lichess.org
Donate

How do you visualise chess positions?

I don't know if this is relevant to anyone, and suspect like 2 or 3 people at most will read this, but I figure I'd post it here regardless. I've recently been working on my visualization ability and testing my limits in this regard. It's kind of essential to have this ability if you want to be able to study chess effectively from books. In longer classical games I can, for the most part, now remember and play through games mentally provided there's nothing too crazy going on in a middle game. In shorter time controls, I'm completely helpless.

I've only ever been able to see a small section of the board at any given time. I can actively mentally picture only around a quarter of the board pretty much perfectly regardless of the situation. Any more than that and things start getting more fuzzy. If I "look" at other parts of the board I can "see" them. I know what pieces are where and that they exist, but I can't keep them in my mental stack easily when visualizing. I believe I tend to use the pawn structure as a sort of landscape and remember pieces in relation to it. I believe this works well because I've typically seen pawn structures many, many times allowing me to chunk them more easily in addition to the fact that they don't tend to rapidly change, unlike nearly every other piece.

However, as you might imagine, this makes it difficult when I get down to things like king and pawn races. Since they're on different parts of the board. and since there's not much pawn structure left for reference as a landscape, I can struggle to keep track of my calculations in these. Additionally, the board in my mind starts to warp, in almost non-Euclidean ways, if I try to instead visualize the checkerboard pattern as my reference. Diagonals might curve. A rank might be the 5th on one edge of the board, and the 4th if I try to follow it over. Squares stop being the same size. Etc. One trick I've found that seems to help me greatly in these positions with large empty spaces is to create an artificial landscape over the board in the form of a lattice. For me, I find it helps to visualize the squares along the 2nd and 7th ranks and the b and g files and to remember pieces and pawns in relationship to this rather than an actual "chessboard" pattern. I mentally highlight the squares along these two ranks and two files as a purplish pink personally, and remember where the pieces stand in relation to this framework. It's slow, and far from perfect, as I have to mentally check where the pieces all lie in relation to this framework, as if setting them up on a parallel board, but it seems to greatly reduce any warping of the chessboard that occurs if I try to visualize it normally for whatever reason. I've been toying with the idea of trying to do this for all positions from the get go, but it's really only needed for the late endgames, as the piece locations don't shift for me, only the structure of the chessboard itself really tends to warp badly. It does make it easier to trace pieces across the chessboard however, which can be difficult since I can really only visualize about a quarter of the board normally. Additionally, it would likely stop me having to mentally check where pieces lie in relation to this different board structure as I'd already know. If anyone else has these "chessboard warping" issues, I'd recommend trying to visualize things with this set of highlighted squares, and remembering where your pieces lay in relation to this structure for a little bit to see if it helps. It was pretty much instant for me, taking only a day or so before I could notice a difference.

Additionally, my father has aphantasia. I certainly don't have aphantasia, but if visualization ability is genetically influenced, at least in part, it might explain why I've always struggled in this regard with chess.

I'd also be interested in knowing how this affects the rate of blunders that are just careless oversights. e.g. missing seeing something is attacked, missing that a piece is hanging, etc. While certainly overrated, I at least would like to believe I'm decent at chess. Far from good, but decent. However, this is something I struggle with greatly. Every 2 or 3 blitz games of 3+2 I'll just miss that a piece or pawn is attacked or hanging. I've noticed though that this rate varies greatly even between players at similar skill levels. I'm in the more blunder heavy side I feel. Some players feel like they have little positional or tactical strength, but seem to almost never blunder. Likewise other players are much more peaky. I feel like this may be related to an inability to mentally keep the whole board continuously "present" in my mind rather than breaking it up and looking at individual pieces of it. I'd be interested in seeing if those are at all correlated, although it might be a little difficult to test in a non subjective manner.

If anyone has any suggestions for practice for better visualization or questions, let me know and I'll try to answer them.
@LanceFairfield This conversation made me think that maybe learning with a custom board could help with visualization and found an easy way to do it here www.youtube.com/watch?v=hpShZFp7hYs although I used stylus (github.com/openstyles/stylus/wiki/Usercss) because stylish was showing me server errors.

So naturally after reading your comment (especially because I'm relating to your experience a lot) I had to try your mental board next. This is what it could look like imgur.com/a/lankEsV

I have no idea whether training with it online could help burn it into memory and help with blindfold visualization but now I'm intrigued by the idea.

At first it was thinking of some kind of gradient map or dividing the board into quarters with each corner having its own color. I'm struggling with coordinates so trying to think with annotations (i.e. there is a knight on g6) is pretty much useless to me in a real game. Maybe having a color associated with specific parts of the board could help.
First, I would like to point out aphantasia may affect as little as 1% of the population. It is hard for me to believe therefore that it has a significant impact for most chess players. It is not irrelevant per se, but certainly an outlying question.

Some questions that could be considered: do you employ approach V or M? Is it some mixture of both? Or something else entirely?

- I believe generally I employ some form of M, but it is hard for me to say with certainty. It is likely a combination, as I will explain below. I generally think though, that I group information together in clumps rather than as specifically as you made it sound in the explanation of approach 'M'.

If it is V: how does the picture of the position in your mind look like? Is it lively/detailed or rather abstract / does it only contain the necessary information?

- When I play blindfold, which I have done quite successfully against players with a bit of a rating gap, I tend to picture the board. I often find I do this during classical games of considerable length, as well. I can walk away and still analyze, or potentially recall a position or the entire sequence of moves days later. However, I do believe I visualize it more abstractly and without much detail. I don't know how this factors into the rest of my visualization.

If it is M: describe what kind of facts you store in your memory, how you recall them, etc. And if you don't have aphantasia, why do you think you gravitate towards this approach? Is a chess position simply a too complex/detailed scene to visualise fully?

- This part is hard, but I think I store patterns and ideas. I will look at relative position, factors such as piece placement, structure, combinations of pieces, board geometry. I do not know how I "store" them. I imagine my chess visualization is rather suboptimal and disorganized. Recall is probably based on pattern recognition. Humans tend to be excellent at that.

I think more people use this method than you may realize. It is more about the WAY people think and natural thought structures of our neurology and psychology than it is a choice in method. It isn't that I, or others, are incapable of raw visualization. I think it is far less efficient for most players, below master level, to employ this method regularly. It takes far too much time. Although, it now makes me wonder if this isn't why one of my students regularly gets into time trouble. Perhaps she visualizes using the 'V' approach.

To break it down more, I would say I use the 'M' method as a reference point to guide my cursory analysis, while I reserve 'V' for when I need to specifically calculate deeply. I determine that by position complexity and how persuasive my candidate moves appear to be, although I don't have a specific method at this time for determining those parameters. I think I choose this way because it is efficient. It saves time to clump information together and identify patterns.

I am rated ~2000 USCF btw.

You might consider talking to NM Elijah Logozar. I know he is highly interested in this type of research. He has a fair amount of sources on the way we learn and visualize in chess. I know Aagaard is an inspiration for a lot of what he knows. Regrettably, I have neglected to better structure my visualization, calculation, and analysis structure with recommended materials to date. It may guide you some to look into that.
I think yet another approach would be to not only think about position itself but how the position was arrived at - that helps me memorize the board because i can trace back to moves and reconstruct the board from the sequence of moves. I recall seeing some post about chess players being able to memorize positions that happen in the real games much better than random positions. I think our mind acquires patterns and compresses them as we learn and internalize chess over time especially if starting young. As a result those patterns can help us memorize the board by mere recognition of the position already stored with some modifications. In any case fascinating topic worth discussing!
@LanceFairfield said in #41:

>
> I've only ever been able to see a small section of the board at any given time. I can actively mentally picture only around a quarter of the board pretty much perfectly regardless of the situation. Any more than that and things start getting more fuzzy. If I "look" at other parts of the board I can "see" them. I know what pieces are where and that they exist, but I can't keep them in my mental stack easily when visualizing. I believe I tend to use the pawn structure as a sort of landscape and remember pieces

That is how vision works. Try looking at chess board.Not in your head put real one with pieces on it and try too all of there without focusing on some point. Does not work that way. You have look at diagonal of the bishop, king position etc separately to really see wha is there. So doing mentally cannot really be a full board vision either.
@iakov98
This is a very interesting topic and I'd like to share how blindfold works for me. For context, I am around 1900-2000 and am usually able to play a full game blindfold. Just recently I won a 50-move rapid game blindfold against a ~1750 who had vision of the board.
I don't think the ability to play blindfold is equivalent to be strong in calculating tactics. I think I'm at most average in calculation (for my rating) due to lack of motivation/concentration.

For me blindfold works mostly with *pattern recognition* and *knowledge* and I only fall back to visualising a part of the board when a pattern is unknown to me.
By *pattern recognition* I mean that typical setups like a castled king with fianchetto'ed bishop are familiar, so I don't need to memorize the individual pieces but just the pattern. On the Queen side I might memorize the pawn structure a2-b2-c3 vs a6-b5 as one familiar pattern. I remember it as a pattern, but if action starts in this area of the board, I will visualise it.
By *knowledge* I mean that a Bd3 aims at h7, if a knight on c6 is captued, bxc6 and dxc6 are two options to recapture, etc.

Let me give an example how this works in practice. Let's say the game starts 1.e4 c5 2.Nf3. I had this position thousands of times on the board and know that this position is fine for both sides, what are the possible moves and ideas. It wouldn't add any benefit for me to visualise it (which takes me energy). If the game continues along typical paths, I continue like this and can play in blitz mode.
What happens if there is an unfamiliar move? Let's say Black replies 2...f7-f6. Firstly, I apply knowledge: I know that f6 opens up the h5-e8 diagonal (which I then visualise automatically) and that a Bc4 could later aim at f7. The other implications of 2...f7-f6 is that it attacks g5 and e5. The field g5 is relevant for later Ng5 or Bg5 ideas, which I don't care so much now. The other attacked field e5 is in the center, which is more relevant in the opening. I will now actively visualise the center: e4, c5, f6 which attacks e5, Nf3 which can't got to e5.
The visualisation helps to save this pattern temporarily and I will continue with the main plan of 3.d4.

In the middle game the possible moves and ideas are dictated by the pawn structure (Philidor). Whenever the pawn structure changes significantly, I take a moment visualise the new structure so that I know which lines and diagonals have opened up and which pieces are affected, what are the new possible captures, checks, threats.

Similarly in the end game, if I need to calculate a pawn race or if I am constructing a mating pattern, I will actively visualise the relevant part of the board.

Lastly, I want to be clear what I mean by visualising a "part of the board". I never visualise all 64 squares at once, but I also don't cut the board into 4x4 areas. Instead I look at all pieces which affect a certain idea or area of the board, even if it's a long range bishop or rook, but I completely disregard any pawns and pieces which are far away or haven't moved or haven't been relevant in a long time.
And that's exactly why oversights can happen eventually! :-)
@i-bex said in #42:
> @LanceFairfield This conversation made me think that maybe learning with a custom board could help with visualization and found an easy way to do it here
> although I used stylus (github.com/openstyles/stylus/wiki/Usercss) because stylish was showing me server errors.
>
> So naturally after reading your comment (especially because I'm relating to your experience a lot) I had to try your mental board next. This is what it could look like imgur.com/a/lankEsV
>
> I have no idea whether training with it online could help burn it into memory and help with blindfold visualization but now I'm intrigued by the idea.
>
> At first it was thinking of some kind of gradient map or dividing the board into quarters with each corner having its own color. I'm struggling with coordinates so trying to think with annotations (i.e. there is a knight on g6) is pretty much useless to me in a real game. Maybe having a color associated with specific parts of the board could help.

Yeah, thanks for the link! I've been tempted to look into creating something like that, but I'm still not sure whether it would help or hurt long term. I'd have to switch back to a normal board after all when playing otb. I haven't really decided yet. I mentally use much more distinct shade for those ranks and files, something that REALLY contrasts, but maybe that's partly because I'm color blind lol. I don't want to have to think at all to tell what's what. The reason I went with the structure I did is also because every square is now only at most 1 square away from my lattice pattern (in the case of the center 4 squares) and everything else is either on, or adjacent to it, while only having to keep really two lines in mind as reference points. Thanks again though!

@petri999 said in #45:
> That is how vision works. Try looking at chess board.Not in your head put real one with pieces on it and try too all of there without focusing on some point. Does not work that way. You have look at diagonal of the bishop, king position etc separately to really see wha is there. So doing mentally cannot really be a full board vision either.

I understand what you are saying, but it's far less the case when the pieces are visible. I focus on certain things, but I have other things in the background that I'm much more aware of mentally. Additionally the size of the area I'm "looking" at and focused on mentally is smaller and more rigid of an area than what I can do with a chessboard when I can see the pieces.
@LanceFairfield I actually started with way more vivid colors imgur.com/a/7Dgnhea but felt it's too strong and distracting. Who knows whether practicing with it could actually be detrimental for OTB. You might create some internal references that are suddenly missing on the board but maybe it's not too hard to imagine them while playing.
I have an outstanding memory, however my chess instincts aren't very good. Thus I lean on the M method heavily to memorize 1000+ chess positions from the openings and their themes/important squares. I try to guide every game into a position that is familiar to me. My chess game plan is all about controlling the direction of the game from move 1. Though I do agree that compared to my friend, who focuses more on getting to the middle game quickly and soundly, I am consistently behind on the clock. Of course he has a knack for choosing quiet sidelines that take me out of my prep. When it comes to actually visualizing a combination, I feel as if there is no other way than just to move the pieces in your mind. You know how when you click on a piece, and the dots appear to give you the allowable squares? I will pause in the middle of my visualization and "click" on friendly and enemy knights or other pieces (in my mind) to see how and when they could influence my attack/defense. My method isn't the best for sure, but it is the most fun for me!

P.s. This post pertains to my over the board and online games
I visualise the position/my calculations on the chessboard, for some reason that is way easier for me than visualising in my mind. Maybe that is trainable? When visualising on the chessboard I can go very deep. When doing it in my mind it requires more focus and I can still remember where the pieces are but I usually just don't completely see the position.

This topic has been archived and can no longer be replied to.