your number of candidate moves, which obviously makes your job easier.

I have the same problem.

At RHP I can shuffle my pieces around 'til I find the combo I want. IRL I have to glare at the board, and try to figure out what position I have after a few moves ahead in the combo. Often, after a 'genous sac', I find one of his bishops preventing the grand finale, and the game is ruined.

Perhaps I play too much here.

It is very difficult - I have found that playing blitz and doing puzzles have improved my 'feel' for the positions so I usually play an instinctive move that looks ok and has some sort of positional merit but spotting long and difficult combinations are hard, especially in the positions where you have been working towards a long term strategic advantage or are having to defend against an aggressive opponent and are not expecting any sort of attacking chances yourself.

The best way is not by solving puzzles, because you know there is a combination to be found and tend to look for the obviously attacking moves, but actually by getting a database game, playing through to somewhere in the middlegame and stopping and examining the position - best done on a board so you are not staring hard at the screen.
This is really good for improving your calculation powers, but you do have to work at it - you need quite a long time of thinking - start off by writing down lists of candidate moves and then analyse them.

I got this idea from 'think like a grandmaster by alexander kotov' and it was probably the most important factor in my moving from 1800 to 2000.

Thanks a lot for all of your replys! Mr. Chex, I will see if I can locate that book at my local library, and will try some of those exerceises you ghave suggested time permitting.

I was lucky to have my first serious chess book be "improve your chess now" by Tisdall. I think he section on calculation got me started on the right foot. Based from Kotov's "Think Like a Grandmaster" Tisdall gives a few example problems and then discusses the issues involved with calculation.

I often make it a habit to do a stoyko like exercise. That is I take a diagram from a chess book, informant, nic whatever and sit down for 15-30 minutes analyzing as deeply as possible the position writing down everything I see.

Less often, I do this but I setup 4 boards and give myself 1 hour to complete all four boards. I do this when there is more than 2-3 weeks till my next OTB tournament.

I think tactics problems are useful as well, but the stoyko exercise is superior (imo) to improve calculation.

this is all after one has gotten the color of squares, diagonals, knight patterns etc all worked out. If you can't answer what color square c4 is, or actually see all the squares a knight can attack sitting on e5 than that is probably where you want to start.

Originally posted by nimzo5
I was lucky to have my first serious chess book be "improve your chess now" by Tisdall. I think he section on calculation got me started on the right foot. Based from Kotov's "Think Like a Grandmaster" Tisdall gives a few example problems and then discusses the issues involved with calculation.

I often make it a habit to do a stoyko like exercise. That is ...[text shortened]... e squares a knight can attack sitting on e5 than that is probably where you want to start.

Interesting. I can "calculate" (a1 is black, b1 is white, c1 is black, c2 is white ... c4 is white) what color square c4 is, but can't "see" it in my head. Same for knight moves, etc.

Some say they can see the entire board and pieces when they calculate blind, but most see some limited view of the board or even simply the lines of force that pieces have.

Another way to build up visualization is to play through a chess book without using a board- meaning you have to visualize the moves in between diagrams. This is fairly hard (depending on the complexity of the position) pretty tiring exercise but I think it helps.

Originally posted by Tyrannosauruschex
start off by writing down lists of candidate moves and then analyse them

I think it's better to complete all your analysis before putting pen to paper. Analyse fully... write down... then check. Otherwise the pen and paper are too much of an aid.

Originally posted by jekeckel
I can "calculate" (a1 is black, b1 is white, c1 is black, c2 is white ... c4 is white) what color square c4 is, but can't "see" it in my head.

I've heard one GM say that the stronger a player becomes, the more abstract their mental visualisation of the board becomes. For example, they may use more relationships and patterns such as "I know I'm castled kingside with a fianchetto bishop there", "my knight is attacking his pawn which is protected by the bishop", etc. Aiming for a photographic representation isn't necessarily the way this develops.

Compare this with something you are very familiar with such as the layout of your house. By-all-means, you can visualise how it looks using some form of images in your mind, but its probably easier and clearer when you think of how things are arranged in relation to each other, e.g. "a chair next to the window next to...."

this idea is called chunking. Interestingly, there have been studies that show stronger players do less calculation (see De Groot etc on this) in fact it is players around 1900-2200 who do the most calculation during a game of chess.

Originally posted by nimzo5
this idea is called chunking. Interestingly, there have been studies that show stronger players do less calculation (see De Groot etc on this) in fact it is players around 1900-2200 who do the most calculation during a game of chess.

Doing "more" calculation is the wrong term, as it would be obviously wrong to claim stronger players do less calculation. If I remember correctly, the study's conclusion was that weaker players calculate about the same or more width, and stronger players kept their branches thinner but went a lot deeper in them.

I'm in that range you mention and I'm pretty sure Kramnik is doing more calculation than me 🙂

Of course a gm should be able to calculate much deeper- but I think the study (and there have been several) that a gm calculates far less frequently (not depth of calculation) during a game than does an IM etc on down. At around the expert level (2000 fide) they found that a player is calculating on nearly every move. This goes down as the rating goes up.

De Groot found similar results back in the day when he tested various masters- he was surprised out how much a master (and we are talking, keres, alekhine who took part in the test) is actually using memory to solve problems and not calculation.