Tuesday, November 13, 2007

Don't Analyse Unnecessary Tactics

Tom and I have been discussing tactics recently. I told him I was trying to eliminate tactical sequences from my games as much as possible – and he told me it was not possible to be a strong player without a thorough grounding in the tactical elements of the game.

He’s right of course but what I had in mind is encapsulated in this quote,

“ analysis is an error-prone activity. Overlooking one important finesse can completely change the result of the analysis. If it is possible to decide on your move on purely positional considerations then you should do so; it’s quicker and more reliable.”
John Nunn, Secrets of Practical Chess p.21 Gambit 1998

In short, Don’t Analyse Unnecessary Tactics (or DAUT).

For a long time I didn’t understand positional chess at all. If I wasn’t playing for a direct attack against my opponent’s king I felt uncomfortable and never really knew what to do. I just bluffed my way through games waiting for opportunities to win material through tactical tricks.

I remember distinctly the game that changed my mind. It was played in the Surrey League against a player who has been rated in the 150s for many years (and still is).

I'll put the game up to play through first then go through my thoughts during the game in some detail below.

I was White and the game opened,

1. e4 e6, 2. d4 d5, 3. Nc3 Nf6, 4. Bg5 Be7, 5. e5 Nfd7, 6. Bxe7 Qxe7, 7. f4 a6, 8. Nf3 c5, 9. Qd2 Nc6, 10. 0-0-0 b5, 11. dxc5 Nxc5, 12. Bd3 Bb7

I’m not sure castling queenside is the most accurate idea but I was playing for a direct attack, possibly with a Bxh7+ sacrifice or perhaps a pawn storm with h2-h4-h5-h6 and maybe hurling the g-pawn forward too. Unfortunately for me, Black has avoided this possibility through the simple expedient of leaving his king in the centre for the time being.

Not knowing how to proceed, I remembered the suggestion that if you can’t think what to do you should improve your worst placed piece (who originally said that btw?). The trouble was all my bits seemed to be on good squares already. Eventually, completely happy with everything else, I decided that my best bet would be to bounce my knight around to occupy d4.

13. Ne2 Nb4

A small piece of tactics – one of the very few in the entire game. Of course, if White takes the knight on b4 he loses his Queen to … Nxd3+. Not a bad trap to set in itself but the trouble is White can get out of it by playing a move he’d like to anyway and in getting his king of the semi-open c-file he shows Black has just lost time.

14. Kb1 Nbxd3, 15. cxd3

I was much happier now and at this point I realised I had a ready made plan to follow. I was going to stick a knight on d4, try to exchange all the heavy pieces along the open c-file then win an ending of knight against bad bishop. I also saw that if I left my pawn on d3 it would cover the important e4 and c4 squares, preventing Black’s knight getting into my position.

15. … Rc8, 16. Ned4 g6

I was already set on my plan so it took me a few moments to work out why Black played this move. He wants to prevent me playing f5 but as I’ve explained that wasn’t on my agenda at all. In any event, I really think Black should have castled then gone for counter play with … f6 at some point. He spends most of the game trying to dig himself into a solid bunker but he leaves himself without any active play at all – just the sort of position I most prefer.

17. Rc1 Kd7, 18. Rc2 Na4, 19. Rhc1 Rxc2, 20. Rxc2 Rc8, 21. Rxc8+ Kxc8, 22. Nb3 Kb8, 23. Qc2 h6, 24. g3

Evidently with his last move Black wants to move his queen to the c-file without allowing a knight to g5. He’s been, and continues to be, very compliant with my plans but if he hadn’t swapped rooks he’d have left me the open file all to myself.

At this point I wanted to swing my other knight to d4 but I didn’t care to give him any chance of messing me around on the kingside by playing his queen to h4. I knew I could always return to my long term plan because it was so firmly embedded in the fabric of the position. As a result I didn’t feel the need to rush things (a crime of which I’m often guilty).

24. … Qc7, 25. Qxc7+ Kxc7, 26. Nfd4 Kb6, 27. Kc2 Nc5

So I’d achieved the first part of my plan and I remember sitting there wondering how I was going to force him to exchange knights when he suddenly played this move. Perhaps he thought the knight on the rim was dim but I was very happy getting to my knight v bishop ending.

Incidentally, the game’s second tactical line is here.

28. Nxc5 Kxc5, 29. Nb3+ Kb4??, 30. a3+ Ka4, 31. Nc5+ wins the bishop on b7 so Black can’t penetrate with his King and I can exchange knights in comfort.

28. Nxc5 Kxc5, 29. Nb3+ Kb6, 30. Kc3 a5, 31. a3 Bc6, 32. Nd4 Bd7

Black played this move instantly which surprised me somewhat. Surely he’s not worried about me exchanging my knight for his horrible bishop (something I’d never have done). Whatever the objective merits of the king and pawn endgame, I can’t believe Black has more practical drawing chances there than in the minor piece ending.

Incidentally, I played the knight to d4 because, aside from being an obviously good square, I wanted to avoid Black playing …d5-d4, sacrificing the pawn to open up the line for the bishop. I thought this might be a good practical chance for him – even though I assume the ending must be lost. At least he gets some play that way though.

33. b4 Be8, 34. Nc2 a4

Black offered me a draw at this point and, perhaps trying to bluff me that he was unconcerned about the position, wandered off to get up a cup of tea. I was going to turn him down instantly but then suddenly decided to pretend to consider the offer as he left the table. I liked the idea of him going to the kitchen and coming back all the while hoping I’d accepted the draw only to be deflated when he got back to the board.

In fact I’d already worked out the winning plan. I’d played 33. b4 to block the queenside then I was going to work my knight around to g5 to attack h6. He would have to play …h5 and then I could re-route the knight to g5 to attack f7. Fool proof. The only problem was I couldn’t work out the knight’s tour sequence to get my horsey from d4 to where I needed to be. After a few minutes of failure it suddenly occurred to me to start at g5 and work backwards – and as soon as I’d had the thought the solution came to me in a flash.

First though, there’s time to improve the position of my king. There’s no need to rush – as is demonstrated by Black spending the remainder of the game just shuffling his king and bishop back and forward.

35. Kd4 Kc6, 36. Ne3 h5,

I had time to start really enjoying myself at this point. Back at move 33 I’d initially intended to get around to f3 by playing Ne3-back to c2-e1-f3. When I got here though I decided it would be prettier to avoid going back to a square the knight had already visited. I was really getting into the knight’s tour theme.

37. Ng2 Bd7, 38. Nh4 Be8, 39. Nf3 Kb6, 40. Ng5 Kc6, 41. Nh7 Kb6, 42. Nf6 Bc6

And now I just need to lose a tempo. I chose h4 because I liked the idea of all my pawns being on dark squares. Perhaps h3 might have been more cautious though.

43. h4 Bb7, 44. Ne8 1-0

Black resigned because after White plays Nd6 the pawn on f7 is going to fall.

So there you go. A very satisfying win that contained only a couple of tactical sequences – neither of which actually happened in the game. OK, my opponent was rather helpful in that he chose to try to sit back and try to construct an impregnable defence rather than seek active counter-play but this game convinced me I had some positional ability after all.

The game remains one of my favourites. It’s particularly pleasing to look back at a more or less full board at move 15 then see the game unfold more or less entirely as I’d wanted it to – even finishing off with an ending position of the type I’d envisioned some 25 moves ahead of time.


ejh said...

I used to play that line as Black and I like it, but it's very easy to find yourself in a bad-French position if you're not careful in the early middlegame.

Ryan Emmett said...

Nice game. I admire that sort of remorseless chess, following a long-term plan, than I do flashy tactics.

Jonathan B said...

Thanks Ryan,

I agree with you about the "remorseless chess, following a long-term plan". Mostly I find it more admirable because it's so hard to do. The game I posted today was rare - usually of course I bugger things up along the way.

ejh said...

Do note that not analysing unnecessary tactics is dependent on having something constructive to do otherwise. A lot people think or hope) it means "avoid analysing tactics if you possibly can".

joe s said...

Nunn's quote is interesting but it was made in 1988. Would he still agree that analysis is error prone when computers can now run through multiple variations in a fraction of a second without any mistakes. Perhaps he means that we humans are error prone. It is surely time for Nunn's book to be revised and updated ?!

Jonathan B said...

Hi Joe,

You make an interesting point.

The context of Nunn's book makes it clear that he is talking about analysis being an error strewn activity for human beings rather than silicon life forms. I'm sure this will remain so (well it will for me) regardless of how well computers calculate - although I wonder if computer assistance/training can improve human calculating ability.

btw: I think that 1988 date may be a typo on my part. 1998 sounds much more likely to me. I'm going to check.

Jonathan B said...

Yes, Nunn's book was first published in 1998 and reprinted the same year. I'll ammend the post now.

Still a decade ago though so your comment is still valid.

ejh said...

Incidentally Tom's warning note is absolutely right. You can't cut out tactics and you shouldm't try to - if you do try, you'll end up having to try dodgy tactics when you're under the cosh because you avoided calculating good lines when you were ahead.

The point is not to waste your time analysing (or playing) speculative lines when you have good positive and simpler alternatives available (a possible example is Horton-Smith, 4NCL 2005, which you can find on Chessbase Online Database in which I cast away half a point in a very strong position against a 200+ player by trying a flashy line rather than play the solid moves that would have maintained my dominance). This is of course a much harder thing to do than to tell people to do.

It's also hard because very often, indeed normally, you can't simply tell what's necessary and what isn't, and you do often have to go into complex lines where you're making a bet on the outcome rather than just let the opponent bluff you (or simply equalise).

It's not straightforward. But there are ways to play sensibly without being afraid of tactics.

Jonathan B said...

I have a very good example of me failing to make the correct judgement from my most recent game.

I'll write it up for the blog for next week.