Find the Grandmaster Move

Find the Grandmaster Move

| 27 | Strategy

In our previous installment, we discussed the concept of pattern recognition and this time we present a method for developing overall chess skills without having to rely on static generalizations or idealized simplicity (patterns).

Playing chess requires a variety of skills and there is an exercise recommended by professionals and chess coaches alike that works miracles both for our positional and general chess understanding: the guessing game, where we take an annotated game of a strong chess player and try to guess the moves played one-by-one.


And why do we need positional feel or flair? Because we can’t calculate everything, and our sense of the position clues us in on what to calculate. Positional understanding may be described as our first filter, our subconscious grasp, towards a position where the first candidate moves for calculation announce themselves.

 "Guessing" might be too strong as we are not guessing in the lottery sense of the word; we try to find, we look for, assess, evaluate, compare, etc. and all of these involve calculation on same level. Intuition, for instance, is nothing but calculations of which our brain (or mind, if this feels less alienating) only serves us the conclusion (Krogius, 1976, p. 45).

We recommend guessing moves of the old-school players such as Rubinstein, Capablanca, Alekhine, Botvinnik, Smyslov but also modern players like Keres, Spassky, Fischer and Karpov etc., players renowned for their positional flair. At first, it might be hard to guess correctly but with some practice we may even improve upon the move played and our positional understanding will improve as this exercise works all chess muscles at the same time: the art of the exchange, piece mobility, cooperation and coordination between pieces, understading of the initiative (how to seize it and how keep it), calculation, prophylaxis, attack and defence, endgame and middlegame skills, familiarity with plans and strategies, grabbing space, determination, resoluteness and chess intuition.

We should:

  • Cover up both players’ moves and uncover the moves one by one, since by catching a glimpse of the opponent’s move we might get clued in on the move played.
  • Try to make the game our own, i.e. treat it as a real (tournament) game, as it is just too easy to sacrifice someone else’s wood.
  • Guess the move before reading the annotations or see the diagram, as these as well may clue us in.


We recommend spending at least two minutes per move, since this makes even short games a decent training session but with practice we will guess faster (and more correctly). If our move doesn’t match the text, we put it in brackets or parentheses. To ease into this exercise, instead of only one move, why not make a list of for instance four possible candidate moves and if the move played matches one of ours, we pat ourselves on the back and grant ourselves a point as the move at least showed up on our radar, another benefit of this exercise. The stronger and more refined our positional understanding becomes, the fewer moves will be on our list.

How about turning our guessing game into a sport: how many correct moves in a row? Do we guess better with White or Black? By timing ourselves we can see if we speed up over time. Say we work our way down from five minutes per move to 30 seconds and within that time we have to move. The guessing game may be used as a diagnostic tool: Do we more easily guess attacking moves or defensive moves? And what about long, slow manoeuvring games?

What if we slip up and see the move? Two things might be done:

  • Leave the field blank and skip it (we don’t calculate the move in our score).
  • Declare we may have a move of our own (though this might be difficult, since we’re already influenced by the move seen).

And, by the way, we seize the opportunity to strike a blow for descriptive notation, often frowned upon, but which in this exercise takes centerstage in all its glory and is much to be preferred to the modern algebraic or figurine notation: if we by accident catch a glimpse of a move, our guess may be wrong since based on what we think we saw but not the move actually played.

After sliding a sheet of paper into the book to cover the moves before opening it, not to catch a glimpse, we’re all set and ready.


Let’s start with the opening.

We may either skip a few opening moves and start from the early middlegame or guess the opening moves as well, strengthening or reactivating our opening knowledge, perhaps even learn a new opening along the way or improving on existing theory. Familiarity with the variation, not only the opening, the time period the game was played and the GM’s preferences and predilections, may facilitate our guessing.

Having only the name of the opening, and not the variation, for instance the French, grants four possibilities, 3.Nc3, 3.Nd2, 3.e5 and 3.exd5, whereas having the name of the variation as well, like the Tarrasch, we may know 3.Nd2 is played and get a hit or a match, for free, as it were.

By guessing moves in the opening, two possible priorities seem at work, pulling our ego in two opposite directions:

  • Match our move with the move played.
  • Aspire to play the better move, regardless of fashion or previous history.

For example, in the Grünfeld-Indian, players once went for the Russian variation with 4.Nf3 Bg7 and 5.Qb3, instead of 5.cxd5 but since we’re updated on the theory, we play 5.cxd5 instead, and don’t match the text, thus getting fewer hits but striving to play the better move. However, what if we guess the correct move and it turns out to be a gross blunder? Well, take comfort in the privilege of blundering like a GM. On encountering move repetition, our GM may be trying to gain time, we should still prefer moves that carry the game forward for every move.

We try to guess the best move regardless of the player because if guessing the moves of a Bent Larsen or a Nimzowitsch we might be inclined to pick moves we would not otherwise play just to match these players’ scoresheet. Assume we think g4-g5, gaining kingside space, a good move, and which turns out to be best. But no! Our GM played Nb6-a8 for the sake of artistic eccentricity.

The guessing game improves our technique (perhaps especially in the endgame) since familiarity with master games increases our knowledge, which in turn equals improved technique, as technique (and technology) derives from the Greek term techne, meaning know-how, skill, proficiency, craftsmanship, craft, or art in producing, making or doing something as opposed to and different from epistēmē as disinterested understanding.

Yet another possible benefit is an improved ability to stay calm and focussed and find good or useful moves under time pressure, for instance when playing on the increment only (or with a huge crowd watching). Also,the guessing game develops endurance and perseverance and thus prepares us for long and hard games.


This exercise teaches us to trust ourselves and our instincts. If we first think of b2-b4 but write down a different move but where our first impulse does prove correct, we stick to our first instinct—the pawn move.

What happens might be described as follows: our brain triggers a move—b4—or we get a glimpse of it from the corner of our eye so to speak, before the move turns into a more conscious experience, then we become conscious of the move and seek/concoct a justification: "I want to play this because…" and we may become entangled in long-winded discussions with ourselves as to why we want to play this or that, but forget about all that. (Remember from our previous article: move first—explanation later.)

Interesting is when our explanation for our move is at odds with what still turns out to be the correct move. The reason we stick to our first impulse, is to learn when to trust our instincts. We may throw out a brilliant move without knowing why, or being able to put into words, why a specific move is to be played or preferred. Note how you feel when your instincts/intuition prove correct, as this is the feeling we want to nurture and cultivate, and over time, intuitive precision develops because this exercise also works our perception.

It is said that the correct move often occurs to the GM in a few seconds and we can try to notice, after a while of training, if the first move occurring to us is the correct one and if so, it is a sure sign our perception, understanding and intuition are improving.

Calculating our score

So, we have worked our way through a game and are ready to calculate our score, an example illustrated by the scoresheet below. (We don’t have to use a scoresheet, regular A4 sheets of paper, napkins, envelopes, post-it etc. work as well but for the sake of clarity of presentation we use a scoresheet.)

Downtown Chess Club San Antonio, TX

Event: World Championship, 18th

Round: 4

White: Mikhail Botvinnik


Black: Samuel Reshevsky


Opening: Nimzo-Indian E51



1 d4

23 (Nxd4)v

2 c4

24 (h3)v

3 Nc3

25 h3

4 e3

26 Nd5

5 (Nf3)v

27 Re2

6 Nf3

28 (Nh4)v

7 (Bd3)v

29 gxf5

8 Bb2

30 Qxe4

9 Bd3

31 (Rd3)v

10 Bxc4

32 Nf4

11 (0-0)v


12 b5


13 Bd3


14 Bxb5


15 (Bd3)v


16 Qd3


17 (Bxd7)v


18 Bxd7


19 0-0


20 (Rfd1)


21 Bc1


22 (d5)v


(Scoresheet courtesy of RC Gonzalez and by kind permission of Robert Nomad, San Antonio Chess Club, TX, our modifications.)

In our example we have guessed all White’s moves, including the opening, and the first step is to count how many of the moves match Botvinnik’s. Remembering that bracketed moves are moves not matching the GM’s, we count 21 matches for a total of 32. Then we calculate our score: 21 x 100/32 (or 21/0.32) = 65.62 percent. 

In the second step we return to our bracketed or parenthesised moves and notice the 'v' next to the move. After finishing guessing, we run our checkmarked moves by an engine and mark those moves equal to or better than the GM’s (we get back to equal/better in a moment). We count the number of checkmarked moves (10) and add these to our previous score (21) bringing our score up to a subsidiary or alternative total of 31 of 32 and 96.87 percent. 

However, if we exclude the first six opening moves, starting on move seven, the math looks like this: 16 (moves that match) x 100/26 (or 16/0.26) = 61.53 percent or, with our better suggestions (nine, as the first checkmarked move falls outside our guessing list): added (16+9): 25 x 100/26 (or 25/0.26) = 96.15 percent

We checkmark a move when equal or better but what count as equal? With the entry of chess engines, equal  comes with a certain numerical tolerance between 0.00 and 0.30 (Komodo 10) and whether we prefer our move to be identical to the hundredth of the evaluation of the GM’s move or within the tolerance depend on personal preferences but once our move turns +/= we checkmark it. (The space between 0.00 and +/= makes up "the edge."

Playing the guessing game might give us an idea of how strong "old-school" players actually were. Assuming their move better than ours, how big is the difference—in terms of chess understanding and knowledge—between their (better) move and our move? What did they know that we don’t and how much of an effort would it take us to get there?

Checking our moves against an engine, we may learn not to trust the annotator (but ourselves) and the difference between human moves and computer moves i.e. learn to see unexpected possibilities, be open to weird-looking moves, etc.

In conclusion: we try not to get discouraged or bogged down in musings as to why if our move does not match the GM’s but trust our judgment and keep on guessing.


Krogius, N. (1976). Psychology in chess. New York, NY: RHM Press.


  • A heartfelt thank you to friend and colleague Victoria W. Guadagno for assistance and proofreading.
  • Robert Nomad for kind permission to use the scoresheet.
More from Vik-Hansen
Pattern Recognition—Fact Or Fiction?

Pattern Recognition—Fact Or Fiction?

Why Chess Players Blunder (Part 2)

Why Chess Players Blunder (Part 2)