How to See Three Moves Ahead in Chess

Calculation in chess requires visualization. At least it does for us humans. I don't know what those computers are doing when they calculate, it boggles the mind to think of it. But we actually have to see in our mind's eye the piece on the new square, and then imagine all the possibilities for our opponent, and visualize the moves, and so forth. Training exercizes will help us realize that we can do this, and going further than we think we can will make us accept nothing less when we play in our next tournament. So let us begin an excersize. Look at the following diagram. Keep in mind that the further along we are into a game the more there is to calculate. For a simpler position with less to calculate check out How To Think Three Moves Ahead. But in this next exercize we will choose a more complicated position so we will have more to calculate.

A candidate move is a move that you are considering. Before I tell you what Topolov (an attacking player) played here, take a moment to determine what candidate moves you would come up with. Here is a list of all the possible moves you can made in this position. This is move 11 in the game between Topolov and World Champion Anand, taking place in 2005. I will go through the list from left to right listing every possible move for white. Try to visualize each one of them. Just glance briefly, and briefly answer a possible response for each. Get a sense of which moves are worth seriously considering. Many strong players go further in their calculations even on ridiculous looking moves.

11. a3, 11. a4, 11. b4, 11.d5 11. dxc5, 11. g4 11. h3, 11. h4 These are all the possible pawn moves. Now on to the rook.

11. Rb1, 11. Rc1, 11. Rd1 Those are all the rook moves. Onto the Queen.

11. Qb1, 11. Qb2, 11. Qc1, 11. Qc3, 11. Qd1, 11. Qd3, 11. Qe4, 11. Qf5, 11. Qg6, 11. Qxh7, That's all the Queen moves. Now the Bishop.

11. Bc3, 11. Bb4, 11. Ba5, 11. Bc1, 11. Be3, 11. Bf4. That's all the Bishop moves. Remember to visualize the piece on the square, and to briefly asses it sitting there. Don't think it is silly or a waste of time to consider moves that look absurd, by this practice you will always be thorough in your consideration of the position. And so far as how much time it takes, remember that many players, especially impatient young players, lose the game with plenty of time left on their clock. So, onto the king.

11. Kd1, 11. Ke2. That was short.

11. Be2, 11. Bd3, 11. Bg2, 11. Bh3. Notice the lack of mobility for certain pieces where they sit? Not that this is necessarily the most important consideration in this position, especially considering Topolov's choice of moves here. (That's a hint).

11. Rg1. We're almost done, now onto the knights.

11. Nfe5, 11. Nfg1, 11. Nfh4.

11. Nge3, 11. Nxe6, 11. Ngxf7, 11. Ngxh7, 11. Ngh3. Now we know every possible move. There's quite a lot of them, let's count them. Total is 41 possible moves for White. Which of these should become candidate moves?

How do we decide which will be the candidate moves? What should be the criterion? This could almost be considered a philosophical question, except that computers seem to know the "truth" or do they? Because there are always faster and deeper and more amazing computers beign built. And besides, we are not computers. We can feed the information to them and see what they "think" but we are still human, and must ask human questions about which moves to consider. So with that in mind, let's move on to how to determine the candidate moves.

What do you observe about this position. Let's just list these observations. If you have any to add you can put them in the comments.

1. White cannot legally castle this move.

2. White's bishop is still on its original square.

3. Black hasn't castled either but can legally castle next move.

4. There is a white pawn on d4 and it is under by a pawn and the black queen.

5. The pawn could capture c5, and be recaptured by the bishop or the pawn.

6. Black's light square bishop is pinning the knight on f3, and aiming at the knight on f3, and the knight on g5 is the only piece gaurding it, and the knight on g5 could be kicked by pawn to h6 (and I get the feeling Topolov could care less).

7. Although White cannot castle, there are no immediate (meaning next move) checks to White's king, nor any routes through which any of Black's pieces can immediately (meaning next move or two) begin an attack on White's king.

8. The black king is the only one defending f7.

There are plenty more observations to make. There are many more than we think there are, there always are, that's why great players stare at the board so long. Don't believe me? Let's make a few more.

9. The bishop currently has no way to get past c4 on on the f1-a6 diagonal, but that could change if the pawn on d4 moved to d5, if 12. exd5, cxd5.

10. If the knight captured the pawn on f7 it would aim at the queen. If the king captured the knight on f7 it could be checked by the knight on f3.

11. It would take one tempo for Black's bishop to capture the rook on h1.

12. If the pawn on e6 were to move the f5 square would be available for White's queen.

Think that's the end? I bet it's not. Let's see if there are more.

13. A white rook on d1 would be on the same file as Black's queen.

14. White's dark squared bishop could protect the pawn on d4 by moving to c3, but it would have to be the one that recaptured, not the knight on f3 if you wanted to stop the bishop on b7 from capturing the rook on h1.

There are plenty of trivial, or at least what we might assume are trivial observations too. Let's find some of those.

15. Pawn to a3 will guard the b4 square but give up control of the b3 square.

16. If Black develops his knight to c6 he will block the bishop from the diagonal and unpin the knight on f3. (Perhaps not so trivial)?

17. Bishop e2 guards the knight but doesn't unpin it.

Do you think you could ever make every possible observation there is to be made about this position? We will stop here, knowing we have not covered everything, but giving you a sense of what it means to make observations about a position. Next we will choose a plan, then the candidate moves will be based on some sort of plan. It's this or react willy nilly to immediate threats with no plan in mind. Hardly what Topolov was doing.

 Here are some possible plans. I'm just throwing them out there, these are not necessarily the best, or what Topolov was thinking.

1. Guard the knight by placing the bishop on g2 and castling next move. Lose the pawn on d4. Horrible choice, but at least it saves the knight on f3 from the guard of it being chased away with h6 next move.

2. We can guard the pawn on d4 by playing Bc3. Of course this leaves h6 which kicks the guard away from the knight on f3.

3. We can capture the pawn on c5 with the pawn on d4, and then when we lose the knight.

4. We can move the queen to d3 to guard the pawn but then the knight has to recapture the pawn and then the knight hangs. So we can dismiss this.

5. We can push the pawn, like Topolov did.

 He opened up the diagonal for the bishop when he recaptured with cxd5. He temporarily blocked the diagonal. He made the f5 square available when the pawn on e6 captures. He opened up the a1-h8 diagonal for his bishop if he needs it. Let's look at Blacks next move.

 See if you can figure out why Black doesn't recapture. We have calculated only one and a half moves thus far, and asking ourselves why he wouldn't recapture. We will stop here knowing that we haven't fully calculated the three moves, but you have seen more about what it means to see three moves ahead. To see the rest of the game click here. Feel free to leave comments below with your ideas. We did not use a computer to analyze the game because I wanted to develop our own analysis skills. But Computer analysis can be very helpful as well.