Chess calculation and how you think ?

I've just completed a chess puzzle and it was one of those forced sequences where at the end of the sequence you're up material.

I'm curious how you guys tot thinks up in your mind ,personally i do a running total in my head plus or minus as pieces come of the board much the same way a chess engine would score it , but i find doing the maths whilst visualising is difficult .

so how do you lot do it , is your vision that good that you can just look at the ending position and then tot things up?

The goal of daily practice is primarily to build up mental resources, alongside being sharp to the fact of the pieces on the board. As conscious efforts and certain patterns become subconscious knowledge, you will not have to calculate as much any more. However, there will be scenarios where that still happens.

My point is, it just gets easier over time as you build up your skill. When you practice puzzles you practice a whole range of skills that align with it. Counting and visualization are part of it, altho they level up slower as the main practice is the tactical motifs themselves.

I do not have a method to offer, but if anyone does I'd also like to hear it.

@sausage4mash
@bak000za

Most puzzles can be solved by knowing the most common tactical motifs and by analyzing the starting position carefully to find all weak spots. Then you must find the most logical candidate moves and play them in the most forcing order.

I made 2 studies about that stuff. Maybe it is as helpful for you as it was for me.

Not that my method is any good, but: I calculate out all the exchanges first, takes, takes, takes, and so on, and then I say, let's see I took a knight, a queen, and a bishop; he took a rook, a queen and a knight. So, down an exchange if I do it that way.
Then I might look at another order of captures, and tally up.
Naturally we are not only looking at material, but after running the totals, I consider other factors.
I find it confusing sometimes, as you say, especially if material of different value is being taken in different order. Such as, when you sac your queen, then have to make sure you have enough material for it. Or if you let your queen be taken first, then need to make sure the exchanges are forced up until you take the opposing queen.
Not sure that helps, but you asked, so I expounded.
Of course, most of the responders are likely much better than I am.

hi sparowe , i too adopt that method on the shorter lines and revert to tallying up as i go, when my memory fails me .

I used to calculate without actually seeing the board in my mind, but just considering a sequence of moves (just the relationships between the pieces), it works fast and it's not so hard to do but the price is many blind spots.

What I do now is visualization, I'm still struggling a lot about it but I find it more efficient and accurate. Visualizing forcing lines and trying to assess if the outcome is any good for me. It's good practice for endgames, middlegames are kind of very messy, I often end up unable to say if I'm better or not.

I try to greatly limit my search depth when ever possible. I try to play chess at a depth of about three moves or six ply, Unless a longer forcing sequence jumps out at me.

Three move chess is good for going wider and looking at more candidates. Things like intermediate moves and temporarily leaving pieces hanging for other gains become normal using a shorter line.

Every line should end in an evaluation either tentative ( my pieces are a bit better here) or absolute( I'm up a pawn for no compensation here) so that you can put that line to the side and move on.

Most of us train the sword more than the shield, so improvement in calculation should likely start with our weakest link: opponents idea plans and threats.

Finally we play chess positions much better when we are familiar with them. Becoming familiar with the positions that occur from the openings you play will cause you to calculate good ideas for both sides in the early middle game. When your opponent plays a bad idea for that position ,you can delve into your familiarity with that position to chose a better idea. The result is your play will advance faster than your opponents.

As mentioned we train attacking more than defending so if you get a good position by move twenty it will be much easier for you to play and calculate than your opponent.

Visualization is the least developed essential skill of most club players, because it is hard and takes time to develop.
I think you can work on visualization in the endgame, since there are less pieces it's a good training ground.

Calculation and tactics make heavy use of forcing moves if there are no forcing moves Calculation is no longer your primary tool for evaluating the position.

1. Short lines, look wide.
2. End each line with an evaluation
3. Focus on opponents moves and intentions at every step
4. Become familiar with many early middle games positions that arise from your openings.
5.It's easier to calculate in good positions
6. Practice visualization while learning endgames.
7. If there is nothing to calculate, don't calculate.

@comfortable
2. End each line with an evaluation

That is something i would like to train, as other train tactics using puzzles. Do you know of attempts to train that aspect, evaluation of a FEN.

I play only correspondance, and sometimes gets hypnotized by the sound the board makes (slightly kidding), i mean i get lost in deep lines, and somehow, after a lot of that, loose my ability to tell if i like a position or not.

I am not well educated in chess as many of you may be. i only can remember principles, hard to learn openings, and i believe in my own brain statistical engine, comforted with strategy principles i cam across here or there and remembered.

but when i calculate too many lines, i loose that instinctive (?) evaluation which could mean stopping a variation earlier, and instead consider "wider" as you put it, nicely (it suggests that you project variation lines in some sort of metric space, i could be inferring too much).

so how to train that step two? focusing on that, and train evaluation of improvements along a 6 ply search, or even shorter, if the exercise can be made.

tests showing to FENs reachable from each other by few moves, couldn't exercises train people to choose the most promising one? Could be embedded in teaching positional principles, or just building subconscious skills, or a judicious balance of both.

It would be a big job to create test's showing potential fen's reachable threw a whole game. You could do it like a chose your own adventure game.

If variation A appeals to you go to chapter 21
If variation B appeals to you go to chapter 7
Etc..

It could be done fairly well using lichess studies if you limited the choices down to three for a few positions in an annotated game. It could get out of hand quickly though, with to many positions

For a wider search you could have one position with eight candidate moves and you advance to a certain chapter to learn about the resulting position and how play might develop.

chessgym.net has a feature called tactics:positional, where you analyse a given position for as long as you want then guess the best move in the position. They then give you the top twenty engine moves from best to blunder with a line that you can play through for each top twenty moves. You can see how close you got to engine best,while practicing analysis.

@sausage4mash

Calculating the moves and looking which stones disappear from the board is the normal method if it's about material. That's the way I do it and everybody else I talked about this. I talked with a lot of people about this. When you get stronger you do this nearly automatically and at the end of the calculation process the question of evaluatiing the position gets more important. Have you reached a calm point without checks and exchanges? Is the line a progress to the actual position?

Thinking about thinking in chess was a 'hobby inside the hobby' for over 40 years. I could write a little essay about it. But as I'm no master I recommend two sources. Kotov - Think like a GM and more modern and more practically Tisdall - Improve your chess now. People often prefer Kotov, because he sells a normative approach. Tisdall sticks more to the natural thinking of humans. That's why it is easier to dislike him. Have this in mind, if you read critics about the books.