Candidate moves
You might recall the following situation: after a game, you look at it with the help of an engine, and more than once, it asks for a move that you didn´t even imagine during the game. Or this one: the engine suggests a move that you, unfortunately, saw but discarded during the game. The latter can be particularly painful when you had a 50/50 choice, i.e. only 2 moves were worthy of consideration (=candidate moves), and you went down the wrong path.
Well, we will never be able to match the engine when it comes to sheer calculation, but at least we can try to surpass other humans at it! Here I will look at 4 cases related to candidate moves, all of which are very frequent in practice.
a) Few candidate moves. Defensive situations (king in check etc.), endgames.
b) Many candidate moves. When we are attacking, chaotic positions etc.
c) Move orders. We spot an idea, but only a correct move order will do the trick.
d) Secrets. Only after having checked move "a", we are able to find move "b".
You can find the other 3 articles here: Candidate moves II , Candidate moves III , and Candidate moves IV . They touch upon the b), c) and d) cases above, while this one focuses on a).
Note that these cases are often intertwined. Let´s see a simple illustration of the first case.
White to move. We soon realize that our only hope is to sacrifice the bishop for the g2 pawn, and eliminate the e6 pawn either by our king, or by d4-d5. 1.Ke5 is illegal, but an evident candidate move is 1.d5, getting half the job done. Following 1...exd5 (or 1...e5?! 2.d6) 2.Kxd5, what will Black do? Approach the king, it must be. After 2...Kg5 3.Ke4 Kg4 the g2-pawn is safe, that is bad news for us.
Let's go back and look for more candidate moves. Any other forcing option? 1. Be3+ Kg6 would just help Black´s king. 1.Kf3 looks illogical, decentralizing our king. But at least it gets our main enemy, the g2 pawn, off the board. After 1...g1Q 2.Bxg1 Bxg1, any need to look further?
Ah, 3.Ke4!, the king is back. Now if 3...Kg6, then 4.d5 forces the draw. As for 3...Bh2, after 4.d5 e5 5.d6 Black´s king is outside the square, so the bishop must let off the e5-pawn. Great!
So, with some discipline we are able to make the right choice here, 1.Kf3!, and save the game. Note, that in situations with little time on the clock, this decision-making process must be executed in a swift way. Thus it makes sense to practice this kind of calculation exercises a lot.
Here is a similar case, based on a recent game of Peter Svidler.
Black to play. Evidently, we are in trouble, that c-pawn is about to queen. Not many candidate moves to speak of! Suddenly we spot a nice tactic: 1...Rxc7 2.Rxe8+ Kf7! and we recover the piece next move. But...horror! If we put ourselves in White´s place, we soon find 3.Re7+!, a typical simplification ploy in endgames. After 3...Rxe7 4.Nxe7 Kxe7 5.Ke2 Kd6 6.Kd3 Kc5 7.g4! White imposes a deadly zugzwang. So...time to go back. Any other candidate moves? Well, i) checks, ii) captures and iii) threats should always be considered in the context of finding candidate moves.
Thus we should check 1...Rf7+. White replies 2.Kg1. No mate on e1 of course, since the e8 rook is pinned. What to do? Let us check the Rxc7 idea again! After 2...Rxc7 3.Rxe8+ Kf7, things have changed with the king on g1: if now 4.Re7+?? Rxe7 5.Nxe7 Kxe7 6.Kf2 Kd6 7.Ke2 Kc5 8.Kd3 h5! it is White who suffers the zugzwang, and loses. Thus 4.Re4 is better, but after 4...Kxg6 5.Rxd4 Rc6 (or 5...Kf6) Black has reasonable chances for a draw despite the minus pawn, thanks to his active pieces.
Summing up: in tough positions we should be attentive to any existing candidate moves, such as 1...Rf7+! in this case. They might make a huge difference.
A final example, from the recent Georgian Championship:
White has a lot of activity in exchange for the minus pawn. We immediately spot the threat Rd7, and after a short while, we also notice Nd6 attacking e8 while threatening a fork on f5. Not so many options here for Black. 1...Qxe6 looks very bad due to 2.Nxg5 with a double attack at e6 and h7. As for the knight on a6, it won´t be able to stop 2.Nd6. More sensible seems 1...Rad8, parrying both threats and bringing the last piece to the game. No more thinking needed here?
One more look! After 1...Rad8, it turns out that White could insist on his 2nd threat by 2.Nd6! intending 2...Rxd6 3.Rxd6 Qxd6 4.Qxe8. Any other candidate move? Well, since the rook was hanging on e8 in the previous line, why not give 1...Red8 a try. Apparently, White can´t win material anymore. He could perhaps try 2.Nd6 anyway, with the idea 2...Rxd6 3.Rxd6 Qxd6 4.Qf7+ followed by 5.Qxf6+. Anyway, there are not so many pieces left on the board anymore, besides we don´t have any better option (= rejection logics), so we must go this way. White might certainly have a forced draw there, or even a win!, but this is not really relevant for our decision.
As you can see, the Georgian GM correctly opted for 1...Red8!, and in fact he even managed to win the game in the end, after that White rejected a draw by repetition. It is interesting to notice, that 2.h4!? might have been White´s most critical try, instead of the tempting 2.Nd6. No matter what happens then, 1...Red8 remains Black´s only playable move here.
I hope you have enjoyed these examples. In order to enhance your training, here are few related exercises.
1. Black to play. (2-3 moves)
--------------------------------
2. Black to play. (2-3 moves)
--------------------------------
3. Black to play. (2-3 moves)
--------------------------------
