Who do you believe - me or the computer?

After I lost the game which was the subject of Tuesday's column - and during it - I assumed that it had followed the trajectory my games habitually follow, to wit that a good opening had led to a promising middlegame, which I had then thrown away, as I usually do, in a tactical skirmish.

I was surprised, therefore, to look at the game with Rybka and find that in the early middlegame, before the aforesaid tactical skirmish (in which I did indeed throw the game away) it liked White's position a good deal more than it liked mine.

The game began 1.d4 d5 2.Nf3 Nf6 3.e3 c6 4.Bd3 Bg4 5.c4 e6 6.Qb3 dxc4!


7.Bxc4 (7.Qxb7 loses to the two captures followed, after Qxa8, by ...Qc7) 7...Bxf3 8.gxf3 Qb6 9.Rg1 Nbd7 10.e4 Qxb3 11.axb3 g6 12.Nc3 Bg7 13.Be3


13...O-O 14.f4 a6 (perilously close to draughts) 15.h4 Rfd8 16.Ra5 b5 17.Bd3 Bf8 18.Ke2 Bb4 19.Ra2 c5 20.dxc5?! reaching the position we interrogated earlier this week.

As I say, I was rather expecting Rybka to show, if anything, a small advantage to Black for most of the play given above and I was surprised - almost offended - to discover the opposite. Rybka likes White's position a lot: for example, in the diagram position after 13.Be3, it swiftly shows a White edge of more than +0.80, and at 17-ply it's still as high as +0.77, indicating that Black is really up against it.

Now I'm not tied to computer assessments. There's not much point in observing that, say, Fritz gives +0.47 for White, if you don't know why this is: nor is that +0.47 line necessarily better than another one which is given as +0.15, just because the computer thinks it is. Programs possess algorithms, not the mind of God.

Nevertheless, I've learned to respect these assessments, especially if you can't in the short term find a line which causes the computer to revise them. Certainly I'd be unhappy to play against that 0.77 in a correspondence game, even if I couldn't see how it was going to increase, let alone that it would lead to a win for White.

Still, during and immediately after the game I held a completely different view and I suspect I would take the same approach to a similar position if it occurred in another game. I thought White was jumbled up, lacking mobility, obstructed by his own pawns, while Black, notwithstanding his lack of space, had a more harmonious set-up, with time to organise a break (as he duly did). I don't suppose my play was anywhere close to optimal - nor my opponent's - but was I really so wrong in my assessment?

Perhaps. Perhaps I was misled by my relative familiarity with Slav positions, and hence felt comfortable even though strictly speaking and objectively, I should not have.

Perhaps I have a certain stylistic preference for certain sorts of positions, playing for a break against a possibly overloaded centre, which seem to me more attractive than they really are. (I recall watching live this encounter - also in the Slav - and seeing Black get pasted despite White's adoption of a set-up that I found, impressionistically, unappealing.)

Perhaps I was misled by the chessplayer's natural optimism that derives from the fact that we make our moves (the process Rowson terms Narrative) and hence we naturally see ourselves as in command of the game - while the opponents merely respond, as best they can, to our initiative.

Or, perhaps, I was just suffering from the typical response "he's gone wrong so I must be better" that occurs when the opponent seems to have missed a tactic (as he had in the first diagram, after 6...dxc4!). This is not at all uncommon in my games, as, even though I say so myself, my knowledge of the opening outlasts that of my opponent rather more often than not. In this instance, I'd had the position after 6.Qb3 before (I'm sure the game is in a comments box somewhere, though I cannot find it now) and had discovered 6...dxc4 with the computer's help when looking through the game afterwards. The opponent was the first to be surprised.

Hence I often think I'm better... because I am better? Except that perhaps I'm not. Perhaps, even when the opponent is surprised, I'm only comfortably equal. I do, after all, have Black. Or perhaps I was better - or I was equal - for a bit, but I subsequently let it slip. It can be hard, without a blunder to crudely make the point, to understand that your position has deteriorated. But of course it happens all the time.

But there is another possibility. Which is that my judgement is right and the computer's is wrong and that the position was, in fact, perfectly OK for Black. I don't think I was so far out in my assessment, not as far as the computer thinks I was at any rate. So I appeal to the court of public opinion, unreliable though it may be. Which side of the board would you prefer? And how far does the computer's assessment affect your choice?