Thursday, August 18, 2011

Peak Rating

Bishop vs Knight takes me to the summit


I'm afraid this game was not particularly remarkable except for raising my rating to 1800, a relative peak for the last 13 years. However, it does show some typical endgame ideas and techniques. The game is Ken Ho - Jesse Nicholas, from the recently concluded Reubens-Landey qualifier for the Boylston Chess Club championship. The two time controls were 40 moves/90 minutes, then 20 minutes for the rest of game.

After 40...h6 (concluding the first time control)

I think at this point I had about 30 minutes left for the rest of the game, and Jesse had perhaps about 50. Given the relative paucity of pieces on the board, I chose to invest a significant chunk of time before my next move to get a deeper feeling of my chances for a win, while reserving enough time for myself to minimize the chance of flagging.

I felt generally that if the minor pieces were exchanged, at worst I should draw. However, I also considered that I wouldn't have enough time to calculate a win by force in assorted king and pawn endings, if one actually existed.

I looked at 41 Kh5.  If 41 ... Ng8? 42 Bh7 will win Black's h6 pawn, with a likely win. Alternatively, if 41 ... Nf5 42 Bxf5 Kxf5 43 Kxh6 Kf6 and White wins with his extra pawn and the power of zugzwang, e.g., 44 e4 a6 45 h5, when Black must give way.  So I concluded that Black would play 41 ... Kg7 with a draw by repetition. The White king would dance between g4 and h5, the Black king between f6 and g7.

41 Kh5 Kg7 42 Kg4 Kf6 43 Kh5 I offered a draw with this move, which was declined.

43 ... Kg7 44 Kg4


Here Jesse took some time on his next move. Since he had declined my earlier draw offer, I prepared to claim one myself by 3-time repetition, and I was reviewing the position to make sure such a claim would be valid. Naturally, I expected 44...Kf6, whereupon I intended to stop the clock and indicate that I would play 45 Kh5.
(I just looked it up, and apparently the correct procedure is to first write down your move [to which you are committed, if your claim is denied], then stop the clock and make your claim. Hi, Matt!  :-)  )

However, Jesse quite unexpectedly played:

44 ... Ng6??

Black has yielded the f5 entry square to White's king, which at best appears to lead to a most unfavorable exchange of White's h-pawn for Black's e-pawn.

45 Kf5 Nxh4+ 46 Kxe5

The White bishop is already covering all of the Black knight's possible squares.

46 ... Ng6+

If 47 Bxg6 Kxg6 48 Kd6. If Black's king doesn't move from g6, White's e-pawn will queen with check even though it will queen one move later than Black's h-pawn will. In such a case, I expected that I should either be able to win the queen with a skewer check OR exchange queens down on h1 with Black's king recapturing, in either case winning easily.

However, I think I felt that line might have somewhat more murky considerations if Black moves his king on his 48th move, as I would have expected Jesse to do. So I considered that I had a more straightforward win by staying in a bishop vs knight ending. White enjoys the traditional bishop advantage: long distance control of both wings.

General considerations: Avoid hanging game-losing material, choose my king's squares to avoid damaging knight forks. More concretely: queen my a-pawn, or force the knight to sacrifice itself for it, while my bishop guards my remaining e-pawn and prevents Black's h-pawn from queening (the bishop can sacrifice itself to do so, if needed).

47 Kd6 Nf4 48 e3 Ne2 49 Kc6 Nc3 50 Bf3

The bishop again controls many of the Black knight's possible squares.

50 ... Kf6 51 Kb7 1-0


Well, now that I'm at 1800 again (well, on September 1st, anyway), I'm ready for next year's Reubens-Landey....

2 comments:

Anonymous said...

Congratulations Ken! I'm on 1798 myself, the third time in the past three years I've been over 1780 without going over (though I managed it in my youth).

Rick Massimo

Ken Ho said...

Rick,

It's been so long, I can't quite tell for sure, but...
I think the air is better up here.
;-)

Come on up!

"To 1800...and beyond!"