from my point of view the "sac" only works in a blitz game. move 13, c'mon, that can't resist analysis, even it's kasparov.

but very, very impressive and maybe not so unusal for blitzing....

The queen's advantage is maneuverability and speed; so if your opponent sacrifices that for initiative, you now have that advantage over him during the process of defending against whatever attack he's planning. In the case of sacrificing a rook for a minor piece, you ought to link up your rooks ASAP, since that is the major advantage you gain against a player who only has one rook left when you still have two. If he sacrifices a minor piece for a pawn, you ought to bring a minor piece into instant defense of whatever file/square he just opened up. That way if he starts to bring his own non-pawn pieces into play, you can (theoretically) maintain an advantage in army size at the point of attack.

To me, that is what makes Kasparov's sacrifice in this game so brilliant... not necessarily brilliant in conception, but certainly brilliant in execution. He may or may not have known exactly what he was doing, but by God he made it work. He was willing to sacrifice his queen for minor pieces/pawns AND the placement of his knights in the front center...because by doing so, he negated the maneuverability advantage of Kramnik's queen. The key to the sacrifice, and any sacrifice, is being able to use your tactical advantage to overcome your strategical disadvantage. In this case, Kasparov's tactically placed knights overcame the strategical disadvantage of not having the speedy, maneuverable queen.

--Nakamura

Also, real (FIDE) 2200 is much stronger than USCF 2200 the latter uses logistics curve instead of correct Gaussian curve and it has a bias due to "rating floor" garbage. "Life master" is nonsense one must keep up with chess to prove it over lifetime, ability wanes with age. In his "My Story" video super-genius Kasparov pulls rank on mere genius Plaskett by saying no-one but he and Karpov fully knew what was going on in their match. Although arrogant he is at least partially correct and Plaskett cedes his point. Brilliant people realize there are others above them and show humility to it, whereas average people are often clueless or jealous.

The discussion about the numbers of players and "inflation" reminded me of something I read a while ago on <chessbase> by Jeff Sonas. Have you seen it?

http://www.chessbase.com/newsdetail...

GM Bent Larsen may have been a 2600s player in his era in the late 1960s, but I have no doubt he would be 2700s in this era; his games indicate he was calculating more variations faster than even many of today's 2700s.

Ratings do not measure chess strength. They are mere descriptors and predictors of chess players' performances in the near past and near future in the same quasi-equilibrium group. If all of today's 2700 players were to suddenly play like patzers, but retain their relative strengths to each other and play no one else outside their group, they may retain their 2700 ratings, but they will still be playing like patzers.

that reminds me of a tourney in Denmark 20 years ago.

A really old guy was playing; he had an ELO of about 2400-2450. He lost some games, drew some and won a scintillating brilliancy. The 2600 GMs had a reverance for this 2400 "guy", despite their higher ratings and the old guy not being a pro player.

<I should point out that there were barely 1,500 active players on the January 1975 list, and that number more than tripled in ten years, to more than 4,600 active players on the January 1985 list. Nevertheless there was no inflation (using my meaning of the term). So the argument that inflation is a natural result of the general advance of chess knowledge would not explain why there was no inflation across those ten years.>

Yes, by itself this neither explains nor differs with the basic premise regarding how a bell curve fills out with increasing population, since the population considered is still quite small and a second factor is also important: Knowledge/dissemination.

During that time dissemination of higher level chess knowledge to the pool of FIDE members (or overall chess population for a better metric) was minimal, it was in fact stymied for political reasons. Moreover, the growth in knowledge has exploded with computers via engines and internet. This also has a primary effect.

The population growth of FIDE members (assuming all are active) from 1500 in 1975 to 4600 in 1985 reflects an annual increase of just under 12% (actually, 11.856%). With the same 12% growth rate this computes (from 1500 in 1975) to 45,000 in 2005 close to the actual number of active FIDE members then.

With 30 times the population the variance of a normal distribution increases by 30 times, or it's standard deviation "sigma" increases by the square root of this which is about 5.5 a large number now.

It is quite easy to fit a bell curve to a population size with a mean rating and standard deviation. To do this "back of the envelope" simply note the following ONE-SIDED probabilities http://en.wikipedia.org/wiki/68-95-...

P(2sigma) = 0.02250

P(3sigma) = 0.00135

P(4sigma) = 0.00003

As an (zero-order) approximate example, suppose the mean is 1800 and sigma is 250 so that 2800 is 4 sigmas away.

Kramnik vs Kasparov, 1994.
YOU ARE PLAYING THE ROLE OF KASPAROV.
Your score: 55 (par = 53)