Go back
Rating guru help? win % and such vs ratings.

Rating guru help? win % and such vs ratings.

Only Chess

s
Fast and Curious

slatington, pa, usa

Joined
28 Dec 04
Moves
53321
Clock
22 Nov 12
Vote Up
Vote Down

I have been playing some teaching games against a guy rated about 400 points lower and it seems maybe he has learned a bit, his rating has gone up about 100 points.

My question is this: I am about 1650 and he averages about 1250 so a 400 point dif.

So far we have played 57 games where he won one game and lost the rest.

Given our 400 point difference, am I under performing for that rating difference?

Also, just on the math involved, would the win or lose % be different for that pairing vs if I played the same number of games against MY 400 point difference, in this case, me playing a 2050 player, do the numbers work out exactly the same or are there differences based on the different numbers even though the difference is the same?

So it looks like one loss out of 57 games gives me a loss number of 1.75 percent or thereabouts. Is that where I should be given the rating difference of 400 points?

LB

Joined
25 Jun 10
Moves
12789
Clock
22 Nov 12
Vote Up
Vote Down

Originally posted by sonhouse
I have been playing some teaching games against a guy rated about 400 points lower and it seems maybe he has learned a bit, his rating has gone up about 100 points.

My question is this: I am about 1650 and he averages about 1250 so a 400 point dif.

So far we have played 57 games where he won one game and lost the rest.

Given our 400 point differenc ...[text shortened]... .75 percent or thereabouts. Is that where I should be given the rating difference of 400 points?
FAQ

According to the above formula the "win expectancy" with a 400 point difference is 0.91 or 91% so you are ahead of expectation unless there are draws you have not told us about. In every ten games he should win one or draw two on average.

s
Fast and Curious

slatington, pa, usa

Joined
28 Dec 04
Moves
53321
Clock
22 Nov 12
Vote Up
Vote Down

Originally posted by Long Barrow
FAQ

According to the above formula the "win expectancy" with a 400 point difference is 0.91 or 91% so you are ahead of expectation unless there are draws you have not told us about. In every ten games he should win one or draw two on average.
No draws, just wins and losses. Are the numbers exactly the same for the opposite, where I am 1650 and opponent is 2050?

LB

Joined
25 Jun 10
Moves
12789
Clock
22 Nov 12
Vote Up
Vote Down

Originally posted by sonhouse
No draws, just wins and losses. Are the numbers exactly the same for the opposite, where I am 1650 and opponent is 2050?
Yes, at 2100+ the formula awards elo points differently but as far as I understand it the win expectancy remains the same.

h
peacedog's keeper

Joined
15 Jan 11
Moves
13975
Clock
22 Nov 12
Vote Up
Vote Down

Although in theory the win/loss ratio between 1200-1600 and 1600-2000 should be the same, it most likely isn't in reality. The higher up you go the harder it is to gain points. A difference of 100 points is meaningless sub 2000. But over it shows quite a difference in strength.

g
Mad Murdock

I forgot

Joined
05 May 05
Moves
20526
Clock
22 Nov 12
1 edit

I have read an interesting article about that a while ago. Here is the link:

http://www.chessbase.com/newsdetail.asp?newsid=7114

HikaruShindo

Joined
08 Apr 12
Moves
68553
Clock
22 Nov 12
Vote Up
Vote Down

Thanks for the article, gambit05.

s
Fast and Curious

slatington, pa, usa

Joined
28 Dec 04
Moves
53321
Clock
23 Nov 12
Vote Up
Vote Down

Originally posted by HikaruShindo
Thanks for the article, gambit05.
Very interesting. I wonder if FIDE will heed this new analysis and include it in the rating system?

Cookies help us deliver our Services. By using our Services or clicking I agree, you agree to our use of cookies. Learn More.