hmm yes good point, can't believe no-one else has made that point in 1000 posts. Incredible.

I feel sorry for anyone reading 1000 posts in this thread. I read the first page and a couple of the last pages. I don't feel much smarter.

Tmb86 wrote:

hmm yes good point, can't believe no-one else has made that point in 1000 posts. Incredible.

I feel sorry for anyone reading 1000 posts in this thread. I read the first page and a couple of the last pages. I don't feel much smarter.

tieics wrote:

I think if there is 300 ELO point difference between two players it's no point playing the game. Practically the player with lower ELO has 0 winning chance.

I f you want to troll, you are doing it RONG. You should start a thread with the title

"if there is 300 ELO point difference between two players it's no point playing the game. Practically the player with lower ELO has 0 winning chance.**"**

Genius! I predict 1000 flame posts, and you wont even have to respond.

totally possibly, if the 2700 had a connection interupted thing, which happens about 5 times per day for me

scheppy wrote:
totally possibly, if the 2700 had a connection interupted thing, which happens about 5 times per day for me

True, Scheppy. * If* they were talking about online chess. But they were talking about an official, rated OTB game between two adults with established ratings.

Tmb86 wrote:

hmm yes good point, can't believe no-one else has made that point in 1000 posts. Incredible.

Is this sarcastic? The point had been made, although I admit the poster above you did so rather eloquently.

"Is there any chance an ** average **high school football team could beat an NFL team?

Is there any chance an ** average **high school soccer team could beat a Champions League group winner?"

Well, first of all it's easier to blunder the magnitude of a mate in 1 in chess than it is in football or soccer (what would be the equivalent, really? You can allow a free touchdown, but that's no mate in 1 blunder).

But even if we were going with this, I'm not sure we can really know that there is a zero chance of either of the things you mentioned happening, either. Again, intuitively it may seem it must be impossible, but if one were to have billions of tries, maybe a very rare, unlikely occurrence might occur.

Yes, jbskaggs.

I think it's more likely that the +2700 player has a stroke and either dies or is unable to finish the game than for the game to be lost. At least GMs have been known to die at a chess board. +2700s have not been known to lose to 1300s.

Are the odds greater than 0? **Yes**.

Are they so insignificant that the odds may as well be 0? **Yes**.

If the game is part of a simul the odds greatly increase. If the game is part of a simul with at least 100 players the odds greatly increase.

If it just happens that a 2700 is paired with a 1300 say in the 3rd round of a rated tournament then my guess would be the 1300 rated player would have more than 1/2 of one percent chance.

Last month Chess Life featured a game where a 1000 rated player defeats a 2000 rated player.

Certainly giving hope to the 1700's out there. And since 1300 defeating a 1700 happens quite regularly, then by the law of symetry we can say a 1300 can defeat a 2700.

Q.E.D.

(Feel free to refer to this as "Ubik's Axiom" It really requires no evidence, nor a proof, since this is in the form of a self-evident axiomatic Euclidean truth. However, you can of course deduce further conclusions from this law, which I leave as a excercise to the student.)

In a game with 1300 and 2700 rated player

50% both players win.

50% GM Wins x2.

(ignore stalemates and draw)

SmyslovFan wrote:

Are the odds greater than 0? Yes.

Are they so insignificant that the odds may as well be 0? Yes.

The clearest and simplest answer to the original question.

Ubik42 yazmış:

Last month Chess Life featured a game where a 1000 rated player defeats a 2000 rated player.

Certainly giving hope to the 1700's out there. And since 1300 defeating a 1700 happens quite regularly, then by the law of symetry we can say a 1300 can defeat a 2700.

Q.E.D.

(Feel free to refer to this as "Ubik's Axiom" It really requires no evidence, nor a proof, since this is in the form of a self-evident axiomatic Euclidean truth. However, you can of course deduce further conclusions from this law, which I leave as a excercise to the student.)

If we accept the axiom above, the general conclusion of the topic would be flawed as well.

tieics wrote:
I think if there is 300 ELO point difference between two players it's no point playing the game. Practically the player with lower ELO has 0 winning chance.

I won against a guy rated 509 points ahead of me. With black. After 23 moves.

Ziryab wrote:

SmyslovFan wrote:

Are the odds greater than 0? Yes.

Are they so insignificant that the odds may as well be 0? Yes.

The clearest and simplest answer to the original question.

Actually, no, unless the odds are 1/infinity. I can handle 1*10^(-23).

Rasparovov wrote:

tieics wrote:
I think if there is 300 ELO point difference between two players it's no point playing the game. Practically the player with lower ELO has 0 winning chance.

I won against a guy rated 509 points ahead of me. With black. After 23 moves.

There have been upsets of1000 or so points rating difference (my personal record is 600-700). However, as previously stated, once you get to a 1400 point difference with a game between a master and an amateur the chances are insignificant.

(btw, 1/infinity is 0 as a limit, 'cause infinity's not actually a number)

I got the statistics from the FIDE website, which has a link to the Elo probability table:

http://www.chesselo.com/probabil.html