On Jan 9, 3:04 pm, "Chess One" wrote:

That is untrue. You say above, as your basis, that if I draw 20 times
against 2,000 rated opponents I will not be at 2,000 level. Why the hell
not? That is exactly the level of player with which I can achieve an even
result, and is this not the entire function of the rating system to
determine? At least I always took it to be so.

Bingo. We now know the basis of P Innes' claim to be "nearly an IM
2450." He drew one - probably in an offhand game.

Chess One wrote:
"Kenneth Sloan" wrote:
Similarly, a win says that one player is better - but doesn't say
how much better.

But given 20 results, it does. PI

Twenty wins says that one player is a lot better but still doesn't say
how much better. It could be any IM playing against a random club
player or Kasparov playing against a total beginner.


Dave.

--
David Richerby Flammable Monk (TM): it's like a man
www.chiark.greenend.org.uk/~davidr/ of God but it burns really easily!

"David Richerby" wrote in message
...
Chess One wrote:
"Kenneth Sloan" wrote:
Similarly, a win says that one player is better - but doesn't say
how much better.

But given 20 results, it does. PI

Twenty wins says that one player is a lot better but still doesn't say
how much better.

Quite so, but 20 results distributed against a range of opponents and
including wins, draws, losses, will. Let us not argue by exception, or by
extremes. The rating system is a socialised and averaged one, right? So
let's make that the default, or basis for things.

What can be determined against 20 mixed results is, IMO, a very fair idea of
a players current strength against the general rating pool. I would think it
would be within 100 points of the result of 40 such games.

Phil Innes

It could be any IM playing against a random club
player or Kasparov playing against a total beginner.


Dave.

--
David Richerby Flammable Monk (TM): it's like a
man
www.chiark.greenend.org.uk/~davidr/ of God but it burns really
easily!