Email 9: Ken responds, regarding the COA

Date: May 2, 2011

First Name: Ken
Last Name: Wilsdon
Country: Canada
Subject: Ken responds, regarding the COA

Hi Graham,

Glad you are enjoying H&M-S. Don't knock you COA. It does a lot of the mundane grunt work that H&M-S and Seirawan recommend in a tiny fraction of the time it takes a human to calculate it. Space and Material (Force) are two important elements in developing a plan.

For a Beta product, you are to be commended. Now you just need to add Pawn Structure and Time (I know, easier said than done!), and tweak a few things currently, and you will have a truly excellent product. Already it is a product to be proud of, and a great help to newcomers to the game.

Wow, there is a lot to comment on from your email. Soon we will need to write a book! Here goes.

H&M-S gives each of their point count items listed on page 8 as one unit, and three uncompensated advantages equals one pawn. Thus he has multiplied the common relative values by three in order to accommodate his point count system. The bottom paragraph on page 10 and the top paragraph on page 11 point this out. He has not deviated from what is said in Wiki, except for the value of the king at 9 (three in common counting). The sixth paragraph on page 11 gives their rationale on why he values the King, which is in keeping with your comments.

The relative value of the pawns and pieces is a very interesting topic, with quite a bit of variety. Larry Evans (LE) in New Ideas in Chess (p.77) places the Bishops and Knights at 3 1/2 each, and the Queen at 10 (!), the King with no value. I can't remember which book I read, but it said that modern theory would place the value closer to Knight 3 1/2 and Bishop 3 3/4 (because generally having the Bishop pair is better than a Knight and a Bishop - though that is somewhat situational. H&M-S gives a point for having the two Bishops, or having a Bishop over a Knight. In closed games, I want my Knights a lot more than my Bishops, at least until the position opens up, and then I wish for my Bishops back!).

World Correspondence Chess Champion Hans Berliner (HB) in his book The System, p. 14, has done a lot of work in this area, and gives a fractional analysis of the pieces. His system allows relative values that are well known when exchanging, which some of the others do not do as well:
  1. 3 minor pieces about equals Queen and pawn.
  2. 2 Rooks are better than Queen and pawn
  3. Rook and Bishop are only a little worse than a Queen
  4. Knight and pawn or Bishop and pawn are less than a Rook.
He adds or subtracts based on mobility of the pieces, which your COA program already accounts for. The problem with his system is that it is fractional, so no ordinary human could keep track of the values (but a computer program could).

On the following pages he highlights his views of the value of each pawn, giving greater weight to those that are on centre files and those farther from home up to the 6th rank (after that, I presume it would be worth at least a piece, as you would not want it to Queen, and would be willing to sacrifice a lot to stop it). This fits in well with the point count system, where a pawn on 4th versus a pawn on 3rd is an advantage, as well as an advanced pawn is an advantage.

In chapter 3 he deals with "chunks", as that was something that interested you in a previous email.

None of these works save H&M-S gives a value for the King. I guess in some ways I understand, because you cannot actually "take" the King, because the game would be over! Yet the King does have offensive and defensive capabilities, as you point out, because it moves similar to a Queen, but just one square in any direction. Thus its mobility (8) is greater than a pawn (3 or 4 depending on whether it is the first move, and whether there is something diagonally to attack), the same as a Knight (optimally 8 - just different squares), less than a Bishop (up to 13) and can move in more directions, so I would put its value over that of a pawn, but I am unsure if I would say it is more powerful than a Knight or Bishop, at least in the Opening or Middlegame.

As you move to the endgame, I think it could have a value of close to 4, as it often is more powerful than either minor piece, the opponent generally needing two of them to entangle the King.

In the Opening and Middlegame, I think King position and safety are more important than the King's strength. Part of King safety has to do with how well he defends pawns and pieces around him, and what kind of holes he cannot defend. Without the kings being valued, someone could have a zero value in the traditional material value system (just the king on the Board), but if you include a value for the King, you would never have a zero.

If both sides are zero, it is definitely a draw! Once you move to the Endgame, the King moves all over the board, and King position still is critical, but King safety takes a back seat. GM Viktor Moskalenko (VM) in his book Revolutionize Your Chess, places King position as a 5th critical element (or "Touchstone" as he calls it) in the game, along with STFP as per LE (we'll soon have enough initials to make our own glossary!), but he rearranges things differently.

Yasser Seirawan is an excellent chess player and teacher, and you have done well to model your site after his teachings. As far as the spread sheet, you have done a lot of work! It's going to take me more time to look through all that you have done there, and comment on them. I would suggest that for your own repertoire benefit, when we start to play we start with one of the lines you have (are?) already analyzing in the King's Indian (KI).

While I don't normally play d4 and c4 together in the opening, because of the vast number of responses Black has at his disposal, I am not afraid to try my hand against the KI using one of these positions, and we could analyze it further than the opening. It would be good for me to try a different opening to learn new ideas.

As to knowing what the reason for each move is in an opening, the "How to Open a Chess Game" will help with a lot of lines being explained by Grand Masters. Down the line, you may be interested in Irving Chernev's classic, Logical Chess Move by Move, which has helped tens of thousands with this issue, as well as a couple of other books along a similar vein, that explain why each move is made using all of the strategies we have been looking at throughout an entire game.

I know you must have a lot on your plate with all I am throwing at you and with your own goals and projects in this realm (when your University dissertation is complete! Then I'll have to call you Doctor Graham!), but when you are ready, I can send some of them to you if you don't have them or can't find them in print.

I think I will wait let's say a week or so to send a link, as you will need some time, I think, before we start our game. What do you think?

Until next time,

Best wishes,