On May 11, 11:06 am, raylopez99 wrote:

1) Would you feel equally confident if we only gave crafty 11 ply? 10?
8? 4? Where do you draw the line? What non-arbitrary criteria are you
using to suggest that 12-ply is meaningful whereas 3 ply, obviously,
would not be?

2) What objective criteria are you using to define "extremely close"
such that you don't trust the computer's ability to rank players
properly?

I'm very curious to hear your answers to these questions.


In truth, nobody in this thread really knows, and indeed further
research is needed. But the burden of persuasion is on Camp #1 to
make their case--that so called "positional sacrifice" positions are
rather common in a game of chess and that chess is NOT largely tactics

This is absurd. The burden of proof, if indeed there
is one here, is on those who would maintain that a
mediocre chess program can *accurately* rank the
world champions.

My view is that the greater the superiority of the
chess engine over those it judges, the smaller this
burden of proof becomes; the less important the
dead-on accuracy of move-ranking becomes as a
requirement to accurately rank the players relative
to one another. An inferior program is good for one
thing, though: spotting gross blunders in the games
of the world champions. A perfect example would be
the famous game where BOTH human players
overlooked a simple tactic involving moves: 1. Q-h8+
(gives away the Queen) ...Kxh8, 2. Nxf7+ (capturing a
pawn) ...K-moves, 3.NxQ, netting a pawn. A human
is *far* more likely to make this oversight than even a
mediocre program like Crafty_at_12 plys.


(these are the assumptions behind their claims--I claim the
contrary). History has shown otherwise.

Nonsense. History has not yet shown that RL's
ignorant assumption regarding tactics is anything
other than just that: an assumption. The figure often
quoted by GM Tarrasch (99%), was not an actual
measurement, but only a way of making a point.

For purposes of the present discussion, it would be
far more useful to think of chess as "only" 90% tactics.
Even here, it is not really the percentage which is the
main point, but rather it is that tactics take precedence
over many lesser things. It's akin to having nuclear
missiles or, say, an aircraft carrier within range.


Indeed, on the last point,
Kramnik missed a mate in one last year.

This was an anomaly; generally speaking, the
world champions don't overlook mates-on-the-move
in top-level play.


Chess is largely tactics, and
that's why it is fair to have a chess engine rate the champions.

Ah, but "fairness" was never the issue here.

What critics have focused on is the subject of
*accuracy* (something RL obviously knows
nothing about).


You can make 30 brilliant "deep" positional moves in chess, have a clearly
winning position, and still lose a chess game in a mate in one.


Obsessing over GM Kramnik's fluke blunder, I see.
The phrase "grasping at straws" leaps to mind.



That is chess. A PC would score you poorly in such a game,
even though you were "brilliant" up until your blunder

Now that you mention it, GM Kramnik had
already tossed away his win by that point; far
from being rated as brilliant, I think his computer
opponent would likely have ranked him as weaker
than itself, since the human's moves did not as
closely match its own quirks. What about
Crafty? I imagine it would rank most other
programs as superior to human GMs for similar
reasons, including those who are objectively
superior.


(and perhaps unappreciated by
the PC, though I have argued in this thread that PCs are in fact not
so bad at rating positions that require positional moves, even
exchange sacs).

The question is not "are they bad" at rating positions,
but rather, it is "are they good enough" to *accurately*
rank the world champions. My view is that if they are
good enough, it is probably programs such as Rybka
and Shredder -- essentially, the ones rated at the top
of the heap -- which are good enough for the job. And
we need a decent sample size for this sort of approach
to render meaningful data; at least one of the champs
had only one single world championship match, of
course, against but one single opponent; that is no
way to do this sort of thing properly.


In fact, Camp #1's arguments are better if we were trying to rate
"correspondence chess" champions rather than OTB champions, since in
correspondence chess tactics are much less important than deep
positional moves.

I wouldn't say that. It would be more accurate to
say that in correspondence play, there are fewer
gross tactical blunders and that in general, the
tactics are deeper and better executed. In sum,
it is a closer match to what the chess programs
rate as perfect chess.


But that was not the inquiry of the original
article ranking of champions: it was for OTB world championship play.
However, that said, I would not be surprised that even for
correspondence chess players, rating such players with Fritz 5.31 at 5
seconds a move would give you a pretty clear indication of the best
correspondence chess players, since good positional moves and good
tactical moves are largely one and the same in chess

Not true.

In one of my current games at RedHot, I just
chose to double my Rooks on the only open file,
as opposed to snatching a free pawn and
surrendering that file to the opponent, even
though doing so would not make a whole lot
of difference. The point is, grabbing at the
maximum of material gain will most likely rate
higher with a shallow search by a mediocre
program, such as Fritz 5.31; yet my move is
likely to result in an immediate resignation
because it stomps out any imagined counter
play and thereby underscores the fact that I
am up the exchange for nothing and can win
material almost at will. Suppose Crafty_12 ply
penalizes either move as inferior to the other --
how does this style issue contribute to ranking
the world champs *accurately*?


(again, this goes to chess being 99% tactics).

Personally, I think this obsession with the
figure "99%" may be the root of the problem. If
you can learn to accept that tactics are, let us
say, higher in rank than positional play, but not
so overwhelming as to merit such a figure as this
"99%", then you will finally begin to grasp the
issue.

Are Generals and Colonels 99% of the military?
Do airplanes and jets and missiles constitute 99%
of the armed forces? No, but their greater weight
may make it seem to be such. Suppose that GM
Tarrasch had instead stated that "artillery is 99%
of winning the war" -- would this mean that all other
segments could then be dismissed as utterly
irrelevant? Would you then sit down and just start
counting up beans (here, artillery pieces)?

To sum up, there are *at least* two problems with
the approach taken:

1) sample size too small (except with GMs like
Steinitz and Botvinnik)

2) 12 plys depth of search is likely insufficient


-- help bot