Have we been duped by AlphaGo?

[John Fairbairn]

Of possible interest to some amateurs, an article by Jerry Silman, who seems very highly regarded as a chess teacher:

https://www.chess.com/article/view/long ... nstructive

The core message is play slow if you want to improve.

All the people on the site forum seem to support Silman. Some seem to point to dramatic differences in rating by taking a little more time to think.

[dfan]

Indeed, although all he is doing in this essay is recommending avoiding ridiculously fast time controls (3 minutes a side for the whole game) in favor of at most moderately fast time controls (30 minutes a side for the whole game). I doubt many people here would argue with that.

[Kirby]

Indeed, although all he is doing in this essay is recommending avoiding ridiculously fast time controls (3 minutes a side for the whole game) in favor of at most moderately fast time controls (30 minutes a side for the whole game). I doubt many people here would argue with that.

Agreed. I don't argue with that either.

However, you have to make a decision: have tons of fun playing blitz (without learning much), or be serious and play with longer time controls so you can actually think.

Part of the reason that this thread went on so long is that we are talking past one another. Many that are arguing for longer time controls argue a point similar to the quote above - you need time to actually think to avoid making silly mistakes.
Of course I don't disagree with that. In fact, I've seen professional games played at time controls that even I suspect are probably unreasonable for resulting in quality play (e.g. where they play moves almost immediately in succession).

In many cases, additional time gives you the time needed to avoid some mistakes you might make without thinking of particular variations, evaluating the board, and so on.

Fundamentally, there is some formula for playing a good move. My argument is that (1) we don’t know what that formula is, and (2) I find it unlikely that the formula is as simple as:

Code: Select all

Quality_Of_Play(time) = time*constant

I argue that the formula is very complex. And there are even variables, such as fatigue, physical health of the player, physical stamina, and so on, which may result in additional time having a negative effect on the quality of play.

In short, I don’t argue that additional time cannot be helpful for improving and/or playing quality moves. I argue that the formula for a good move is as simple as that, and that it’s even possible in some cases, that longer time controls could lead to lower quality play.

Above all else, I feel that none of us really know this complicated formula well enough, which is why it’s so easy for us to resort to generalizations.

[Javaness2]

What does this have to do with being "duped by AlphaGo" ?

[Babelardus]

Of possible interest to some amateurs, an article by Jerry Silman, who seems very highly regarded as a chess teacher...

Silman is not the strongest chessplayer who ever lived, and his teaching methods are a bit unconventional, but I can attest that they work. He actually teaches you how to create, and then how to use imbalances on the board. Like this: If you have two knights and no bishops, close the position. If you have two bishops, open it. Make a trade to create a difference that fits your position, or create a difference and then fit your position to it.

It's a bit like "play A to get answer X, play B to get answer Y, then play C to make shape. Then use that shape to attack or defend" in Go.

[hyperpape]

An example from Bill: he says that Ota Yuzo resigned in a position where a modern player would still hope for a mistake: http://lifein19x19.com/forum/viewtopic. ... 61#p186561.