Downing A GM On The Internet

     Today we'll only be going back a stone's throw in time...to 1996. In those days you could by the Zircon II and Jade III chess computers that allowed you to, as per the advertisement, "Train with a real USCF 2241 rated master anytime!" Both models offered 56 levels of play and had LCD lights that gave hints, had training functions, position evaluations and more. The Zircon sold for $185 and the Jade (the travel version) for $159. They ran on either 4 or 6 batteries, but for $16 you could get an adapter. 

     The concept of a machine that plays chess actually dates back centuries. We have all read about Wolfgang von Kempelen's contraption in the 1760s that had a human hidden inside who was pulling levers to move, but it wasn't until the 1950s that computers could play a crude game. By the 1980s computers could conduct basic searches and examine several thousands positions per second and by 1987 they were beginning to compete against humans. 

     It was in 1987 that Deep Thought defeated British IM David Levy. On February 10, 1996, at a convention center in Philadelphia a six game match was held between World Champion Garry Kasparov and Deep Blue, IBM's 2,800-pound supercomputer. 

     Commentator GM Maurice Ashley didn't believe the computer had a chance. Nor did Kasparov who had derided the idea that a computer could defeat a GM before the year 2000. Deep Blue was capable of evaluating 100 million positions per second, but while Kasparov saw fewer, they were better and he won 4–2. A rematch was played in New York City in 1997 and was won by Deep Blue by a score of 3.5-2.5. 

     The USCF's best selling clock was the battery operated Master Quartz (analog) clock that sold for about $85. A bargain plastic clock cost about $60 and if you wanted to go high tech, you could get a digital clock with new and dramatic breakthrough technology for $100. It had a delay option and besides regular time controls (no delay) there were 96 preset time limits. The clock also boasted the option of a sound alarm, or a red light and an on-screen flag to show when the time expired plus there was an optional move counter. A basic set, the now ubiquitous regulation size plastic set with a 3.75" King, weighted and balanced, was $7.95. 

     Internet chess was starting to become popular. If you had a computer and a sufficiently fast modem you could play rated games at any time control with players around the world 24 hours a day 7 days a week. 

     In those days Google didn’t exist and in January 1996 there were only 100,000 websites. The browser of choice was Netscape Navigator. By 1996, Yahoo, created in 1995, was well on its way to becoming one of the major Web portals. In 1996, the search engine Altavista became the exclusive provider for Yahoo. People used dial-up internet connections and they were beginning to switch from 640×480 to 800×600 screen resolutions. 

     In 1996 Chess.Net hosted its first Internet tournament, a $1,000 event with cash prizes and no entry fee that even drew some GMs. The time control was Game 15. 

     Players had to have a "handle" and GMs had to use their last name. Thus the following game pitted Nimrod, a 2232 rated US Master named Rene Phillips against the new US Open Champion Gabriel Schwartzman (born 1976), a Romanian-born American GM. 

     Schwartzman moved to Florida and played in his first tournament at the age of 4 and by the age of 12 he had the FIDE Master title. Three years later, he became an IM and in November 1993, at the age of 17, he was awarded the GM title. 

     In 1988 he finished 2nd behind Judit Polgar in the World Under 12 Championship. He won the 1996 US Open and was the winner of the Internet World Student Championship. He started the world's first interactive chess school in 1996, the Internet Chess Academy. Schwartzman decided to retire from chess in 2000 to focus on a career in business. His USCF rating stands at 2617. 

     In spite of his impressive credentials, in the following game Schwartzman got dinged by Phillips.

[Event "First Internet Tmt, G15"] [Site "?"] [Date "1996.??.??"] [Round "?"] [White "NM Renee Phillips"] [Black "GM Gabriel Schwartzman"] [Result "1-0"] [ECO "A80"] [Annotator "Stockfish 15"] [PlyCount "73"] [EventDate "1996.??.??"] [SourceVersionDate "2022.05.09"] {Dutch Defense} 1. d4 f5 {The Dutch has never been a main line defense and is rarely seen in high level competition. In the past is was sometimes used by Morphy, Alekhine, Najdorf ,Larsen and Nakamura. It's most notable adherent was Botvinnik and Bronstein who also played it in their 1951 World Championship match.} 2. Nf3 { In practice 2.g3 yields somewhat better results.} Nf6 3. Bf4 {A rare sideline. Again, 3.g3 gives better results, but 3.Nf3 is also good.} d6 4. e3 e6 5. Nbd2 Nc6 6. c3 Qe7 (6... Be7 7. Be2 O-O 8. h3 Kh8 9. Qc2 Nd7 {The position is equal, but in Hanel,E-Hechl,G Vienna 1994, black blundered on his next move and lost quickly.}) 7. Bc4 h6 8. h3 e5 {Oddly, this is the same mistake Hechi made in the above quoted game! Here, too, it results in black getting an inferior position. That's odd because it strikes at the center while the seemingly weakening 8...g5 is actually the best move.} (8... g5 9. Bh2 Bd7 10. b4 Nd8 11. a4 Nf7 {and black has completely equalized.}) 9. Bh2 g5 10. Qe2 {This is too passive and allows black to justify his last move.} (10. dxe5 Nxe5 11. Nxe5 dxe5 {and white is better.}) 10... e4 11. Ng1 Be6 {One square too far as will be seen in the note to white's reply. After 11...Bd7 there is no d5 fork.} 12. O-O-O (12. d5 {was his best option. After} Bxd5 13. Bxd5 Nxd5 14. Qb5 Nb6 15. Qxf5) 12... d5 {While hardly a mistake, this eventually causes black some problems because it allows the B on h2 to become active. He probably should have castled.} 13. Bb3 {This is a poor position for the B, so 13.Bb5 was better.} a5 {This is also questionable. Again, black should castle Q-side.} 14. Ba4 Kf7 {This is hard to understand. Simply 14... Bg7 was good.} 15. Nb3 { White is aiming for Q-side play, but the more straightforward attack on the center with 15.f3 was stronger.} (15. f3 Kg8 (15... exf3 16. Ngxf3 Kg8 17. Qb5 Rb8 18. Ne5 Nxe5 19. Bxe5 Bg7 20. Qxa5 {white is better.}) 16. Rf1 Rh7 17. fxe4 Nxe4 18. Kb1 {and white has the more active position.}) 15... Na7 {The N is out of play here so retreating to d8 would have been better, but Schwartzman is setting a little trap.} 16. Nc5 (16. Nxa5 {This is a horrible mistake!} Nc8 {and the N is trapped.} 17. Nxb7 Rxa4 18. b3 Ra7 19. Nc5 {White's 2 Ps for the piece are not sufficient compensation.}) 16... c6 17. Be5 b6 {This loses the game. He could have put up stouter resistance with 17...Bg7} 18. Nxe6 Qxe6 19. g4 {Black's K-side is going to get ripped apart.} b5 {Black has a forlorn hope of getting counterplay on the Q-side} (19... fxg4 20. hxg4 Bd6 21. f4 gxf4 22. exf4 {White has a decisive advantage.}) 20. gxf5 Qxf5 21. Bc2 b4 22. c4 Bg7 23. f3 Kg6 (23... Qe6 24. Rf1 dxc4 25. fxe4 Ke7 26. Nf3 {and white is clearly much better.}) 24. Rf1 {Black must now prevent fxe4.} Qe6 25. fxe4 Nxe4 26. Nf3 Rhf8 27. Bxg7 Kxg7 28. Ne5 Rf6 29. Bxe4 dxe4 30. Rxf6 Qxf6 31. Rf1 Qe6 32. Rf7+ Kg8 33. Qf2 Qxh3 {This allows a mate in 4, but he was dead lost in any case.} 34. Rg7+ {[%mdl 512]} Kh8 35. Rh7+ $3 {[%mdl 576] Decoy} Kxh7 36. Qf7+ Kh8 37. Ng6# {A solid performance against a GM} 1-0