On Apr 26, 9:24 pm, "David Kane" wrote:

So, understanding how chess works, and how chess playing computers
work, and having seen Crafty evaluate pretty good myself, I have to
side with the original article.

I would not go so far as to say that I side with the original argument,
only that Riis' objections were groundless. In fact, the original authors
have done some groundbreaking work on developing a
methodology to rate chess players. It is, at the very least,
very interesting, and a refreshing change from the pseudo-science
historical ELO/chessmetrics stuff. The problem with the work is
that it applies a new method to a very hard problem (ranking
world champions) when they haven't even shown the method's
worth when applied to easy problems (ranking everybody else).

It seems to me that the above comments themselves do
a decent job of showing how the "groundbreaking work"
is little different from ChessMetrics' pseudo-science.

------

In one of the defenses to a criticism, it was argued that
even a weak chess program could be utilized effectively
to rank players, due to a strong correlation of some sort.
But in constructing their example to demonstrate how
this works, the authors (as always) made some invalid
assumptions; in this particular case, that apart from the
single strongest move in a given position, the remaining
choices are distributed or chosen evenly. Obviously,
the remaining move choices are anything but equal, and
how a player chooses among them is a big part of how
strongly they play. The stronger the player, the more
likely he would be to go for #2 as opposed to #10
(granting the oddball assumption of exactly ten choices
per position). All these invalid assumptions come off
as a clueless math major having fun "playing around
with" numbers which just happen to relate to chess.

Thus far, the only works I have seen which are not
seriously flawed in terms of logic and reason, were a
few of the brief criticisms of the published works by
the math whiz-kids.

-- help bot