Friday, June 22, 2007

Mental Muscle: Fast Twitch vs. Slow Twitch

Those of you into exercise know that there are two general categories of skeletal muscle: fast-twitch and slow twitch. Fast-twitch muscle is capable of short bursts of intense power, but tires more quickly; slow-twitch muscle is less powerful but capable of long periods of aerobic exercise. I think that, when it comes to chess, something comparable applies to the muscle between your ears.

The #1 mantra of chess success is "study tactics". To me, however there are two separate components to tactical strength:

  • speed of pattern recognition
  • depth of vision

This is a bit artificial...the two aren't completely separate. Here's a hypothetical example to hopefully clarify what I mean.

You analyze a game of yours with Fritz, and Fritz points out that you could have won a piece by force. The error could fall into one of two broad categories:

  • "Oh, of course! I should have seen that!" (a sign that either you weren't playing "real chess", or that you were playing "real chess" but you missed a relatively simple tactic) This is your standard patzer error that tactics drills a la Heisman and proper thinking technique help to weed out. The aspect of chess strength most at play here is pattern recognition, and quite possibly the speed of calculating short forced sequences.

  • "How the hell was I supposed to see that? That's twelve ply deep!" Assuming it is a fairly forcing sequence, the relevant component of chess strength here is how clearly you can see the board several moves ahead. Even if it's a Kotov-esque tree of variations, visualization is an important component of the calculation process.

For me, the approximate boundary between the two in a middlegame of average complexity is six to eight ply--in other words, falling into the Chernev and Reinfeld "how to see three moves ahead" range of set up the tactic, execute the tactic, collect the booty.

I don't think that the idea of training your "visualization muscles" is often addressed. Instead, we have books or programs such as CT-ART that have a gradient of tactical problems from simpler to harder. Then, to do the harder ones, you need to work that muscle and see deeper. To me, that's a bit like exercising your back by doing squats. Yes, the back muscles will coincidentally get worked out a bit while you develop your legs, but you're not actually targeting them.

I have recently started studying master games in earnest. I try not to spend too much time on each one (I'd say 5-30 minutes, depending on the depth and quality of the annotations), but I usually use a board and pieces and take notes. For most of the annotations I push myself to visualize as best I can the sequence of moves. If it's a bit fuzzy, I may make a move or two on the board and try again. I do the best I can without spending too much time on it, then move on.

This position from Alekhine-Marshall, NY 1924 struck me as an excellent example:

Alekhine has the following side note: "[8. d5] must be done at once, inasmuch as Black, after 8.Be3, would obtain counter-play by means of 8... Bxf3 9. Bxf3 e5
10. fxe5 dxe5 11. d5 Nd4! (12. Bxd4 exd4 13. Qxd4 Nxe4), etc."

Set this position up on a board...go on, do it. Try to visualize down to the final position and get a feeling for what "etc." entails. Try to visualize, for example, why White doesn't play 14.Qxe4. If you can't get it, make a move and try again. Repeat until you get it. I'll add a comment about the final position at the end of my post...spoiler alert.

In the past my eyes would start glazing over at around 11.Nd4 and I'd just continue playing over the main line of the game. Here I said, "this isn't impossible, push yourself a little". While it's not a forcing sequence, it's a logical series of exchanges. With a little effort I was finally able to visualize the position at White's move 14 fairly clearly.

Note that I'm not trying to calculate my way through to the end, just visualize. In some cases, as you work through an annotation you'll be struck by alternatives: "hey, wait, why not _____ ?" If something jumps out at you, then go ahead...take a look, try to quickly figure it out, and if it's unclear write a note and move on. I have found that lengthy, linear analysis, especially by lesser annotators, is often suspect, and you may find that your own ideas are correct. For example, I found one instance in Capablanca's Best Chess Endings where Chernev missed a classic queen sacrifice and double-bishop mate. I was pretty chuffed about that.

I am hoping that by studying master games in this fashion that I'll overcome some of my mental laziness and improve my ability to visualize several moves ahead. One of the few books I've come across that actually touches on this aspect of chess development is Best Lessons of a Chess Coach, where in several critical game positions the reader is pushed to see ahead as far as they can.

I think the study of master games tends to be overlooked for a couple reasons:

  • You aren't getting a concentrated dose of tactics, or endgames, or opening theory, or strategy. So, if you're setting goals like "today I'm going to study tactics", or "this week I'm going to work on rook endgames", etc., you're going to be using other, more specialized, training material.
  • You aren't being spoon-fed the information. You have to actually think about the game on your own and ask your own questions. For example, an annotation may read "obviously forced" or "this alternative would result in a dead-lost endgame" as if it's self-evident, but you actually have to stop, think, and see if you can understand what the annotator is getting at.

Spoiler alert for the above position: I had difficulty visualizing that the e-file was completely vacated of pieces, save for the king. I had a residual shadow in my mind of one of the bishops on the e-file. After working my way through it once, I was able to work forward from the given position and visualize fairly clearly the final position, complete with vacated e-file.


transformation said...

i will have more to say about you soon enough elsewhere, but yours is really one of my absolute favorite blogs right now.

there are those we can lead, and those in turn far ahead of us but perhaps consequently cannot learn from. but so few are our peers, and i can relate to every single word you write.

i am either traveling the exact same ground you travel on, or very similar or did similar things--i.e. 941 games for three years, exactly as you describe here, not to slow or too fast, but massive digestion of protean.

i also love the tone and cleanlyness of all of your blog. you are a logical mind, with a good edge. you are a real asset to me to mirror to. i can read so few here as peers, again, above or below, i feel, or at different junctures of learning.

do you know Batesons idea of type iii learning? i find type iii learning here. the learning of learning.

i have also don tons of physical training, running, climbing, etc. interested in nutrition, and pretty on it.

warmly, dk

Grandpatzer said...

Aw shucks.

I wasn't familiar with Bates III learning until I googled it. Consciously I feel I'm working on "learning how to learn", but maybe unconsciously I'm kicking it up another notch.

IA said...

You said, "I don't think that the idea of training your "visualization muscles" is often addressed" and that you want to "improve my ability to visualize several moves ahead".

Let me tell you about a practical new approach to chess visualization based on 800 positions taken from real games that I discuss on my blog here.

This new approach to chess visualization training will stretch your vision from 4 to 39 half-moves while expanding it from 1 to 2 to 3 sectors of the board.

It might be just the thing you need to work your chess visualization muscles!

Best of everything.