Originally posted by SmittyTimeSkeeter !
RJHinds is the new Skeeter.
I still think I could beat him OTB.
I was trying to remember her name. She was the most awful person, so smug and sanctimonious in her posts, claimed she had irrefutable proof that the Apollo moon landings were fake but wouldn't divulge it, she was a real loon, prolly didn't even know how to play chess but was a top site player for a long time.
Hi Zygers
I've been saying for quite a while now that Carlsen's knack is not playing
like a computer so I am not at all surprised his 1st choice is pretty low.
My theory is (and it is just a theory.)
He is a problem setter up there with Tal, Kasparov, Fischer, Lasker....
There is an element of risk involved but it is not too perceivable OTB.
(A sounder Lasker if you like. Lasker stated to play for a win without any
obvious blunder from your oppoent you must take a risk.)
I'm not saying the rest are computer clones but they are rational thinkers.
and so are, to a certain extent the top computers.
(awful choice of word 'rational' but I'm playing chess at the same time
as writing this and am waiting for them to move. My head is flowing with
positions.....losing positions.)
A box would never play a second choice move even on the basis that it
would set more OTB problems for a player.
Some top players are the same.
Though Carlsen is the top dog grading wise perhaps selecting him for this test
was not the best choice. Go for a player in the top 10-20.
Not a Nacko or Topalov style. A Kramnik style.
Originally posted by KilgoreTrout15But....Skeets left in Jan 2012, you joined in Dec 2012.
Skeeter !
I was trying to remember her name. She was the most awful person, so smug and sanctimonious in her posts, claimed she had irrefutable proof that the Apollo moon landings were fake but wouldn't divulge it, she was a real loon, prolly didn't even know how to play chess but was a top site player for a long time.
A reincarnation or an avid reader of Skeets old posts?
Jesus wept.
Originally posted by greenpawn34The marble bit was an analogy to the players being out in the open, not hidden, and therefore supposedly easier to detect cheating, but there is one hitch in his scheme: There are many engines out there today, Hiarc, Fritz, Houdini, Chessmaster, etc., and each one gives a bit different set of moves. So in order to account for those you either have to have those engines loaded directly in a computer or have a simulator that simulates the exact set of moves each one would make.
Ken Regan - Fide's man to catch cheats using computer match up.
http://rjlipton.wordpress.com/2013/09/17/littlewoods-law/
Can anyone put in laymans terms what on earth he is going on about.
The marbles scenario confused me (mind you easily done) even more.
So in order for that to be effective, you would have to have computers capable of making multiple assessments of positions for at least a dozen players in a round robin or hundreds or more in an open.
This might be able to be done but it seems to me it would require a super computer of the highest order, such as the ones currently at the top of the world list of computing speed, like the Chinese Tianhe-2 which puts out 33 petaflops per second.
So that computer could probably do the job but at this stage in technology you are talking about 100 million clams.
If they get that down to 1000 bucks it might be something FIDE or USCF or tournie directors can think about using but till then...
Look at how long the mods used to spend finding the cheats here, till they gave up.
Originally posted by greenpawn34He could have said a bit more about it, but it sounded like they did not have the resources to test every single game. So they test notable performances - scores that stand out in the cross table. The notable performers are the shiny marbles.
Ken Regan - Fide's man to catch cheats using computer match up.
http://rjlipton.wordpress.com/2013/09/17/littlewoods-law/
Can anyone put in laymans terms what on earth he is going on about.
The marbles scenario confused me (mind you easily done) even more.
But there are two problems: 1) You can win a class prize just by having a decent score, and thus allow yourself to lose some games to avoid detection; 2) Statistically, it is possible that a player may have a have a higher than normal engine matchup in a single tournament even though they didn't cheat. To avoid wrongfully branding them a cheat, the statistical certainty must be higher. But then some cheats will not be detected.