Computers and Correspondence Chess
an article by Jim
Brennan
I would like to shed some light on the use of computers and their strengths and weaknesses
at correspondence chess. I have played with and without computer assistance and will
use the games I have played to illustrate the benefits and drawbacks of computer usage. I will also comment on the touchy issue of morality concerning computer assistance.
According to some testing done at IECG the top commercially available computers are playing around 2200-2300 rating at correspondence chess. This means there is a great resource available for whoever wants to use it, almost like having a master to help
analyze your positions. As well there are the huge database programs readily available so that you can play master strength openings up to point as well. Conceivably a person using a computer will play even higher than the computer on its own. It require
s a person with a better understanding of the game than the computer has but this is probably the case for B level and up.
Does this mean a plethora of master strength correspondence players? Well
probably not since to improve on a computers play a great deal of
analysis and study is necessary and this is really what seperates the best
from the mediocre without computers. Also there is the point that the stronger players who use a computer should be able to play that much better than weaker ones who do. Perhaps
overall playing strength will increase and ratings will decrease to compensate as more and more players
use computers. I think the general acceptance and use of computers is inevitable though and so the best players will still be the best except everyone will have more resources and chess theory will pe pushed at an accelerated pace.
When playing a game with computer assistance I would basically analyse the position with a normal set and then on the computer trying out different moves I thought playable making sure I wasn't missing any short range tactics.The computer can be a
very good tool in this regard, if you have a fair understanding of the
game, since you can efectively narrow its search and use it to consider
likely positions much further than it normally could on its own. The
drawback of this is that it is entirely possible to be come too engrossed in what the computers evaluation of a position is and lose sight of a deeper plan that is the human strength of the equation.
In the following game moves such as 17. e6 and 18.... h6 are moves that, regardless of their merit,a computer could not come up with because the material justification for the moves is beyond computer move horizons. Being able to confidently play move
s, such as these, that a computer evaluates as bad is essential if you are going to use computer assistance. Good experience with playing against computers or analysing the games they play is also a must because in certain positions they are phenomenally
strong and others, such as closed positions, they are very weak and the computers value must be judged accordingly. A good example of these strengths and weaknesses is the Kasparov - Deepblue match. Game 1 saw the computer at its best in a tactical fight
with lots of targets the last game saw the computer at its worse in a closed position where it got pushed off the board.
Ian Rout (Aus)- Jim Brennan(Can)
1.d4 Nf6 2. c4 c5 3. d5 b5 4. cxb5 a6 5. Nc3 axb5 6. e4 b4 7. Nb5
(this initiates the Zaitsev variation of the Benko gambit) d6 8. Bf4
Nbd7 ( this is an old line played in the late eighties. I beleive g5 is the current theory move to allow Nxe4 after Bxg5)
9. Nf3 Ba6 10. e5 Qa5 11. a4 bxa3+( this move may not be good. A few games in my database have this move and all games black lost.I however played a better move in Bc6 on move 16 instead of a queen move which loses quickly.) 12. Bd2 Qb6 13. Rxa3
Nxd5 14. Bc4 Bb7
15. Rxa8+ Bxa8 16. Qa4 Bc6 17. e6 ( this is really a great move as well as being gutsy I realy didn't consider it much as I didn't figure white could give up another pawn but this moves freezes the black kingside which never completely recovers) 1
7...fxe6 18.0-0 h6 ( this was the hardest decision of the game to play h6 as it furthers weakens the light squares and is played with the realization that the b and rook won't move for a long time.The idea is simply to stop the f3 knight from reaching
e6 via g5 as my analysis and instinct told me would be deadly)19. Ba5 Qa6 ( I almost played Qb7 which loses to a nice attack on the light squares after Bxd5 followed by Qg4. On a6 it attacks the a5 bishop which stops the same idea)
20. Re1 Kf7 21. Rxe6 N7b6 22. Ne5+(diagram1)

I found the surrounding of the b5 knight t be quite remarkable. Whites play has been excellent up to this point in my opinion) 22....Kxe6 ( I felt I had to capture the impudent rook but considered N7f6 as well where white would retreat the rook and
play b4 next with advantage) 23. Bxd5+ ( but here with Nc7+ after extensive analysis I beleive white has a won game.The main line goes 23. Nc7+ Kxe5 24.Bc3+ Nxc3 25.QxQ a possible continuation is ...Nxc4 26.Qxc4 Ne4 27.Qe6+ Kd4

White keeps a crushingly better position as the black king rook and bishop are spectating while the king is out on a walk. Don't send refutations untill you have analyzed the position after QxQ for upwards of 12
hrs. I wouldn't claim I should have lost unless I was sure! Its seems appropriate that Zaitsevs B5 knight finally breaks out of the quadrupal pin to win. White doesn't have enough compesation after Bxd5+ as the game shows) 23...Nxd5 24. Nxc6 Kf7 25. Nd
8+ Kg8 26. Qa2 Kh7 27. Ne6 Qxb5 28. Qxd5 Qxb2 (it is still dangerous for Black the queen covers key squares from b2) 29. g4 g6 30. h4 Qa1+ 31. Kg2 Qxa532. h5 Rg8 33. hxg6+ Kh8 34. g5 Rxg6 35. f4 Rxe6. 0-1 ( the f8 bishop never moved! not recc
omended strategy)
This game the computer allowed me a bit of false security in choosing the opening variation after bxa e.p. +, it never had a clue as to how precarious the black position became around move 23 and it didn't recognize the winning line I cited either.
Also several times during the game moves suggested by the computer I reguarded as losing outright. The computer was still a big help help though since it allowed me to check the pertinent lines during the very tactical middle of the game.This shows the d
ouble edged nature of using a computer for assistance.
I also played a correspondence match against a computer. I requested this match to get a better idea of my strength versus a computer.
This match was therefore played without computer assistance.The first game is an example of how not to play against a computer.The computer is the Tasc R30 2.5 stand alone, it is rated 2357 on the SSDF list and has an estimated strength of 2550 USCF for
over the board chess I beleive the strongest stand alone computer available. The computer was evaluating approximately 22,000,000 (22 million) positions before each move. This information was witheld untill well into the match as it was part of an ongoing
study by IECG which as since lapsed due to their difficulties this
summer. I have to say I underestimated the machine at first untill I
realized I was losing the first game and really bore down after that in the second. Perhaps the knowledge of the stre
ngth of the computer would have helped me but that is no excuse.The computer plays excellently in this game and refutes whites line of play after Bh6. It is no slouch and I was soundly outplayed.
Jim Brennan ----- Computer 5
1. d4 d5 2. c4 dxc4 3. Nf3 Nf6 4. e3 e6 5. Bxc4 c5 6. Qe2 a6 7. O-O b5
8. Bb3 Bb7 9. Rd1 Nbd7 10. e4 (an interesting line of the Queens Gambit Accepted) cxd4 11. e5 Nd5 12. Rxd4 Be7 13. Rg4 Rg8 (this move surprised me and I thought it was just bad since the black king is kept in the center)
14. Bh6 (I have to say this wrong in retrospect because g5 is playable and the bishop is in danger of being trapped probably the plan Bd2 and Nc3 is best completing development before trying to exploit black's king position.During the game I though
t I had some advantage at this point but probably it is equal.) g5 15. h4 Nf4 16. Qe3 Bxf3 17. Qxf3 Ng6 18. Rxg5

(I played this missing the fact that 20...Nxh4 was possible because of Blac
k's 19th. White would have good compensation after a queen move instead. If white doesn't give up the exchange a pawn will be lost. White's position is difficult in any case.) Rc8!(this inbetween move throws a wrench in whites plan since the queen
no longer attacks the rook) 19. Nc3( if white instead plays Qh5 than Rc1+ ties white up to much)Bxg5
20. Bxg5 Nxh4 21. Bxh4 Qxh4 22. Ne4 ( white appears to have some compensation but it in reality it is minimal) Ke7 23. Nd6 Rcf8 24. Qb7 Rb8 25. Qc6
Nxe5 26. Nf5+ Kf6!( this forces the queens off) 27. Qxe6+ fxe6 28. Nxh4 Rbc8 29. Rd1 Rgd8 30. Re1 d4 31. g3 Rxh4 (0-1) A brutally accurate game by the computer in a type of position that is favorable to it.
Computer 5 - Jim Brennan
1. d4 Nf6 2. c4 g6 3. Nc3 Bg7 4. e4 d6 5. Nf3 O-O 6. Be2 e5 7. O-O Nc6 8. d5
8... Ne7 9. Nd2 a5 10. b3 ( this follows Karpov - Kasparov IV 17th match game up to the 16th move but here b3 was thought to have lost a tempo since b4 is played 3 moves later) 10.Rb1 c5 11. Rb1 Ne8 12. a3 f5 13. b4 axb4 14. axb4 b6
15. Qb3
( this position arose in Taimonov-Kavalek,Wijk aan Zee 1970 but with white to move! There black was equal in a few moves so with an extra move black should be at least equal now . It is
unusual that Karpov would play an inferior line 22 years later in a world championship match. It is also strange that this computer should play the same way, perhaps the opening book work was a little superficial) Nf6 16. Rd1( the computer starts
to show signs of not knowing what is going on) Bh6 17. Bd3 Nh5 18. Nf3 f4 ( I played this move based largely on computers' infamous handling of closed positions and realizing things were not so rosy in game ones more open position .I judged that my
bishop would be better than the white one even if passive so decied not to trade before f4, also a bishop trade on c1 speeds up whites play and does nothing for blacks. )19. Bb2 g5 20. h3 g4 21. hxg4 21. Bxg4 (objectively this position may be slig
htly better for black but considering white is a computer the closed position is strongly in favour of the human) 22. Be2 Nf6 ( the knight comes back to hold a defender to e4 and centralize) 23. Ra1(the computer grabs space on the queen side
but there is nothing happening there with no targets and no pawns to advance) Qd7 24. Ra3 Rxa3 25. Qxa3 25. Kh8 (this moves starts a reorganization of blacks pieces so as to pressure the white kingside each piece clears a square for another) 26
. bxc5 bxc5 27. Ra1 Rg8 28. Qa6 (this sets a trap with a threat of taking on e5 but the computer needs to think of defending the kingside) Ng6 29. Kf1 Bf8 30. Qb6 Qg7 31. Qb3 Qh6 (black has reorganized while white has accomplished nothing on th
e queenside) 32. Kg1
Bxf3 33. Bxf3 Nh4 34. Kf1 Ng4 35. Bxg4 Rxg4 36. Nd1 Nxg2
( blacks patient slow build up has won a pawn but the extra rook pawn is notoriously hard to promote) 37. Qf3 (now black could force a queen trade with 37...Qh1+ 38.Ke2 Qe1+ 39.Kd3 Qf1+ 40.Kd2 Rg3! 41.Qe2 QxQ but the ensuing ending is very hard
if not impossible to win.Whites pieces would be free to harass the black king and d6 pawn. With the queens still on h
owever whites pieces are needed to defend the king and black has the resources necessary to usher the hpawn home.) Rg8 38. Ke2 Nh4 39. Qh1 Qh5+ 40. Kd3 Nf3 41. Ra8 Qg4! ( I was happy with this multipurpose one square move. The queen breaks the pin
on the knight, clears the hfile for the pawn to advance, gets an iron grip
on the gfile and takes control of the key f3 square) 42. Nc3 h5 43.
Nb1(white is lost but a human would probably try Nb5 threatening the d6
pawn.Nb1 allows black to continue unmole
sted) h4 44. Nd2 Nxd2 (tempting was the idea Ng1-h3 but white can organize a blockade of sorts in this line)45. Kxd2 f3!( keeping the rook away from the hpawn, h3 right away would lose the pawn to Ra3) 46. Kd3 h3

47.Ra1 Qh5 (a blunder now would be Qg2 allowing Rg1! drawing after Qxg1 Qxh3+ Kg7 Qd7+ with a perpetual check) 48.Bc1 h2 49.Be3 Rg2 (the computer now feels some security perhaps in stoping the pawn, black i
s however effectively up a rook and pawn in this position as the white queen and the bishop can't move) 50.Rb1 Kg8 (getiing off the h file to allow the queen to go in for the kill .If the queen moves now white could play Rg1 forcing defence of the
hpawn) 51.Ra1 Qh7 ( a nice backward queen move preparing a transfer to the queenside) 52.Kc3 Qxe4 0-1(the computer presumably pushes some greater material loss off the horizon with the king move however the end will come quickly now since bl
ack will play e4 and activate the dormant bishop to go after the white king. White has been in near zugzwang for the last few moves with nothing constructive to do. All around probably one of my best games, accurate while steering the position well into t
he computers weakness. Thanks to Davd Griffith who lent his time and computer to make this match possible.
This match illustrates well the type of positions that favour and disfavour a computer. Game one the computer had lots of scope for its pieces and some good targets the only bad thing was its king position which I failed to exploit. Game two saw a closed
position where the computer tended to drift shuffling pieces and getting a lost position. These computer weaknesses and strengths are not revelations but it is interesting to note that they still exist even at the highest levels and knowledge of them is
essential for one using a computer. Strategys in matches or tournaments where computers are allowed or being used could be weighted one way or the other accordingly.
You contribute more to a game the less assistance of any type that you use but this has to be weighed with practicality and quality and judged according to your specific reasons for playing correspondence chess. Creation , accomplishment and competition
are key reasons to play correspondence chess and these are enhanced the more of oneself that is put into each game but the computer pushes these to a new level and those that want to compete on equal footing have to realize it.
Computers are here and they are being used.They necessarily must be
allowed in world class play because their use can't be stopped. We have to deal with the situation and realize that the computer, like it or not, is a part of correspondence chess, to think otherwise is delusional.
Is it moral to use computer assitance? This requires a definition of morality. I think morality is defined as what is acceptable to the individual as being appropriate and ethical
but, since there is a dependence on opinion, morality is a changing
entity. In a situation so dependent on opinion we can only inform
ourselves and make a choice. I beleive the morality of computer use has
necessarily come into being for correspondence chess. Having said that, there is still room for people to compete without computers but it has to be done with the realization that they might be playing with a handicap and tournaments or matches dissallow
ing computers are on the honor system alone.
We have to play with what suits our own values and morals being honest with ourselves . I beleive computers lead us closer to the truths of a position while still leaving ample room for human contribution
and should be regaurded as another tool that pushes competition to higher levels. I don't recomend using computer assistance to everyone and I think one should consider their stand alone strength against their computer when deciding if to use one. If you
don't think you can find better moves than your computer suggests by itself then you probably shouldn't use it since you won't get much out of the game. If you think you can play better strategically and positionally than a computer than use one.
If you think you are better tactically than today's computers you are mistaken.
An interesting and informative article related to this is
Computers In CC an article by Dr.Stephan Busemann an I.C.C.F. Grandmaster.
All games available at in our own N.S. Games collection
| NS Chess
|NS Chess News
|NS Players Forum
|Tournaments
|Young NS
| Links
|Chess FAQ
|Games
|Web Tools
|NS & CFC|