### Forums

### Most Recent

- Where I can find free downloadable chess engines rating ELO 1600-1900
- How do u find a girlfriend on chess.com?
- Is the Sicilian meant for chess experts only?
- 2/27/2017 - Where There's A Will, There's A Way
- The Top 5 Most Underrated Chess Openings!
- Share your most embarrassing games here
- game abortion
- Which rating pool is an absolute beast and hardest to crack !?
- Would prime Karpov have beaten Kasparov?
- When do you know the right time to resign?

### Forum Legend

- Following
- New Comments
- Locked Topic
- Pinned Topic

there is a chance if the lower rated player had another very strong player subtuting for them

I doubt that you could convince the stronger player to play enough games as it would take before he would lose a game but I think that if they played for a year or so (say 12 games a day) eventually the 2700 player would be so bored that he might start experimenting and then by the end of the second year finally lose a game.He would likely be glad that the chore was over by that time.(For that matter it may have driven him to drink by then which would also help).

What needs to be factored in is the ability of the 1300 to improve

after 3000 games :)

lol

Aha!

That'swhat those initials really stand for!@ Dude_3 a.k.a the 10 year old kid who is correcting me.

There was absolutely nothing wrong with my calculations! ... simply because I didn't make any. You are quite right though, my reasoning was flawed, the odds of tossing 100 heads are much greater than a random mover beating a GM. As I am a renowned genius, the only possible excuse would be that I was tired. :)

Of course, the difference being the number of available permutations. With a coin throw we have a 0.5 probability on each 'move' of hitting the head, whereas we could estimate in every chess position there are on average around 30 bad moves which can be played, and maybe just a couple of moves of a high enough calibre to pose the GM a challenge.. i.e. a 0.07 probability on each move.

My answer to RetiFan's original question is "no", at 2700 vs 1300 it will be 100-0. Below a limit of approx. 2035 vs 1300 the weaker might have a chance, according to expected scores as follows :

2035-1920 vs 1300 score 99-1

1920-1860 vs 1300 score 98-2

1860-1820 vs 1300 score 97-3

You can find this list ( 51 lines ) in Hamlin's Dictionary of Chess, for example also :

1545-1535 vs 1300 score 8 - 2 at 10 games,

1415-1405 vs 1300 score 6-1/2 - 3-1/2

1340-1335 vs 1300 score 5-1/2 - 4-1/2

Some matches, a.o. Fisher-Spassky ended according to this list.

Of course, this list is just something someone once calculated, but essentially such list can be correct, based on played matches.

i would beg to differ on said statistics its possible however highly improbible these numbers dont mean anything it could be never or more than or barely a few the questionw as simply is it possible and the answer is yes

The list is based on the rating difference between players and gives an expected score. A score of 8-2 is expected by approx. 240 points difference, e.g. also for 2320 vs 2080 and 1140 vs 900. And yes, I am the first one to admit that this (old) list needs an up to date revision, based on facts (results of matches). And two tolerances have to be taken into account, the 240 points at around 1400-1500 level might have another result than at 2400-2500 level ; furthermore such score as 6-4 can easily become half a point higher or lower. The best is to draw a graph, with lines in it.

Being a Dutchman, I have to translate to understand texts like yours, and having some experience I prefer clear, plain lines, so reading --quote-- it could be "never" or "more than" or "barely a few" --unquote-- is difficult to understand what exactly you mean.

Nevertheless, thanks for your reaction, this makes Chess-dot-com an interesting site !

@RomyGer

Don't worry, I don't understand him either.

I'd guess this is roughly equivalent of sending your sunday league pub team to the Neu Camp to face FC Barcelona. So really, a percentage which approximates to zero.

I'd guess this is roughly equivalent of sending your sunday league pub team to the Neu Camp to face FC Barcelona. So really, a percentage which approximates to zero.

with some of today's refs......

To be a reminder; the actual calculation indicates that odds are

0.0125%(1 game in 10000).That 1 in 10000 is going to be something like the GM faints at the board. Not that he genuinely gets outplayed.

No that is an actual level of odds that 1300 rated player catches a GM level mistake, 2700 doesn't play perfect but close to it you know.

No that is an actual level of odds that 1300 rated player catches a GM level mistake, 2700 doesn't play perfect but close to it you know.

I think about the only chance is to catch him out with an opening novelty which the GM doesn't know or underestimates. Anything over 10 moves and the 1300 wll be so far behind he wont be able to win with an extra queen.

No that is an actual level of odds that 1300 rated player catches a GM level mistake, 2700 doesn't play perfect but close to it you know.

I think about the only chance is to catch him out with an opening novelty which the GM doesn't know or underestimates. Anything over 10 moves and the 1300 wll be so far behind he wont be able to win with an extra queen.

My probability function also says that the chances are decreasing when game is getting longer and longer. But the problem is finding an underestimated novelty against 2700 rated player is probably harder, they just memorize the moves of the opening you choose .

Nevertheless, if there is a moment to strike at the GM, it is at the opening .

@RetiFan, I'm not convinced that the ELO expectancy curve means much at the extremes. You can use it to roughly predict the score between a 2700 and a 2500 (75-25), but between a 2700 and a 1300 the curve is so close to straightening out that it may as well be straight.

I'm wondering:

Will the 1300 have a better chance if:

"This whole topic is just silly. Just because something is theoretically possible , it doesnt mean it can actually happen."

Ok, I fully admit this is pure trolling, but, I think when you say "can actually happen," you really mean "

willactually happen," right? Because "theoretically possible" is defined in that it can happen.Again, I'm trying to be annoying on purpose, so don't take it personally :)

This whole topic is just silly. Just because something is theoretically possible , it doesnt mean it can actually happen. It is theoretically possible to flip a coin 500 times in a row and have it come up heads each time. In reality entropy will reach maximum , the universe will go dark and end before something like this could happen.

If a 2700 player plays like a 2700 player there is no chance, none, zero, nada. Now if you want to suppose something silly like the 1300 player shoots the 2700 player and kills him in the middle of the game, then yeah the 1300 player will win.

If I was sleep deprived, drunk, distracted, not thinking clearly and generally off in the ozone, I still couldnt loose to a 1300 player and that is no disprespect to the 1300 player. Its kind of like all five foot six of me playing Lebron James in basketball and winning. Unless he gets hit by a bus forget about it.

Cheers, Becky

PS. im talking about otb games. Online doesnt count for diddly.

Unfortunately, let Tails be T and Heads be H:

I threw a coin 18 times. The sequence THTTHHTHHHTTHTHTTH, in this order, have 1/(2^18) chance of appearing, but this was what has happened! So I don't buy your explanation either.