Is there any chance that a 1300 rated player can beat a 2700 rated player?

Clever method!

e^((2700-1300)/100)?? Where does it come from? I'm interested because my initial calculation showed 1 win in 8000 games.

the 2700 might forget to eat for a few days and faint during the game 

It seems pretty arbitrary, Retifan, e^rating difference / 100.

Without using the childish infiniti type of "anything is possible" arguement, there is NO WAY a 1300 beats a 2700. No chance.

@ Krames, amen.

Wrong, Krames. It is possible, and the odds are likely not dissimilar from those of winning the lottery. Personally, I consider those odds not worth bothering with - so I don't do the lottery, but that doesn't mean people don't win it. 

There do exist things which are impossible (probability - 0), this is not one of them.

p.s. if that's childish, maybe an adult should be able to spell argument?
 

Tmb86 , if you compare a game of chess with a lottery, you are essentially saying a game of chess is no different from a flip of a coin, or a series of flips of a coin, which is not the case. Chance may play a great role in a game of chess played between very low rated players who do not understand chess beyond basic moves, but nobody who can play chess logically is going to make purely random moves. Not even the 1300, and certainly not the 2700.

Tmb86, I don't think you can press probability into the service you've tried to in this case.  I think there is something different in saying "Anand draws 99% of his games" and "99% of people who play Anand should expect to draw him."  Stats relating to GM performance are not going to correspond to games they play against players in different strata.  

Beyond all that, I don't think "chance" or "probability" in the sense you're describing (e.g. comparing chess to the lottery) is a legitimate way of discussing chess.  Any single game of chess is devoid of chance in any real sense, in that both players are possessed of equal information and have even control of the outcome of the game.  There is no roll of the dice or selection of numbers to be matched by a random draw.  

In short Tmb, I think what you've said is basically a non-sequitur.

Beck15

The point is, there exists a probability of simply outplaying a 2700 by sheer luck alone. Quite clearly that probability is miniscule, but it is non-zero. 

Galileo182

I have discussed this point at length already. The element of skill involved in chess only serves to increase the likelihood that a 1300 will beat the 2700, as compared to a random move generator (which will also eventually win).

Also, I wish I could remember the quote - but I remember a GM stating how few people realise how much luck is involved in chess. 

No, not really, the skill differential is far larger than the luck differential

TMB...I think we can all agree with the notion your are presenting, you just need to stop calling it probability.  It is logically permissible, but I don't think there is any objective way probability can be used to assign a value 0-1 to any single game between a super GM and a 1300 player.  

As for the quote about luck, I don't think it would prove anything anyway.  I would be willing to bet the GM is not talking about over the board play, but rather the circumstances surrounding a game.  The objective quality of a move's accuracy is almost certainly not that to which the GM is ascribing the importance of luck or chance.  I wish I were a GM so I could say this with absolute certainty, but lacking any title as I do, I am still confident GMs do not make moves without a reason.  Reason is the opposite of chance in this context.

"No, not really, the skill differential is far larger than the luck differential"

Ahh, ok. 

galileo,

Thanks, and I understand your point as well. That is why I thought it was clearer to put the GM against a random move generater, and therefore people could see that despite all the intricacies and skill in chess, there exists a chance of blindly knocking out the perfect game. There will also exist a fairly definite probability of this happening, not perfectly definite, because it depends on:

a) the length of the game
b) the number of available moves in each position and
c) the number of moves of a high enough calibre in each position.

As I have said before, I am no expert in this area - I cannot even recall the equation to calculate it. But it would be of a similar form to that of calculating the probability of tossing 10 heads in a row. The differences being:

a) we don't know how many heads are needed,
b) we don't know how many faces the coin has, and
c) we don't know how many heads there are.

Adding the 1300 into the mix complicates things, and it is arguable as to whether he will achieve the feat before the generator. 

I'm inclined to think chess is more deterministic than probabilistic. It's true that the outcome of all positions is beyond mankind's ken as of yet, but that does not make the game probabilistic (there is no throw of the dice). In such a case, you cannot simply go about applying the rules of probability, at least not when the game is played by strong players whose moves are not slaves of random chance.

Edit: I understand what you are saying, but in your case where a random move generator playes a successive strings of perfect moves from start to end is not really a 1300 player. Sorry, does not apply to the thread.

...

Beck15: Although there are disadvantages to being a 1300 compared to a generator, keep in mind that there are advantages as well. For example, a 1300 is more likely to take advantage of a missed mate in 1. So it's not so simple.

So basically, all the 1300 would have to do is make the 2700 scared of him. Hmm...

"Resign right now, or your family shall die."?

I'm pretty sure that is against the USCF/FIDE rules...

(Dang it. :( )

Hey don't be weird. It's more like "Resign now, or the puppy gets it!"

I think USCF regards stating a fact as legal. We all shall die eventually.

Elubas, talking during the game is discouraged according to USCF regulation 20G Annoying behavior prohibited and 20I The director has the option of banning all talking in the tournament room, among others.