On May 1, 7:51 pm, raylopez99 wrote:


Ponder this: if Crafty, Fruit, Toga, Shredder, Rybka, and other VERY
strong UCI engines *ALL* rate the best players ever as Capa, Kramnik,
Karpov, Kasparov, *in that order* (cuz 'puters don't lie!), what does
that say?

It says that all these programs contain the same or
very similar flaws; that their overall results mimic
one another as a result of what HAL9000 might refer
to thus: "weak minds think alike".

I would be willing to bet that the programs listed
above would not all rank the world champions in
the same was as crippled-Crafty, provided we can
cut out the middleman, the human who tweaks
and fiddles with the objective results. An example
was the guy who called it a "tie" when GM Steinitz
crushed everyone in one category, but he "felt"
that GM Fischer should get a freebie for coming in
a distant second. (BTW, GM Steinitz was the only
player whose sample size was probably adequate.)


If Fruit, Toga, Shredder, Rybka *ALL* rate the very best players as
"Capa, Kramnik, Kasparov,Karpov, *in that order*", but Crafty puts
Karpov before Kasparov, does this mean that the top two players are
indeed Capa and Kramnik?

No.


If, like another poster says, Tal won a match by playing complications
(but lost the rematch BTW), does this make Tal the 'best' player at
that point, or, over his career, or, do we analyse ALL his games,

Good idea! Sample size is crucial.


including game played when he first learned chess, or do we just count
the 20 games where he became champion? Some players can "peak", but
should not the test be over the player's career (like Lasker)?

Um, all players can peak. Obviously, there was no
good reason to exclude top-level tournament games
which are comparable to the W.C. matches in many
ways.


If Kata Kamsky becomes champion by spiking Nigel Short's orange juice,

I believe it was GM Fischer who drank OJ. And it's
Gata, not Kata. You may have gotten these confused
if while you typed here you were looking at a picture of
the Kama Supra -- Toyota's "sexy" new convertible.


does that make Kamsky the better player, because he used "shock"
tactics?

No.


Is Tal's 'unsound' sacrifices good chess, or shock tactics?

A false dichotomy.


If "Fischer Fear" where 6 great players all choked and got
whitewashed, in Fischer's championship run, a true test of Fischer?

Six? Can you list all these chokers?


Or just good players choking?


These championship matches are peculiar in
that each yields a single "winner" and one "loser".
If it should happen that the weaker of the two
players falls behind at the start, the result can
be a freaky whitewash, because of desperation
in conjunction with the superior player's ability to
exploit it to the max. In sum, a shot at the title
throws out any consideration for protecting one's
rating. Even with a forced draw in hand, a player
might well try something stupid, if he thinks he
has even the slightest chance for a comeback.


Long story short: we need further research, to the extend anybody is
motivated to do it

For those who support the crippled-Crafty method,
I would suggest beginning by taking the weakest
chess program available, and comparing its results
on the same data set to crippled-Crafty's. If you
get a "match", this would seem to confirm the thesis
that even patzers can rank world champions alike.


(since let's face it, computer chess outside of
writing chess programs for the masses is not exactly the best funded
field anymore), but at the moment, the tentative research shows:

!!! Greatest chess players ever? Capa, Kramnik, Karpov, Kasparov, *in
that order* (cuz 'puters don't lie!) !!!

No, it doesn't. Crippled-Crafty did not demonstrate
anything whatever about the "greatness" of those
players, but it may possibly have determined which
ones were most prone to blunder. In order to check
this, we would need to see how the results were
handled.

-- help bot