Monday, August 10, 2009

International Master Justin Sarkar's "Perfect Game"

I first met International Master Justin Sarkar at the November 2006 Kings Island tournament, where Sarkar tied for first place after defeating Grandmaster Zviad Izoria, a very strong player who outrated Sarkar by nearly 300 points at that time; such a difference means that Izoria would be expected to win roughly 75% of his games versus players of Sarkar's rating. After the tournament ended, IM Sarkar enthralled National Master Jerry Hanken and me with a fascinating analysis of his victory, including insights into how a top level player thinks during such an encounter. Hanken subsequently wrote about Sarkar for both Chess Life magazine and Chess Life Online.

Sarkar is a highly talented player but he has traversed a very challenging path to attain the IM title and you can learn about some of the details of his struggle with autism by clicking on the links in the previous paragraph. I spoke with Sarkar several times in the past few days at the 110th U.S. Open Chess Tournament. Sarkar said that originally he had not planned to play in the U.S. Open but then at the last moment he decided to jump into the fray. After taking byes for the first three rounds and reeling off five straight wins Sarkar needed just one more victory to tie for first place and earn a spot in the 2010 U.S. Championship. Sarkar is particularly proud of his eighth round triumph over Robert O'Donnell, a strong Expert from Michigan who for many years held a National Master level rating. Sarkar is soft spoken and modest about his achievements, so when he earnestly told me that he had played a "perfect game" versus O'Donnell and then lamented that most people don't understand how difficult it is to play a perfect game of chess I knew that I had to see the moves for myself:

IM Justin Sarkar - Robert O'Donnell [E32]
U.S. Open 8/8/09 (8)

1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.Qc2 0-0 5.e4 Bxc3+!? 5...d5 6.e5 Ne4 is the normal continuation. 6.bxc3 White has a slight edge now; he has obtained a massive pawn center without having to spend a tempo by playing a3. 6...d6 7.e5 dxe5 8.dxe5 Nfd7 9.Nf3 Nc6 10.Ba3 In the only other game I could find in this line, Sergey Beavenets defeated Viacheslav Ragozin, a famous Grandmaster who is renowned for his theoretical knowledge and who helped train World Champion Mikhail Botvinnik:

1. d4 Nf6 2. c4 e6 3. Nc3 Bb4 4. Qc2 Nc6 5. Nf3 O-O 6. e4 Bxc3+ 7. bxc3 d6 8.e5 dxe5 9. dxe5 Nd7 10. Ba3 Ne7 11. Rd1 h6 12. Qe4 Re8 13. Bd3 Nf8 14. O-O Rb8 15. Qg4 Bd7 16. Bc1 Nf5 17. Qh3 Ng6 18. Be4 Qe7 19. g4 Ba4 20. gxf5 exf5 21.Qxf5 Bxd1 22. Rxd1 Qe6 23. Qxe6 Rxe6 24. Rd7 Nxe5 25. Nxe5 Rxe5 26. f3 Rbe8 27.Rxc7 f5 28. Bd5+ Kh8 29. Bf4 Re1+ 30. Kg2 R8e2+ 31. Kh3 g5 32. Bd6 Rg1 33. Re7Rxe7 34. Bxe7 b6 35. c5 bxc5 36. Bxc5 Rd1 37. c4 a6 38. Bb7 Rd3 39. a4 1-0 (1935 Soviet Championship).

10...Ndxe5 11.Nxe5 Nxe5 12.Rd1 Qh4 13.Rd4 Qf6 14.Bxf8 Kxf8 15.Qxh7 Ng6 16.h4 e5 17.Rd2 e4 18.h5 Qxc3 19.hxg6 Qc1+ 20.Ke2 Qxc4+ 21.Kd1 Bg4+ 22.f3 Qa4+ 23.Ke1 e3 24.Rd3 Qb4+ 25.Ke2 Qb2+ 26.Kxe3 Re8+ 27.Kf4 Qe5+ 28.Kxg4 Qe6+ 29.Kg3 Qe5+ 30.f4 1-0


Note how Sarkar eschewed taking the f8 Rook on move 13 and instead gained a tempo by activating his Rook while harassing O'Donnell's Queen. O'Donnell could have offered stronger resistance at certain points but Sarkar is understandably pleased with this game because of how relentlessly and logically he played after O'Donnell's slight inaccuracy on move five; O'Donnell's rating places him above the 97th percentile of U.S. players, so defeating a player of that caliber is not nearly as easy as Sarkar made it seem to be.

Grandmaster Jesse Kraai beat Sarkar in the last round, thereby sharing first place with five others and dropping Sarkar to a tie for 18th place. Kraai's rating had plateaued between 2400 and 2500 for about eight years before recently jumping to nearly 2600. Kraai has described how he attained the Grandmaster title at a relatively late age (the 37 year old achieved that goal two years ago, but most GMs reach that level in their teens or twenties) by virtue of hard work and I have much respect for anyone who plays chess that well because I know how difficult it is to do that. However, I also sympathize with Sarkar's quite understandable feeling that his accomplishments in the face of adversity have not received their just due; in the Hanken article cited above, Sarkar notes that his great performance in the 2003 U.S. Championship did not attract much attention and on Sunday he told me that he had hoped to jump start his quest for the Grandmaster title by finishing the U.S. Open on a strong note. Sarkar's living situation and his state of mind make it difficult for him to study and prepare the way that most top flight players do, so he relies heavily on natural talent, trusting his instincts to help him figure things out over the board (as opposed to making the in depth pre-game preparation that is de rigueur at the IM and GM level). Sarkar has defeated many top GMs and certainly has the ability to attain that coveted title; hopefully he will produce more "perfect games" en route to achieving that goal.

1/9/10 Addendum:

IM Sarkar recently emailed me some further insights about his "perfect game" and I am happy to share his perspective with my readers; he feels that it is very important to understand that he not only played this game very accurately but that he did so very quickly. IM Sarkar writes, "I didn't exactly blitz out the game, though I was comfortably within the 3 minutes a move avg and like I said was even up over an hour on the clock. My 16th move (h4) was important to find and recognize over the board as the best plan in the position. I took one of my longer thinks on move 24 (Rd3) as I had a nearly equally promising continuation. The computer at first seemed liked the other, then with more time to think liked my move even better. Note that it's a bit scary, as it involves 'almost getting mated'. It's good I was able to accurately calculate this with confidence, even though I saw another 'less scary' way that I also felt was most certainly winning. And in general, when he gave me the knight on g6 for an attempted perpetual check or attack on my king, I had to accurately foresee the best way out of the series of checks."