Time Spent Evaluating

In his book The Seven Deadly Chess Sins GM Rowson states that when you have an analytical position that spending more time analyzing can be very fruitful (up until about 20 minutes when you can get tired and confused). However, he notes that in non-analytical positions (e.g. White plays 1.a4 and Black does not know what to do), that spending more time on your judgment will not make it better. So you want to play judgmental (non-analytical) moves relatively quickly using the principles you know and how they apply to the position. Save your time in slow games for analytical moves, when more is at stake (criticality assessment) and you may need time to calculate carefully.

This is old news. But yesterday I was giving a lesson to a Class A player (1800-2000) and he found he was spending too much time evaluating (not analyzing) positions during critical moves.

I define evaluation as determining "Which side stands better, how much better, and why". My four evaluation criteria are:

1. Material

2. King Safety

3. Army Activity

4. Pawn Structure

For more on this see Evaluation Criteria (http://www.chesscafe.com/text/heisman27.pdf) and other columns on this such as Activity is the Real Goal (http://www.chesscafe.com/text/heisman63.pdf) which discusses how an item like "space" fits into this list.

My student and I discussed his problem of taking too much time evaluating and I told him something he did not know:

Never spend inordinate amounts of time deciding between evaluation of candidate moves! There are only two main possibilities:

1. The evaluation of your candidates is so close you can't tell the difference. In this case, if you are sure the difference is not major, the moves must be close in value. Therefore, it likely doesn't make much difference which one you play! Chose one by some quick tiebreak procedure (see below), or
2. The evaluation of the candidates is not close in value. Then you are done, as you have identified the better move.

In either case you should not take much time.

Tiebreak on relatively similar evaluations might be:

1. Which line looks more fun, 
2. Which line is more/less drawish,
3. Which line's evaluation you trust more, etc.