The Longest Plan In Chess

Chess is well known as a game of strategy. This is why it is prominently pictured in all kinds of television commercials that have to showcase long-term planning (e.g. insurance, banking, etc.).

It is a popular myth among people who don't play chess that grandmasters routinely anticipate what's going to happen in their games dozens moves in advance!

So it begs the question: how far can chess players see the future of the game?

A long, logical plan implemented in a game always gives great aesthetic pleasure.

Read the next paragraph from "Zurich 1953," the iconic book by Bronstein, after White's move 20.b3:

"There exists a widespread, and therefore dangerous, misconception that the win is automatic once you are a pawn ahead. As a matter of fact, Black's chief advantage in this position lies not so much in his plus pawn, which he is still far from exploiting, as in his control of most of the center squares: d4, d5, c5, f4 and f5.

White has his counterchances: a queenside pawn majority and the d-file. How many similar games have been drawn because of inexact play! Smyslov, however, manages such endings with an iron hand. His plan may be divided into the following phases:

1. The immediate exchange of one rook, leaving the other to restrain White's queenside pawns and attack the c- and e-pawns.

2. Deflecting White's rook to the h-file by the threat to create an outside passed pawn, and then occupying the d-file with his own rook.

3. Advancing the g-pawn to g4, undermining the e-pawn's support, which is the f3-pawn.

4. Tying up White's pieces by attacking the e-pawn.

5. Sending his king in to pick off the weak pawns.

As we shall see, a simple winning plan -- for a Smyslov, naturally!"

It took Black about 20 moves to implement his plan!

Now let's follow another world champion: Alexander Alekhine. Here is what he writes after his move 15...fxe5:

"The play in this ending is by no means so simple as it appears -- especially for White. Black's plan, which will prove completely successful, consists of the following parts: (1) exchange one pair of rooks; (2) transfer the king to e6 where, being defended by the e-pawn, it can prevent the invasion at d7 by the remaining White rook; (3) operating with the rook on the open g-file and advancing the h-pawn, force the opening of the h-file; (4) after this White's king, and possibly his bishop, will be tied to the defense of h1 and h2 against invasion by the rook; (5) Black meanwhile, by advancing his a- and b-pawns, will sooner or later also open one of the files on the Q-side; (6) since at this point his king will still be on the opposite wing, White will be unable to prevent the invasion of the first or second rank by the black rook. It must be admitted that, had White from the very beginning realized that there was a real danger of him losing this ending, by careful defense he might have been able to save the game. But what happened was that Black played according to a definite plan, whereas White played only with the conviction that the game was bound to end in a draw. The result was an instructive series of typical patterns and stratagems, much more useful to students of the game than the so-called 'brilliances' of short one-sided games."

It took about 30 moves for Black to finish his plan!

But there is a proverbial fly in the ointment. IM Mark Dvoretsky rightfully mentions that White could easily prevent the execution of Black's plan on the very first stage by playing either 17. f4! and if 17...e4?! then 18. f5! or even 16.f4! right away.

Moreover, Dvoretsky expressed an opinion that the whole monumental plan was created by Alekhine after the game, when he annotated the game for his book, which is some sort of chess reverse engineering.

It seems like GM Bronstein agrees with him:

"Due to Tarrasch an idea grew up that is still prevalent nowadays, the idea that there are the so-called logical games in which one side carries out a logical plan from beginning to end rather like a theorem in geometry. I do not think that there are such games between opponents of the same strength and the annotator who gives that impressions is often the winner of the game who makes out that what happened is what he wanted to happen."

It is difficult to disagree with the logical conclusion of both Bronstein and Dvoretsky.  Does it mean that planning 20-30 moves in advance is absolutely impossible? Let's take a look at the next amazing game:

When the game was played, some kibitzers were wondering why Nakamura was basically doing nothing by moving his bishop back and forth starting from move 42. He did it for 31 moves in a row! If it feels like deja vu, check here.

Then, out of blue, Nakamura broke his own fortress by playing 115...f5.  Did he change his mind and suddenly start playing for a win?  

Fortunately, Nakamura provided an answer himself after the match: "Rybka (the engine that assisted Nakamura during the match) was saying that in the second game if I did not go f5, White goes Re3 and g4 with an advantage of 0.9. Clearly, Stockfish is better."

Now it becomes clear what was going on in this mammoth game. In the position on the diagram after Black's move 40...Qd6, the only possible plan for White was the g4 break. And so Stockfish started to slowly prepare this break using the peculiar computer logic.  

When on move 115 Black realized that the g4 break is finally unstoppable, he was forced to play the preemptive 115...f5, which eventually led to his defeat. 

If Stockfish started its plan on move 40 and was about to finish it on move 115, then we have a record breaking 75-move-long plan. 

Is it chess of the future?