Hey guys I have seen an 800 beat a 2400!!
...and then you woke up, or not?
no i am serious i have seen an 800 beat a 2400, i saw it!!! his name is george something he is rated 800 and his best win is a 2400
lol, george something.
You can figure out mathematically what the chance to win is using the rating differential chart. Of course it is a small number, but it is not zero.
There is no force that physically prevents the 1300 from choosing the best move in a position (even for the wrong reasons). And nothing prevents the GM from blundering, even repeatedly.
Even .0001% means they just need to play 10,000 games and the 1300 will come away with a win. Winning 10,000 games in a row is a pretty tall order, no matter what your rating is. If you dont like .0001%, then how about .00001%?
I did see an 800 player beat a master once. It was in a junior tournament. The master dropped a piece in the first ten moves or so and resigned. I'll see if I can dig it up. It's in chessbase.
But I think the ratings in that game were provisional. I'll see if I can find it.
lol, why on earth would he resign in that situation? I actually posted a topic asking if you would accept a draw offered by an 800 if you were down a piece against him. The purist in me wanted to take a draw, but now, I don't know what I would do. Houdini could probably beat me a piece down, so why can't I beat an 800 
I did see an 800 player beat a master once. It was in a junior tournament. The master dropped a piece in the first ten moves or so and resigned. I'll see if I can dig it up. It's in chessbase.
But I think the ratings in that game were provisional. I'll see if I can find it.
Mama Ubik taught all the little Ubiks never to resign against a player below 1000.
I've carefully read all the above comments and balanced them
objectively. After long deliberation by the midnight oil and
conference with my colleagues and experts in the field, I've arrived
at the following conclusion: Lollerskates.
For a long time my best win here was against a 2105, he lost a piece and resigned on move 7. I dont know if I could have won if he played on.
From the view of the 2700 rated person, if there is this slight chance that (s)he can slip, and if (s)he is playing against 1300 rated players often, (s)he surely will lose(Murphy's Law).
Firebrand's example is interesting. We all probably think we are "untouchable" against someone 600 points lower rated or something, but if you were suddenly told you have to beat (can't draw!) this one guy (with such a rating differential) one hundred times in a row, in some marathon, you might start to find yourself daunted. Sure, normally things won't go wrong against such an opponent, but if you have to be that consistent, you really don't have room for Murphy's Law.
Or what if there was an experiment where instead of a hundred games, it was a thousand or ten thousand games or something, played over the course of a lifetime at some sort of rate. Your lower rated opponent would know what he has to do. He is confident that there is bound to be a chance somewhere -- maybe you get careless and misplay the opening, not realizing your move isn't so strong after all, and the determined -600 opponent capitalizes and is proud of himself -- now he has an advantage.
I bet over the course of ten thousand games, that kind of thing would inevitably happen -- maybe most of the time you would wiggle out, but there have got to be certain days where you won't be feeling as confident, or get temporarily blind, and you may continue to falter. It would just be immense pressure to be so consistent that you could always come out with a win, even if the position gets drawish, even if you blunder, when you are playing an absolutely absurd amount of games.
In this question's case, maybe it would have to be more than ten thousand, maybe a million, billion (amounts that couldn't be played in a lifetime), or more. But I could see there being a point where you would expect the 1300 to win once.
Of course there's an issue here:
After losing 100 games, is the 1300 still a 1300? Won't he now maybe be beating 1500's if he wasn't so tied up in this match?
I think it has to be a different 1300 every time.
Yes, I am assuming both players's strengths remain constant. It's actually not so implausible to have a 1300 who plays a lot and doesn't improve -- maybe he doesn't even study his games so much or has a serious hope to improve, but just feels like playing the stronger player and seeing if he can get the upset. Of course, the 1300 would learn more about the 2700's style of play, but it could also be argued the 2700 would learn the 1300's style of play better too, to figure out the best way to crush him. So that's the sort of picture I had in my mind.
But yeah, maybe the challenge could just be the 2700 having to play a certain amount of 1300s every day, a different one each time.
Christ, this is still going?
NO!
You can figure out mathematically what the chance to win is using the rating differential chart. Of course it is a small number, but it is not zero.
There is no force that physically prevents the 1300 from choosing the best move in a position (even for the wrong reasons). And nothing prevents the GM from blundering, even repeatedly.
Even .0001% means they just need to play 10,000 games and the 1300 will come away with a win. Winning 10,000 games in a row is a pretty tall order, no matter what your rating is. If you dont like .0001%, then how about .00001%?
A couple of points statistically speaking: first, there is no real way to measure the chance of a 1300 beating anyone over 1700. The chance is as close to zero as it can get at that point (in Elo system), so it is not correct to assume there is "less" of a chance for someone three times the maximum deviation.
And in those maximum+ cases, there is only a "theoretical" chance; it is so low it won't actually happen. The number of games is not a real thing; the astronomical odds are the same for each individual game.
Remember, this isn't a contest of blitz or rapid games, the original question concerned classical chess games.
Perhaps the clearest illustration of this is that it has never happened and, in the real world, it will never happen.
The astronomical odds are the same for each game, yes, but if you play an astronomical number of games...statistically it becomes likely.
The odds against winning the lottery are astronomical for each individual. Yet, someone still manages to win.
Truly, a trained chimpanzee moving pieces randomly could win also, though they should probably pack a flashlight so they can continue playing after the sun has gone nova.
Don't believe that? Picture 20 games with a chimp vs. a GM. Odds are one of the twenty games a chimp has opened with 1. d4 . Black's game could already be in its last throes. Anything to prevent that chimp from playing 2. c4 next?
hmm yes good point, can't believe no-one else has made that point in 1000 posts. Incredible.
Of course, this forum topic can't be locked because this question will always be repeated by different 1300 rated players and this contributes to a test of mine which I will reveal now:
- I made some calculations using a fair level of mathematics involved, and found that the chances are 1 in 8000 which I shared with you guys a 1000 posts ago.
- Still, I wasn't sure that if there is this little chance or not, or actually calculations are wrong in this marginal situation and therefore there could be a better chance.
- The best way of dealing with this uncertainty, is to learn infinite ideas from an inifinite number of comments which eventually lead me to a decision if 1/8000 is final.
- Most of you guys believe that there is this theoretical chance if the 2700 guy doesn't sleep(thanks to IM pfren for starting this classic response which I saw a couple times after him), however 1300 guy doesn't get the chance to exploit this by playing so many games against masters.
- This is the conclusion of this topic until a new one erases it.
- Maybe there is another argument which is better than this, who knows?...
1. A seven year old with tons of talent makes it to a 1300 rating.
2. He studies chess for years; tweaking his system of thinking, studying endgames, strategy, tactics, etc., and in 10 or so years basically masters the game to a great degree.
3. Then he plays a 2700, who somehow blunders in time pressure. Yes, the kid is officially 1300 but it doesn't reflect his current playing strength.
You know what, this is ludicrous. A better question would be can they beat an expert, and if they do win they're clearly not 1300 strength anymore and need to play more games.
Or he could inattentively pull a Kramnik -- could.
I think there is a youtube video somewhere (and I watched a tv program about this) that instructs you to do something like count the amount of dribbles and passes a group of four or five basketball players make. And then as you're doing that a guy dressed as a gorilla or something dances around for a few seconds in the middle of it all, and most people don't notice it because their brain is too focused on something else. (By the way, blocking things out is a good thing -- it's what makes us only pay attention to what is important, but it can sometimes lead to strange misses of obvious but unexpected things)
Remember, this is a guy in a gorilla suit -- you can't get more distinct and obvious than that. But stuff like that can be missed just based on what your mind is paying attention to. You can miss obvious things if you're not looking for them. That kind of inattentive blunder that Petrosian and Kramnik made reminded me of that exercise -- they managed to play a huge blunder despite how much they, and any non-beginner, know better. That is a human factor and it can happen to anyone, no matter how rarely.
That can apply to any position, even trivial ones, even allowing a beginner's stalemate when you are up tons of material -- that may not happen in most chess player's lifetimes, but that doesn't mean it is simply impossible.