Saturday, November 24, 2012

The Customer Is Always Wrong

Well, that's what it feels like in the chess world, anyway. In the chess world, when the score between two opponents is grossly out of line with statistical expectations, the hapless underperformer is sometimes described as being the "customer" of the other player. And in the chess world, when it comes down to the last decision that determines a game's result, the customer is (almost) always wrong. The customer's hard-earned rating points are uncannily drawn toward bad choices (cue the old Saturday Night Live Bad Idea jeans fake commercial).

Before rejoining Boylston tournaments after an extended time away, I had played Walter "The Fighter" Driscoll twice, winning both games, but back then it seems that I had about a 50 point rating edge over him each time. When we met again over the board at Boylston in 2004, I achieved what I thought was a clear advantage, but lost my way and ended up feeling that my position was somewhat uncomfortable.

Ken Ho (1702) - Walter Driscoll (1829)
11 March 2004
After 45...g5

In the above position I played 46 hxg5 and offered a draw, which Walter declined. I think I considered that the most I could get would be a draw, and that I would likely be able to get one, but if he were to accept my offer that would save me effort and unnecessary worry. At an earlier stage of the game when things looked (at least to me) rather worse for Walter, he had offered a draw which I had declined, although I don't think that affected his decision to decline my offer.

After the further moves 46...hxg5 47 Nc2 Kf6 48 Nxb4 Ke5 49 Nd3+ Kd4 50 Ne1 g4 51 Kg3 Kc3:

I played 52 Be6 and again offered a draw, which Walter also declined.

However, after 52...Kd4 53 Bxf5 I "persuaded" him. White's task has always been to eliminate Black's pawns, with both his bishop and knight being expendable to achieve the draw.

So, based on having seen more than Walter had in the above endgame position, I leaped to the wildly irrational conclusion that my practical understanding of chess was greater than his, that I could win the next game (this was the Paramount double round-robin tournament), and lock him in as a long-term customer. Well, somebody was destined to be the customer, but it wasn't Walter. He subsequently handed me 5 losses in a row, and during that period, our ratings have been within ("a mere") 130 or so points of each other.

Now, I understand that Walter's 1800 "exceptional" rating floor from around 2007 originated from a great result at one of the big CCA tournaments, and as he has mentioned to me, he has yet to crack 1900. He's bounced off that rating floor multiple times, but I am no stranger to doing so. My "traditional" rating floor is 1700 (based on getting a 19xx rating), but the last time (okay, okay, the only time) I made it over 1900 was back in late 1991.

Here is my first deposit of rating points at WalterBank®:

Walter Driscoll (1829) - Ken Ho (1702)
30 April 2004
After 16 Be5

16...Bg7? 17 Bxg7 Kxg7 18 Qf4 Qd6 19 Qe3 Rad8 20 Be2 Qf6 21 Rh3 Rh8 22 g4 Bc8 23 Rhh1 Qe6 24 Qxe6 Bxe6 25 c4 c6 26 g5 Rd7 27 Rd2 Rhd8 28 Rhd1 f6 29 f4 fxg5 30 fxg5 Rf8 31 Rf1 Rxf1+ 32 Bxf1 Rd6 33 b3 Bf5 34 Bg2 e6 35 c5 bxc5 36 dxc5 Rxd2 37 Kxd2 Kf7 38 Bxc6 Ke7 39 b4 e5 40 Ke3 1-0

In the postmortem Walter said, "I thought you were going to play 16...Bxe5 17 dxe5 Qxd2+ 18 Rxd2 Rad8 with a draw.". The subsequent open-mouthed silence? Yeah, that was me, making the sound of a player utterly dumbfounded by his lack of vision.

My favored iOS chess app tChess Pro (How's that, dfan?) even says that after 16...Bxe5 17 dxe5 Qe6 18 Qc3 Qxa2 19 Qxc7 Bd5 20 c4, Black has about a third of a pawn advantage (probably insignificant given our skill levels, I suspect). (tChess Pro @ "2500" rating, 30 seconds for Analysis mode limit.)

Customers often achieve perfectly reasonable or even winning positions, but are ever dogged by cruel cosmic chess injustice.

"We seem to be made to suffer. It's our lot in life."
(or was that c3 Sicilian?)

Fortunately I found a customer of my own (Sorry, Larry). Although I started our series off with a typically blunderful loss on my part, I subsequently collected 5 wins and 2 draws, and Larry's rating was similarly within a roughly 100-point range of mine.

Larry Eldridge and I often feed off each other's long thinks. In this game, he "saw" my ~26 minute think on move 9, and "raised" me by taking ~49 minutes for his 10th move.

Ken Ho (1804) - Larry Eldridge (1902)
3 May 2012
Time control: 40/2 d5, G/50 d5
After 19...Bf7

20 Qf3?

I was obsessed with offering to give up a pawn with check. Ah, Capablanca-Tartakower, New York 1924, what folly you inspired in me.... I felt that with the subsequent activity of my pieces combined with the clock situation, even someone who loves pawns like I do could bear to allow Black to take my d4 pawn.

While preparing this blog entry I checked tChess Pro's suggested moves at several junctures, and was rather put out to discover that in my Capablanca-esque dreams I had missed the crushing 20 Rxf7!! Kxf7 21 Qb3+ Ke7 (21...Ke8 22 Bg6+ Kd7 [22...Ke7 23 Qf7+ Kd6 24 Bf4 mate] 23 Qxb7+ and White is clearly winning) 22 Bg5+ and White is clearly winning. So Black's previous move should be punctuated as 19...Bf7??

20...Qxd4+ 21 Be3 Qf6 22 Qe2 Qe5 23 Bf4 Qc5+ 24 Be3

24...Qe5 I declined Larry's draw offer. To reach move 40, I had about 26 minutes left, and he had fewer than 2. Practically speaking, it seemed likely enough that his lack of time could lead to a gross blunder (and with my greater remaining time, it was my intention that such a blunder would not come from me!).

25 Qf2 O-O-O 26 Bf5+ Be6 27 Bxe6+ Qxe6

28 Bxa7 The loss of this pawn introduces queenside weaknesses which I thought would make it easier to win.

28...Be7 29 Bb6 Rdf8 30 Qd4 Rxf1+ 31 Rxf1

I had been looking forward to Black's next move, thinking that my queen's access to the a-file and Black's b-pawn weakness would allow me to bring home the point.

31...Bf6 32 Qa4 Kd7 33 Qd1+

tChess Pro gives 33 Qa7 Qd5 34 Qxb7+ Ke6 35 Re1+ Kf6 36 b4 h4 37 Rf1+ Kg6 38 a4 hxg3, with White having a slightly more than 2 pawn advantage. I had more than 8 minutes to make move 40, but Larry was almost out of time.

I had certainly considered 33 Qa7 earlier, and at this time I don't remember why I didn't play it. I don't normally believe in trying to blitz opponents when their flag is about to fall -- I prefer they self-destruct on their own.

33...Ke7? The Blunder Clock sounds midnight.

34 Re1 Bd4+?

After 34...Be5 35 Bc5+ Kf7 36 Qf3+ Qf6 37 Qe3 Re8 38 Rf1 Bf4 39 Qb3+ Kg6 40 Rxf4 tChess Pro considers White has more than a 4 pawn advantage.

1-0 (time forfeit).

I welcome Larry to continue depositing his rating points here at:
New customers are always welcome.

No comments: