Kasparov vs. The World Team

This month I will publish the first in a series on the World Team vs World Champion Kasparov Internet match and for those who want to peak ahead, there is a site with many of my posts saved. http://www.maths.uq.edu.au/~rwb/kas/bmcc.html . Most initial predictions were jokes that the world team was 2050. Four talented youths; GM Etienne Bacrot of France, Irina Krush and Florin Felecan of the United States and Elisabeth Pahtz of Germany were to guide the world. I did not really find an interest in the promotion and the idea of doing anything as a promotion for free went against my instinct as a chess pro. When the game became a line in my repertoire I began to follow the moves. I was attracted to post on the bulletin board by what I felt was a Kasparov blunder, 16. a4 and the insistence of someone that 16. Ne4 Nxe4 Qxe4 Qb3 wins a pawn. I responded to this person that things were not so simple and black had many dynamic chances, besides the fact if Kasparov wants to play on the flanks we had better act in the center or be beat! I wrote a post to Irina Krush about Ne4! The next day fellow columnists SmartChess Online were using Ne4 as a main line. I expected the other guy to get credit for Ne4, even with bad analysis as SmartChess had done a good job of keeping track of who said what at the Microsoft bulletin board. Later Irina had this to say about a comment I made concerning how integral Ne4 was to their strategy and their independent discovery "Actually it wasn't. I treated it as just another candidate move, did a little analysis - there was some in a FAQ - but didn't think really seriously about it until Alex Khalifman e-mailed me his (correct) opinion about some of my 16...Nd4 analysis - specifically one of the R v B + 2P endgames I was looking at. I switched to 16...Ne4 basically on his advice - analyzed it with Gigi (Kacheishvili) a bit." This extra effort by SmartChess was one of two unexpected events and many people posted Irina must be getting her analysis from Karpov. Let me quote the Microsoft analyst and US Women's champion Irina Krush: "No-one at SCO ever spoke to or contacted Anatoly about the game. I don't know why people keep bringing Karpov up in respect to this game, so I have to keep on denying it." We know their list of dedicated workers included : Irina Krush, Paul Hodges, GMs Kacheishvili, and Henley who put in many hours of effort and World Junior Champion Ilya Guryevich (Junior on ICC) gave Irina some opinions about the middlegame around move 27-29. A second unexpected factor was the organization of our computers at the Computer Chess Team Site. These two vast resources; an integrated BBS where legendary pros like GM Suttles and amateurs could analyze together and computers with 3000 Internet ratings at never seen before depth, provided Kasparov more than he ever imagined.

This natural line 16...Ne4 became the main line, and my threat of mate on h2 with 20...Be5! was the final consolidation of a maneuver that gave us many if not too many choices. The move before Be5 was discovered, GM Bacrot posted he hoped some other analyst could find a defense. The world team's novelty of Qe6 was fully justified due in large part to GM Khalifman's stunning 18...f5. Kasparov had to begin anew fighting for a win. This outline I present here is my favorite as I was the first person to suggest Bf4 was a Kasparov type move and it clearly demonstrates that we were in no way surprised from what looks like a dramatic pawn sacrifice. Lazy or complacent, may be a better word as to why the World Team ignored the Bf4 warning and the Kh1 move of HiArcs. Somehow despite this, we reacted well from what appears to be a dubious position after 33...b4 and the World Team forced several demonstrable draws to date. All these evaluations are subject to microscopic discoveries and rumor has it Kasparov thinks the alternative suggested here Bxg3 loses. I disagree with that and think that any refutation is also subject to the microscope.

Next issue I will delve deeper into Microsoft's helplessness as the potential for Mac users and other non Microsoft platforms to multiple vote began to dominate talk and the resources of the BBS. The users Microsoft claim to support became helpless spectators as claims of stuffing became more outlandish and counter claims stated multiple voting was absolutely not a factor. The fact the insecure voting procedure became an issue at all was distracting enough in itself during a very tense struggle. The game spiraled out of control during a period where there was often no Grandmaster advice at all and the fact Kasparov could change his move over the course of his time, but the World Team analysts could not change theirs once voting started, became a huge factor. This left Kasparov with a 3 to 1 time edge and on move 58 his move was late on top of that, although it seems he sent it on time. SmartChess posted that Mr. Kasparov many times moved early, but it only takes one slip especially in a postal style game.. Microsoft promised to post the move analyzed by Irina Krush Qf5, very early after voting but claimed it was not possible for them to do it when they received the e mails over 3 and a half hours later after 4:00! Somehow they posted late for Bacrot the next day. In Irina's absence the two Grandmaster's advice weighed heavily and both inexplicably missed the 3 move win forced in any line. (Qg1+, Qf2+, Kf6) This is particularly unfortunate as the BBS had posted this winning idea, that was reshaped by IM Regan, days in advance of the vote as had the GM Chess School. I understand trying to match wits with Kasparov, but no matter what your strength the world team resources of BBS, Computer chess team and the GM School site could help. GM Danny King, the paid GM commentator said there was a lively debate about Qe4/Qf5 when there was really only one nut spamming an old post compared with titled players in total agreement and the Computer chess team and my page showing anything but Qf5 was a disaster. GM Bacrot outright recommended Qe4 and the game was over by less than 5 % of the vote. At least the public can take heart in that their vote on Qe4/Qf5 was split, while the official Microsoft GM's were unanimous for the game losing blunder! An argument against lifetime titles if I ever heard one!

The fact is all world team members also had lives to lead and no one can be held accountable for not taking enough time based on these other needs. Kasparov postponed a planned title defense with Anand. The most concentrated Grandmaster effort came from the Russian GM Chess School web site http://www.gmchess.spb.ru/ and even their leader Alexander Khalifman had to take time out to win the FIDE title. Although it took my treasured Bf4 set up to counter American superstar Gata Kamsky in the Queen's Gambit, he fully deserved that title. Gata has been an avid Bf4 fan his whole pro career and this was a fascinating match even if he was not at peak playing form.

Net giveaways have never made much sense to me, but I am glad I went with the flow, I had record vendor income on the Internet Chess Club: http://www.chessclub.com and have a new students from Europe for the first time.

Reuters has reported " NEW YORK (Reuters) - A number of chess players battling against world champion Garry Kasparov on a Microsoft Web site Monday accused the company of mishandling a move that denied them at least a highly respectable draw.

MSNBC the sponsoring web page, published a story about this event that contains a few apparently contrary statements : ``We have the ability to track the votes and we've done that any time there's been any kind of allegation of vote stuffing,'' Diane McDade of Microsoft said, but then they seem to have relied on guesstimates to rule out Qe1: according to program manager Eddie Ranchigoda: "This was the first time a move not recommended got a high percentage," he said. "That was a red flag." If they possessed this technical wizardry why would they have to resort to putting their thumb in the wind? A mother on the alt.gymnastics board asked me about a harasser claiming to be a chess master that was threatening members of the US Olympics team and I remember people having to wait for him to be online to do any real tracking. If Microsoft can track a person from a local phone call or school telnet to a provider and back out to their site, I think they would be bragging about it. There are claims to be able to do this to incoming hits on NT servers, this was posted from their help files to the BBS. Their inability to do this to the Mac and other platforms like Linux, may explain why they had to suspend MAC voting for days.

"YOU LIVE BY THE INTERNET, you die by the Internet," said Microsoft spokeswoman Diane McDade. "This was only possible because of the technology, but unfortunately in this case the technology caused a problem." (Microsoft is a partner in MSNBC.).

Some believe it was caused by people not answering their e-mail and not keeping their web page promises. SmartChess Online posted their e-mails with the time stamps, but Microsoft did not reciprocate. That alone would have removed much of the controversy. Rumor has it the world champ plans to release lines showing no computer glitch was responsible in the outcome, meaning a win after Qf5. We shall see, he sent his new hire MIG to the BBS to advertise and inquire about our Qf5 busts, so at this point, perhaps he isn't 100% certain himself. From the data we had, Qf5 was a draw and I agree with that. The GM School and I discussed one line that they improved which cleared the highest scoring computer line and the other lines had reasonable +110 evaluations which would be expected in a pawn down draw. Of course black is worse and there was no guarantee of the best move being played with children or adults acting like children, stuffing votes. The best analogy I have come up with is that the World Team became a 2850 turtle, as long as we could move straight ahead with a goal in sight we had armor from our computers and progress was easy. Such abrupt changes like ...b5 even if playable were too much of a swing for our system to bear. There is a big difference in overnight computer analysis and week long, walked out lines. Mig even teased of a rematch with Kasparov as black. It would be interesting, but I would want real security that the integrity of one man one vote was protected. If it can't be then we aren't ready for this format yet.

Here was Kasparov's statement on the MSN web page:

'At move 10 I congratulated The World on a great opening novelty, and to that I would like to add it was a phenomenal middlegame, a crazy human endgame - six passed pawns! - and now we have a computerised (sic) queen and pawn ending of mathematical precision. Within our game we have covered all the elements of modern chess. I thank The World for a great fight.

By the way, The World made several inaccuracies in the queen and pawn ending, it was very difficult to play, but after 56...d5 (probably the final decisive mistake) 57 Qd4+, the game was already lost. Although 58...Qf5 was more stubborn than 58...Qe4, it too would have led to defeat as I will show later.

I look forward to our post mortem when the game concludes.'

Garry Kasparov, 20th October, London.

Analyst Elisabeth Pahtz had an interesting observation that may explain the unprecedented popularity of this game: "What other sport could you play on the Internet with just computers and e-mail?" For those intrigued by this format Fritz 6 has taken black against another group of youths at http://fritzgegendeutschland.de/ . Its in German, I couldn't find the English version, but deduced "spiel" was the link to see the board. A C99 main line Ruy Lopez up to Na5 Bc2. This should be an easy translation exercise: 1.e4 e5 2.Sf3 Sc6 3.Lb5 a6 4.La4 Sf6 5.0-0 Le7 6.Te1 b5 7.Lb3 d6 8.c3 0-0 9.h3 Sa5

Here is my 8/19/99 Outline, move 30 seems to be the ½ way point and is a good place to leave the analysis:

Does Crafty agree w/ Zark re Qxf5(!)/b4 in g4

BMcC Zarkov likes Bxg3 1 billion nodes

spider-tl061.proxy.aol.com Thu Aug 19 21:05:51

Quick new outline, I will compare new developments and put out my final version. I think this was very up to date at 5 am. Anyone see any changes or reasons Zark's Bg3 line is no good, pls let me know. Best viewed at:


Kasparov proclaims game "the most complicated and analyzed game in the history of chess," Wall Street Journal 8/12/99 "It's quite annoying from my point of view because if I do a second-quality move in this game, there's just no way out." Does he mean 16. a4?! The game so far: [Site "Microsoft Gaming Zone"] [White "Kasparov, G."] [Black "The World"] [ECO "B52"] [EventDate "1999.??.??"] 1. e4 c5 2. Nf3 d6 3. Bb5+ Bd7 4. Bxd7+ Qxd7 5. c4 Nc6 6. Nc3 Nf6 7. O-O g6 8. d4 cxd4 9. Nxd4 Bg7 10. Nde2 Qe6 {N (Krush) The "World Variation"} 11. Nd5 Qxe4 12. Nc7+ Kd7 13. Nxa8 Qxc4 14. Nb6+ axb6 15. Nc3 Ra8 {(Speelman)} 16. a4 16... Ne4 17. Nxe4 17... Qxe4 18. Qb3 18... f5 (GM School - Khalifman)} 19. Bg5 19... Qb4 {(Jason Van Eaton)} 20. Qf7 Be5 {(Brian McCarthy)} 21. h3 Rxa4 22. Rxa4 Qxa4 23. Qxh7 Bxb2 24. Qxg6 Qe4 25. Qf7 áBd4 26. Qb3 f4 {{Yasha}}27. Qf7 Be5 28. h4 b5 29. h5 Qc4 (above designations as given by analyst US Women's champion Irina Krush: www.smartchess.com):

World Annoys Kasparov! World Bluffs Kasparov!?

Outline 8/19/99:

Predicting 31. Qxe6 Score of Predictions so far 15-1 (Qf5?!) Recommending: 30.... Qe6 31. Qe6+ Ke6 32. g3/4 fg3 33. fg3 Bg3 Garry has tried to sidestep our mountain of ...e6 analysis, but did he do anything else? He transposes to a line I had as recommended from the middle of last week till yesterday. Clearly the answer lies in white's g pawn. Can he play g4 instead of trading? Can he play g3 or g4 once he does play Qxe6? The best way to look at this fascinating ending is by a concept introduced to me by one of my favorite Russian authors: Eugene Znosko-Borovsky, related squares. We have forcing sets of moves that can happen in many different sequences, and GK is a master of seeing the subtle difference. I believe that Garri may have considered Qf7 a harmless prod and that he could retreat to other lines without losing a tempo if needed, but our plans of e6 and Qg4 spoil that. Gary needs a real plan to finish the game, whatever the result, and we need to be as ready as possible

Developments! I just can't convince my computer Bxg3 isn't good after Qxe6 Kex6 32 g3 fg 33 fg, I ran it out to a billion nodes and it liked Bxg3, so I did it again, the result, pv h6 Be5 h7 Bg7 Kg2 b4 Bh6 Bh8 Rf8 Bd4 Be3 Be5 Bf4 Bg7 Kf3 b3 -7 [Zarkov] 1.17 billion nodes. +. The latest try is b4 and usually the 1st thing to look at in all lines, however if both moves are causing decent white positions, we need to think about it very carefully. I will verify this and other new developments for my final Qe6 outline. Zarkov's quick take on the computer chess teams expected line yesterday is in the middle of the other beasts 30.Qf5+ Qe6 31. Qe6+ Ke6 32. g3/4 fg3 33. fg3 b4 34. g4 b3 35. Bf4 and 35...Bd4+ 36.Kg2 b2 37.Kf3 b5 38.Rh1 d5 39.h6 Kf7 40.h7 Bh8.

My current recommendation is not based on any secret knowledge, just trying to direct attention to all playable options. There are many new ideas after Qf5+ Qe6, g4!?, Qxe6 Kxe6 g3 and also ideas of Bc1 if we try f3. All moves have been looked at, but none to the 20 move level we had ...e6. The world has strengthened it defense to Qd3 in the initial line suggested by the Computer Chess Club: 25.Qf7 Bd4 Logray 26. Qb3 26...f4 27. Qf7 Be5 28. h4 b5 29. h5 Qc4 30. Qf5+ Qe6 31. Qd3 Qc4 32. Rd1 Qxd3 33. Rxd3 e6 5/14 +0.17 35+ hours CM6K suggested by Krush to refute 25.Qf7; Crafty rates end pos. +1.33 @ 12ply. now 31...Qg4 and Nd4 are both good plans scoring well on the CC Club. By far our biggest pressing need is deciding whether to play ..b4 or Bxg3 in the g3 line.

MAIN LINE: Garry has entered the old main line, first appearing at the computer chess club site on 8/10/99. After having played out the pawn race in a mock game, I feel confident in black's position. The computer evaluations have been steadily improving since the key Bd4 juncture. IM Orlov feels black is better and said if Kasparov doesn't make a draw he will be playing for the loss! To underline just how ...e6 fit in that exact position Qf8-f5, ...e6 played now is rated at +350! We are left with the pawn race. He repeated Qf7 to fix our weakness and tame our bishop. We have responded by sealing off his queen and bishop so we can try to queen our pawn and discourage any queen trades that bring bad technical endings! Did we succeed? It looks that way do far!

A) 30, Qf5+ Qe6 31.Qg6 b4 32.Rb1 b5 33.Qd3 Qg4 34.Qxb5 Qxg5 35.Qb7+ Ke6 36.Qxc6 Qxh6 37.Rxb4 Qd1+ 38.Kh2 Qh5+ 39.Kg1 full 18 0.00 >20h rb crafty 16.15

B) 30.Qf5+ Qe6 31. Qd3 (Qb1 can also be met by the Qg4-f3 plan) 31. ... Qg4 (It seems Qc4 is not needed but the CC Club suggests b4 as a winning attempt! see B3) 32. Qxb5 f3 33. Qxb7+ Kd8 34. g3 Bxg3 35. Qb6+ Kd7 36. Qb7+ Kd8 14 +0.00 (draw) Aaron Crafty 16.13 SmartFAQ 8/11 Line E5a3) Pawn race looks fine.

B1) 32. Qh3 Qxh3 33. gxh3 Ke6 =

B2) 32. Qxb5 f3 33. Qxb7+ Kd8 34. Be7+!? Nxe7 35. Qxf3 Qxf3 36 gxf3 Nf5 37 Re1 Kd7 38. Kf1 Bf6 (not Nd4 Rxe5! General Moe) 39.f4 Bc3 40.Re4 d5 41.Ra4 Nd4 42.Ra3 Nb5 43.Rb3 Kc6 and Zarkov +58 after 14 million nodes but it is hard to see white winning with his split pawns.

B3) (30.Qf5+ Qe6 31.Qd3) b4 32.Rc1 jb 32...Nd4 33. f3 Qf7 34. Rc4 Ne6 35. Bh4 Qxh5 36. Be1 Bc3 37. Bxc3 bxc3 38. Qxc3 b5 17 -0.63 8h crafty 16.15/solaris SmartFAQ 8/11 Line E5a32; crafty rates end pos. -0.61 @ 11ply

C) 30. Qf5 Qe6 31. g4!? This line is the subject of much debate, I will give Zarkov's take. Nd4 has been hot and cold, Qxf5 risks a possible f6 (Ross Amman) queening, but seems the best until an exact plan is found. Crafty agrees (30.Qf5+ Qe6 31.g4 rb 31...Qxf5 32. gxf5 Nd4 33. Kg2 f3+ 34. Kh3 Nxf5 35. Rb1 e6 36. h6 Nd4 37. Kg4 Kc6 38. Be3 18 -0.12 13h crafty 16.15 ) 31...Qxf5 32.gxf5 Nd4 33.Kg2 Nxf5 34.Re1 Bc3 35.Rb1 Nd4 36.Bxf4 e6 37.h6 b4 Zarkov at 80 million nodes -12, however Zarkov flirts with +08 for a while. This line needs to be clarified, but does not seem dangerous.

C1) Past 90million nodes, Zarkov likes white again, but still close to even. 31...Qxf5 32.gxf5 Nd4 33.Kg2 Nxf5 34.Re1 Bc3 35.Rb1 Nd4 36.h6 Ne6 37.Rxb5 Nxg5 38.Rxg5 e6 +8 97 million nodes.

C2) 30.Qf5+ Qe6 31.g4 b4 32.Kg2 b3 33.h6 b2 (FAQ one line played out on the BBS is : 34.Qxe6+ Kxe6 35.h7 f3+ 36. Kxf3 Nd4+ 37. Ke3 Nc2+ 38. Kd3 Na3 39. f4 Bh8 40. Re1+ Kd7 41. Rxe7+ Kc6 42. Re1 b1=Q+ 43. Rxb1 Nxb1 44. f5 Nc3 45. f6 Na4! 46. f7 Nc5+ 47. Kd2 Ne6 48. Be7 Kd7! draw. "DBC"

D) 30.Qf5+ Qe6 31.Qxe6 Kxe6 32.g4 fxg3 and transposes to below is the current recommendation.

E) 30.Qf5+ Qe6 31.Qxe6 Kxe6 32.g3 fxg3 33.fxg3 Bxg3 (Can we really do this?) 34.h6 Be5 35.h7 Bg7 36.Rf8 b4 37. Rf3 Ne5 38. Rg3 Bh8 (what?! rb) full 14 -0.06 21min crafty 16.15 ("can white win!? we can always play 33...b4, of course" rb )

E1) (30.Qf5+ Qe6 31.Qxe6 Kxe6 32.g3 fxg3 33.fxg3 Bxg3 34.h6 Be5 37.h7 Bg7 38.Rf8 b4 )37.h8=Q Bxh8 38.Rxh8 38...Kd5 39. Kf2 b3 40. Bc1 e5 41. Rh1 b5 42. Rd1+ Kc5 43. Be3+ Kb4 44. Bc1 Nd4 45. Bb2 Kc4 46. Rc1+ Kd5 47.Rd1 Kc5 48. Bxd4+ exd4 49. Rc1+ Kd5 50. Kf3 full 18 -0.08 13h crafty 16.13 (who knows... rb)

E2) (30.Qf5+ Qe6 31.Qxe6+ Kxe6 32. g3 fxg3 33. fxg3) b4 34. Rb1 Bxg3 35.Bd2 Kf5 36.Bxb4 Kg5 37.Kg2 Be5 38.Bd2+ Kxh5 39.Rxb7 d5 -20 CC Club

E2a) (30.Qf5+ Qe6 31.Qxe6+ Kxe6 32.g3 fxg3 33.fxg3 b4) 34.g4 (rb) b3 35.Bf4 Bc3 (35...Bd4+ 36. Kg2 b2 37. g5 Nb4 38. h6 Nd3 39. Kf3 Kf5 40. Bd2 Nc1 41. Rxc1 bxc1=Q 42. Bxc1 Kg6 +0.23 13h crafty ) 36. g5 Nd4 37.g6 Ne2+ 38. Kh1 b2 39. h6 b1=Q 40. Rxb1 Nxf4 41. g7 Bxg7 42. hxg7 Kf7 43. Rxb7 Kxg7 44. Rxe7+ Kf6 full 16 +0.55 13h crafty 16.15 tablebases would probably solve that last position; TB says draw -jb

E2a1) (30.Qf5+ Qe6 31.Qxe6+ Kxe6 32.g3 fxg3 33.fxg3 b4 34. Bf4 Bc3) 35.h6 b3 (12/13 +0.20 3hours -ongoing analysis CM6000 Pentium II 333Mhz CM6k first gave 33. .. Bxg3 34.h6 Be5 35.h7 Bh8 36.Rc8 Ne5 Bd2 Nc4 Bb4+ Kf1 (+0.08) (then it changed line Raimondo D'Ambrosio)

E2b) (30.Qf5+ Qe6 31.Qxe6+ Kxe6 32.g3 fxg3 33.fxg3 b4 34.Bf4 ) Bd4+35.Kg2 b3 36. g4 b2 37. g5 Nb4 38. Bd2 Nd3 39. Rb1 Ne5 40. g6 Nc4 41. Bb4 Kf5 42. Rh1 Ne3+ 43. Kg3 Nc2 18 +0.40 7h crafty 16.15/solaris w/TB end pos. -0.37 @ 14ply jb

E2b1) (30.Qf5+ Qe6 31.Qxe6+ Kxe6 32.g3 fxg3 33.fxg3 b4 34.Bf4 jb 34...Bd4+ 35. Kg2 b3 36. g4 ) Kd5 37. g5 e5 38. Bd2 e4 39. Rd1 Kc4 16 +0.74 7h crafty 16.15/solaris see below

E2b2) 30.Qf5+ Qe6 31. Qe6+ Ke6 32. g3/4 fg3 33. fg3 b4 34. g4 b3 35. Bf4 Bd4+ 36.Kg2 Kd5 37.h6 b2 38.Kf3 Bc3 39.g5 e5 40.Be3 Ne7 41.Rd1+ Kc4 42.Rb1 Ng6 43.Ke4 b5 17 +0.58 12h crafty 16.15 "personally I find it hard to believe that black is holding this " rb.

Conclusion: Garri has left the most analyzed line in the most analyzed game, only to go to one of the next most analyzed continuation. He probably will try a g pawn maneuver as opposed to a queen retreat. We need a complete line vs 32 g3 and the outlook remains positive.

(Computer Chess Club) http://www.egroups.com/wdb?method=reportRows&listname=comp utergang&tbl =1&sortBy=1&sortDir=down&start_at=0&prntRpt=1

Thanks to everyone on the Computer Chess Team!


Here is the remainder of the game score to date from move 30:

30. Qf5+ Qe6 31. Qxe6 Kxe6 32. g3 fg3 33. fg3 b4 (OmniBob/KarlJuhnke) 34. Bf4! Bd4+ 35. Kh1! (recommended by 1 computer: HiArcs) b3 36. g4 Kd5 37. g5 e6 38. h6 Ne7 39. Rd1 e5 40. Be3 Kc4 41.Bxd4 exd4 42. Kg2 b2 43.Kf3 Kc3 44. h7 Ng6 45. Ke4 Kc2 46. Rh1 d3 47. Kf5 b1(Q) 48. Rxb1 KxR 49. Kxg6 d2 50. h8Q d1Q 51. Qh7 b5 52. Kf6 Kb2 53. Qh2+ Ka1 54. Qf4 b4! (McCarthy/ Krush/PKCrafty) 55. Qxb4 Qf3+ 56. Kg7 d5 57. Qd4+ Kb1 58. g6 Qe4?? 59. Qg1+( here Qe1 in protest won the vote but was over ruled by Microsoft!, ) Kb2 60. Qf2+ Kc1 61. Kf6 d4 62. g7 and 1-0 is certain and the World Team appropriately resigned. Kasparov announced a forced mate. Here is a line posted by Peter Karrer: 62. g7 Qc6+ 63. Kg5 Qd5+ 64. Qf5 Qg2+ 65. Kh6+-. And Crafty agrees: 62..Qc6 63. Kg5: depth=15 +6.55 63. ... Qd5+ 64. Qf5 Qg2+ 65. Kh6 Qc6+ 66. Qg6 Qh1+ 67. Qh5 Qc6+ 68. Kh7 Qc2+ 69. Kh8 d3 70. Qh1+ Qd1 71. Qh6+ Kc2 72. Qc6+ Kd2 73. g8=Q Qa1+ 74. Qg7 Qxg7+ 75. Kxg7 Ke3 76. Qe6+ Kd4 Nodes: 607705774 NPS: 10911 Time: 15:28:13.15

Addendum: Since the game's end there have been several developments. Chronologically, people have shown my original idea 21...Rh8 may be the best move after all. Securing the base of the pawn chain was the human touch to a line computers will find eventually. I switched to Rxa4, the move played and one GM Suttles said you should play only if all other moves lost and then voted ...f4, but Rh8 was the move that "sold " Be5!. I became involved in a thread with a DK from the United Kingdom over Rxa4 and he has since identified himself as David Knopfler of the rock band Dire Straits! He humbly proclaimed himself a zero rated player, but was one of the most adept contributors with his computer.

The first wave of 33...Bxg3! refutations have failed miserably and Kasparov tried to back up his claim the World team was lost anyway with 58...Qf5 as well as Qe4, even going so far as to suggest not calling in Qf5 was a ploy: This is a quote from Kasparov's interview on MSNBC:

"I can't rule out, that if I am right, then the scandal about the unsent move [58...Qf5 by Irina Krush] could well be orchestrated. Indeed, ten hours is more than enough to send a move. After all one could have phoned. The fact that the move was not sent can be explained only by the absence

of the desire to forward it, because communication with Microsoft's technical staff was very well organized."

I still don't understand how or why Microsoft posted that Irina's move would be "posted shortly" at about an hour into voting, and then never followed up. They were under no mandate to allow a late move, but once volunteered it is hard to explain why it was never posted. However conspiracy theory aside, it seems Kasparov has a point. This becomes all the more critical because IM Ken Regan has posted a worldwide press release that refutes Kasparov's claim of always winning. From what I have seen, Regan's analysis is sound and most has been computer verified. I will cover the specific tactics when the moves arise and by then a more convincing picture on these critical variations may appear. Here is the critical line forcing a zugzwang and transposing to a line I had posted weeks before with Qb8-Qf4 killing the needed tempo: 58. ... Qf5 59. Kh6 Qe6 60. Qd1+ Kb2 (OK, this is as good as my 60. Qg1+.) 61. Qd2+ Kb1 62. Qd4 Ka2 63. Kg5 Qe7+ 64. Qf6 Qe3+ 65. Qf4 Qg1+ ( 65...Qc3) 66. Kf6 Qb6+ (66. Kh5 67. Kf7 Qb7+ (for maybe why not 67. Kg7, see note 4.) 68. Kg8 Qc8+ 69. Qf8 Qe6+ 70. Kg7 Qe5+ (this and other moves can lead to "G"). And here is his improvement: 1. g7 d4 2. Qa8+ Kb2 3. Qb7+ Ka1 (Club Kasparov gives only ...Kc1) 4. Kf7 Qf5+ 5. Ke7 Qg5+ ( Here 5. Ke8 must be answered by ...Qe6+ as I think ...Qg6+ doesn't work after 6. Qf7. Now one of Black's prayers is 6. Ke8 d3, which may be what Peter Karrer gave, but what I posted about was the Q-side run: ) 6. Kd6 Qg6+! 7. Kc5 Qc2+! (what I gave was the benighted ...Qg5+? 7. Kb4! +-) 8. Kb6 Qg6+ 9. Ka5 =!. He asked "and now maybe 9...d3 is possible?" Moments later I posted a Crafty line (Crafty is available for free as Professor Hyatt's University of Alabama Birmingham site: ftp://ftp.cis.uab.edu/pub/hyatt/) that is 0.00 on my end game table base. He has posted his outline for the draw on the Internet: http://www.cse.buffalo.edu/~regan/chess/GK-ROW/ .The theoretical ball is clearly in Club Kasparov's camp at this point.

In his chat room interview World champion Kasparov made a stunningly honest disclosure he admitted reading the World team strategy board explaining "Host Chris_MSNBC says: Did you come frequently read our comments in the BBS? Host Garry_Kasparov says: Of course I used it to my advantage to look around and follow the discussion on MSN.COM about the game" !! He also gave the team credit for a game at the highest level: " I think that the world deserves to make a draw..."

Finally, I started assigning my version of contributions by the major players of the WT and in my Suggestions for Irina Krush, SmartChess ran a few of the ideas by Anatoly Karpov who disagreed with Kasparov as to the best move at 29. Here is an edited for clarity recap of the thread and an exclusive angle on this discussion between two world champions. My comments on Irina's 4 best contributions are in {} {Here we have a definite K-K difference of opinion! I never disagreed with Suttles and maybe no clear line will be established, (as to 29...Qc4 the move I recommended along with Smartchess vs Qe2, a move chess legend Canadian GM Duncan Suttles analyzed to a draw and Kasparov called an equal game.) but I didn't think this line deserved to be buried on Kasparov's say so alone.

...the choices to me are (of course)a) Qe6, } SmartChess Online wrote: 10...Qe6! would be our choice. Karpov arrived in NY today, admitted he had not looked at the game much after seeing 10...Qe6 and the next few moves reported on the Internet. He described 10...Qe6 as a brilliant concept, and at first glance liked 15...b5. See below for the only other two comments I was able to get out of him, before he had to leave to meet people to go to a hockey game.

{ b) d5 sac vs Qc8 in D, move 51 Qc8 instead of Qh7 a +200 computer position} Irina didn't think it was a big deal as practically any other move than 51.Qh7 (or maybe 51.Qh5) wasn't going to be played. There are some key lines in the 51...Ka1 defense she found which are important contributions to the viability of 51...Ka1 (the "true" draw as she has called it).

{ c) 52...Kc1 (the way AVO tells the story, I now understand why she said the move was established here, however the saved posts don't tell that story. If it was AVO's discovery or hers, I never heard since this was during the credit free part of the game.)} The draw concept with 52...Kc1 was AvO - he had the guts of it worked out before White even played 51.Qh7, when 51...b5 had also been under consideration. IK was the first to be convinced from SCO - she didn't think 52...Kb2 was much good at all, and gave me a bit of a hard time for spending so much time on 52...Ka1, and 52...Kb2. { d) 29...Qc4} I think this move originated might have originated from Spiriev (!) {he was a Hungarian 2200 FIDE player who became frustrated with the system and lashed out at several BBS members, SCO and myself included.} and who then wished to be totally disassociated from it, although it may be pre-dated by an analysis file I have of Henley's - the idea may be around the same time, although Ron thinks someone else mentioned it before him - no matter. What is interesting is that Karpov says that this move preserving the bishop is superior to 29...Qe2, in his opinion. The queen endgame after 50...d1=Q should be a draw - Karpov. Obviously, before he leaves, I will try and get some more! "

Quite a lot of data in a few questions by SCO chief Hodges!

Hosting by WebRing.