Computers and the fringe of theroy

Over the board players: Here are a two ideas you might try with your chess program that will help you prepare for your next tournament, these have been around for awhile, but many times are forgotten.

* The fringe of theory: Set up your board, and play through your favorite openings, as you approach the end of the well annotated moves (normally moves 8-15), and begin the middlegame, play against your computer's top level in these positions for the next 5-10 moves. Write down your computer's moves and compare them to what you would have played. Back up the game to your original position, choose another line of play and do the same thing. Don't forget to play through these positions from both sides of the board, using your computer as an opponent. Doing this with multiple variations of your openings, will allow your computer help you learn how to handle these critical positions as you progress from the opening to the middlegame.

* Set up endgame positions and play against your computer's top level while starting a pawn down. This will be hard on the ego, but will sharpen your ability to defend with inferior material. If you can manage a few draws against your computer this way, there is a good chance you'll save a few points in your tournaments, since your OTB opponent is not likely to be as skilled in this area as your computer.

Nothing personal, but nonsense.
Life is too short to stuff a pepper, so they say:
Most players know that computers play by brute force and not by (sometimes limited) human logic, so therefore computers play a different move at crucial points. A human will not be able to use an advantage learned from a computer when playing against a human. The amount of variations would be horrendous. Perhaps years of learning from a computer to end up in a suitable position against one human in one game? This is a waste of time. Learning to play like a computer against a human brings me back to the pepper problem.
Play real chess against a stronger player if you wish to improve. If you don't have any friends (?!) then always play the same opening as white or play the same defence as black against white opening moves against any opponent (even a computer) and learn from it, for as long as you wish until you move on to another opening.
Is it me?

Originally posted by bill718
Over the board players: Here are a two ideas you might try with your chess program that will help you prepare for your next tournament, these have been around for awhile, but many times are forgotten.

* The fringe of theory: Set up your board, and play through your favorite openings, as you approach the end of the well annotated moves (normally moves 8- ...[text shortened]... urnaments, since your OTB opponent is not likely to be as skilled in this area as your computer.

All well and good but who the heck is TheRoy?

One opening I wouldn't trust computers on is the KID. They constantly over-evaluate the white side . . . . until it's too late.

A famous recent example is Naka vs So last year-

http://www.chessgames.com/perl/chessgame?gid=1800188

Originally posted by SPswindler
One opening I wouldn't trust computers on is the KID. They constantly over-evaluate the white side . . . . until it's too late.

A famous recent example is Naka vs So last year-

http://www.chessgames.com/perl/chessgame?gid=1800188

What does that game between two people have to do with computer chess evaluations of the KID?

This thread is relevant as chess theory is important. Grandmasters use chess programs and analysis engines to train. So how can the advice be nonsense.

Originally posted by knightrider63
This thread is relevant as chess theory is important. Grandmasters use chess programs and analysis engines to train. So how can the advice be nonsense.

To be able to use chess engines effectively for this purpose one has to be quite skilled at the game already. An engine may say a position is equal, some 1600 rated player may see this and head for that position. Little does he know that it is only equal if he plays extremely accurately over the next dozen moves in a very difficult position. How easy a position is to play doesn't factor in to a computers "thinking" in any way. I would rather be a pawn up in a simple position, than +2.00 in an unclear sharp position in which the slightest miscalculation leads to a losing position.

When I get near the end of theory, I look at master games that continue from that position. It is easier to pick up ideas, they think like I do.

Engines are great for finding blunders after a game, but I wouldn't try to play like an engine in the middle game, you have to see too much for that play style to work.