This is me from last Friday, icing my face to help recover from losing the last 4 pieces of wisdom I had. While a few days of applesauce and no exercise or any kind of physical activity certainly did not sound appealing to me, in particular at a time when I would have liked to have played the Chicago Open or the Best of the West Open, I did manage to make the most of it. Luckily my pain medications did not make me too drowsy to study chess, so with nothing else to do I went all out. While I can't disclose all the work I did, I will share a couple insights I came up with-- newly acquired wisdom, if you will. The first came at a very critical moment of a game I was watching live:
After 88 long and grueling moves, this position was reached. Grischuk had been nursing a 2 pawn advantage for some time, but he had been having a hard time making progress because he only had one pawn that could make threats (although the doubled fpawns would later win the game). Here white is faced with a very critical decision- after many moves of maneuvering, he now has a chance to concretely change the nature of the position... in 2 different ways! While I cannot criticize Grischuk for choosing the line of he play he did choose because it led to a won position, while watching the game live I was definitely puzzled by his choice.
Ra5!!?. This move obviously forces the rooks off the board, but I was not entirely sure whether the position could be won at first glance. However, Grischuk did manage to demonstrate a very effective winning plan. White's problem is that he only has one key threat in the position- advancing the dpawn. This is very easily stopped by the black king, who can plop his large rear end on the d7 square and kill any hopes white has of promoting the pawn. So in order to make progress, white needs to stretch the defenses. In order to stretch the defenses, white needs to find a second weakness, which in this case is obvious- the f6 pawn, which is conveniently fixed on a dark square. White's winning plan is as follows:
1. Maneuver the bishop to a square that controls d6
2. Advance d5
At this point, it is clear that the black king cannot approach the d5 pawn, so the white king can leave on his journey without having to worry about losing his lone passer
3. Bring the king around to the kingside, specifically h5
At this point, we reach move 107. Black is left with a serious dilemma. If he allows the white king into g6, the f6 pawn is dead meat, and then with 3 white passers black can basically resign. So, black naturally plays Kf7 to prevent the infiltration of the white king. But here we see exactly what white has done- he has created 2 potential winning plans, both of which require the black king's attention. One of them is Kg6, and the other is d6-d7. Unfortunately for black, the black king does not have a wide enough range to cover both of these avenues of attack, so white has successfully stretched the defenses enough that black cannot respond to all of his threats. The rest was just textbook.
While this was a very satisfying endgame, I am still curious about Grischuk's decision to exchange the rooks. The winning plan does not seem that easy to spot and evaluate correctly- involving a king march all the way to the kingside, and then eventually all the way back- and it seemed after Grischuk had exchanged the rooks it took him some time to find the right plan. With this in mind, I think my preferred 89th move is simpler:
Was Ba5 a better move than Ra5? Probably not, both moves win, but I think Ba5 was a much simpler solution to white's problems. On a mostly unrelated note, I am very glad to see that both Ba5! and Ra5!!? do not make the engines top 3 choices, while in my opinion they are the only 2 moves that will gaurantee a win (the computer prefers to maintain the status quo with useless moves that make no changes to a position that has been largely unchanged for the last 20 moves anyway). This tells me that while the computer will always outcalculate a grandmaster, it won't always understand chess better, and there is a lot of charm and mystery left in our game for humans to solve and understand even if Houdini would clean Carlsen's clock in a 10 game match. In any case, the final lesson, or piece of wisdom, to be gleaned is:
In the endgame, if the main defender of the attacker's winning plan is the defensive King, stretching the defenses to force the king to cover 2 squares separated by at least 2 files is a very effective way to make progress.
And lastly, I'd like to thank both Nakamuras for their efforts- GM Nakamura for resiliently defending a tough endgame and forcing Grischuk to demonstrate the very instructional winning plan in a very conceptual bishop endgame, and Dr. Nakamura, who did an excellent job extracting my wisdom teeth with very little pain or complications.
Best of Luck to All,
GM Sam Shankland