Go back
There are ratings and then there are ratings...

There are ratings and then there are ratings...

Only Chess

M

Joined
30 Dec 04
Moves
8216
Clock
22 Feb 05
Vote Up
Vote Down

No, this is not another 'my rating is x points more/less than my USCF rating so this system needs fixing' rant.

I just noticed something interesting about how the ratings work. Perhaps others have but I didn't see any posts on this.

I got my rating up to about 1591 and then entered a 1400-1599 tourney thinking I had a chance. I quickly found that people in the 1500-1550 range were regularly cleaning my clock with embarassing ease.

After examining my record, I found that I had built up my ranking by playing a lot of players that were 150-200 pts below me. Not by choice, they just happened to be the ones I got matched up with for some reason. As it turns out, all you need to do is average wining 74% of your games against players 200 pts below you to break even. So, I figure I could probably get my ranking up to close to 1650 by playing people in that range. The problem is, I'm not a 1650 player! I'm probably not even a 1550 player based on my recent results.

So the point of all of this is, that playing people well below your ability will artificialy inflate your score. So if that's your goal, great. But it won't make you a better player which is what I'm trying to do and what I think most of us want. A player with a legitimate 1550 rating earned by playing people in his range will probably beat a player with an inflated 1650 rating.

IMHO.

A

Joined
12 Jul 04
Moves
3836
Clock
22 Feb 05
Vote Up
Vote Down

Originally posted by Mavraam
No, this is not another 'my rating is x points more/less than my USCF rating so this system needs fixing' rant.

I just noticed something interesting about how the ratings work. Perhaps others have but I didn't see any posts on this.

I got my rating up to about 1591 and then entered a 1400-1599 tourney thinking I had a chance. I quickly found that ...[text shortened]... y playing people in his range will probably beat a player with an inflated 1650 rating.

IMHO.
Amen... Ratings mean nothing! I have scalps of people rated much higher than me. I don't even look at that number anymore but consciously try to play people whose game is interesting, and stronger than mine. I also look at the percentage of wins whenever I agree to play someone and his/her opposition.

L

Joined
13 Oct 04
Moves
7902
Clock
22 Feb 05
Vote Up
Vote Down

The big difference between playing here and playing otb is that here you can pick your opponents.
When you play a rated OTB tournament you will be playing in your own rating class so you will always play people of about equal strength.
Here I could probably get to 2000 only playing people rated below my strength, but a real 2000 player will still kick my @s.

A

Joined
12 Jul 04
Moves
3836
Clock
22 Feb 05
Vote Up
Vote Down

Originally posted by LordOfTheChessboard
The big difference between playing here and playing otb is that here you can pick your opponents.
When you play a rated OTB tournament you will be playing in your own rating class so you will always play people of about equal strength.
Here I could probably get to 2000 only playing people rated below my strength, but a real 2000 player will still kick my @s.
OTB I play the reserve section whenever I need money 🙂 Otherwise I try to play in the open section and have no choice but to play higher rated players. It depends on what goals you have...

Whenever I need to review a theoretical line I try to find the strongest exponent here at rhp. Despite quite a few defeats, I have learned a lot and it has saved me points in OTB practice.

However, I shall go back to the statement that nobody should be intimidated by a number and play hard no matter what. By the same token, you shouldn't be cocky and take for granted your lower rated opponent.

TSD
The 3rd Coming

London

Joined
23 Aug 04
Moves
25775
Clock
22 Feb 05
Vote Up
Vote Down

Originally posted by Mavraam
So the point of all of this is, that playing people well below your ability will artificialy inflate your score. So if that's your goal, great. But it won't make you a better player which is what I'm trying to do and what I think most of us want. A player with a legitimate 1550 rating earned by playing people in his range will probably beat a player with an inflated 1650 rating.

IMHO.
One of the natural balances to this is occurs in tournaments where early success leads to more difficult play in later rounds.

W
NONE

WORK

Joined
07 Jan 05
Moves
38272
Clock
23 Feb 05
Vote Up
Vote Down

Mavramm I think I might have the same problem.
When I was a little (p) player I usually played 1200 type players and that was what most of my games were. Im at 1561 right now but I feel that might be a bit inflated. I guess I need to set my goals higher!
It sucks knowing that my rating is going to go down but I had better pony up soon and play some harder players.

peace.

Chuck

AThousandYoung
1st Dan TKD Kukkiwon

tinyurl.com/2te6yzdu

Joined
23 Aug 04
Moves
26754
Clock
23 Feb 05
2 edits
Vote Up
Vote Down

Originally posted by Mavraam
No, this is not another 'my rating is x points more/less than my USCF rating so this system needs fixing' rant.

I just noticed something interesting about how the ratings work. Perhaps others have but I didn't see any posts on this. ...[text shortened]... will probably beat a player with an inflated 1650 rating.

IMHO.
The ratings are not perfect, no. However this effect you're describing only goes so far.

I guess this might be the advantage the CC is attempting to use to win clan challenges more than 50% of the time...deliberately and methodically stacking the odds in their favor while appearing on the face of things to have the matches be 'even' (or 'fair' as rauder calls it). This could allow them or anyone else to beat the statistics which predict even rated people will have an even win/loss ratio by adding in a hidden variable that favors them; artificially keeping their ratings low by avoiding playing lower rated people. Interesting.

Obviously if this factor is deliberately manipulated, matching only ratings fails to achieve the goal that people claim to have; making sure the two players are equally skilled in chess.

f
Quack Quack Quack !

Chesstralia

Joined
18 Aug 03
Moves
54533
Clock
23 Feb 05
3 edits
Vote Up
Vote Down

Originally posted by Mavraam
....... After examining my record, I found that I had built up my ranking by playing a lot of players that were 150-200 pts below me. Not by choice, they just happened to be the ones I got matched up with for some reason. As it turns out, all you need to do is average wining 74% of your games against players 200 pts below you to break even. So, I figure I could probably get my ranking up to close to 1650 by playing people in that range. The problem is, I'm not a 1650 player! I'm probably not even a 1550 player based on my recent results. ........


i do not see what you are talking about.
when i look at your ratings graph it looks as though you built your ratiung up to 1601 by beating players rated:
1534, 1555, 1490, 1609, 1457, 1408, 1558, and 1451.
i think you should consider yourself to be a 1600 player!
your rating is presently 1577.
you can still expect to lose sometimes to 1430 players.

f
Quack Quack Quack !

Chesstralia

Joined
18 Aug 03
Moves
54533
Clock
23 Feb 05
4 edits
Vote Up
Vote Down

Originally posted by LordOfTheChessboard
......... Here I could probably get to 2000 only playing people rated below my strength, but a real 2000 player will still kick my @s.
hmmmm, i would like to see you try.
Game 867525
Game 826200
Game 813741

you might find it just as easy: Game 843389

L

Joined
13 Oct 04
Moves
7902
Clock
23 Feb 05
1 edit
Vote Up
Vote Down

Originally posted by flexmore
hmmmm, i would like to see you try.
Game 867525
Game 826200
Game 813741

you might find it just as easy: Game 843389
lol, ....Aspviper is very underrated and he has not played a lot of games here. He easily plays at A class strength (1800-2000)
I could have won the game against the chess express but he really wanted a draw (I don't remember why)
And the game against brakolak does not need an explanation. I had a won possition there but gave it away because I did not take enough time.
Those aren't excusses, just look at aspvipers rating in a couple of months and you can ask the Chess express, im also willing to finnish that game.

Anyway maybe I can't get to 2000 but I can definitly inflate my rating, anyone can.

LB

Joined
02 Nov 04
Moves
2174
Clock
23 Feb 05
Vote Up
Vote Down

Rtg is not a perfect measure of your realy power, of course, but can show your range, with some borders. 100 pts more or less can be casual, but 200 pts is enoguh to say it is another class. When Elo-system was introduced by FIDE, at the begining of 1970s, 200 pts was drafted as an level: 2600 for GM, 2400 IM, 2200 master.

My personal rtg here is real. I won my first 15 games, mostly against oponents <1600 pts, and my rtg arose >1700. Then, I start to play agaist much stronger oponents, mostly 1600-1800 pts, and in the nxt 18 games my result is +8-6=5. So, around 1750 is my real rtg.

o
Paralysed analyst

On a ship of fools

Joined
26 May 04
Moves
25780
Clock
24 Feb 05
Vote Up
Vote Down

I find my rating oscillates around a certain point mainly because of clan match ups, which are reasonably even with whatever rating I had at the time.

When I go down a bit, I get matched against players I'm capable of beating (on average), so I rise again, but after a while I'm therefore assigned matchups against higher people which (again, on average) I lose, pushing me back down a bit...

I'd like to thing I'm improving a bit though, because I managed to sustain a rating above 1200 for quite a while before a recent dip, which I recovered from the other day by beating 2 players ranked above me.

c

Joined
03 Feb 05
Moves
59458
Clock
24 Feb 05
Vote Up
Vote Down

I am new to chess and this site, so I looked up how ratings are calculatated (can't help it...once a mathematician, always a mathematician). Surprisingly, the formula given shows that when two relatively equal players play each other repeatedly, both players ratings will go up over time, assuming each wins his share of games:

Assuming both players are rated at 1600:

For a win, New Rating = 1629
For a loss, New Rating = 1597
For a draw, New Rating = 1613

Regardless of what Greenspan says, it appears that inflation has returned! What am I missing here?

Camilli

AThousandYoung
1st Dan TKD Kukkiwon

tinyurl.com/2te6yzdu

Joined
23 Aug 04
Moves
26754
Clock
24 Feb 05
Vote Up
Vote Down

Originally posted by camilli
I am new to chess and this site, so I looked up how ratings are calculatated (can't help it...once a mathematician, always a mathematician). Surprisingly, the formula given shows that when two relatively equal players play each other repeatedly, both players ratings will go up over time, assuming each wins his share of games:

Assuming both players are r ...[text shortened]... Greenspan says, it appears that inflation has returned! What am I missing here?

Camilli

Interesting.

Well, both players are getting a lot of practice in, so it does kind of make sense...

TSD
The 3rd Coming

London

Joined
23 Aug 04
Moves
25775
Clock
24 Feb 05
Vote Up
Vote Down

Originally posted by camilli

Regardless of what Greenspan says, it appears that inflation has returned! What am I missing here?

Camilli

Some thing has gone awry with your cal here (probably a miss placed bracket in this confusing equation).

The mechanics of the cal means that the winner's rating will always go up the by the number of points that the loser drops.

With a maximum win/loss of 32 points, but only 16 points when the two player have equal ratings.

Here try puting this formula into excel for the new rating (rounding not included):

new rating =rating_mine+32*(score-(1/(10^((rating_op-rating_mine)/400)+1)))

Regards, Dave

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