# Average opponent rating...

Icky Ike
Site Ideas 27 Sep '06 06:58
1. Icky Ike
Pro-Complainer
27 Sep '06 06:581 edit
I think everyone's info should come with a statistic that shows the average rating of the players that player has played in the last 100 games. This provides more meaning to the rating system. What does it matter if someone is a 1500 if he got there only playing 1200s? I know you could just skim through their last 300 games and try to figure it out qualitatively but this would make it much more conveinent and would help with the psychological factor that certain moves can depend on. I don't think the programming would be too hard it's a simple calculation. (r1+r2+r3+...r100)/n
2. 27 Sep '06 12:08
"What does it matter if someone is a 1500 if he got there only playing 1200s?"

The very idea with the rating system is that it just doesn't matter if you play with 12ers or 18ers. You eventually find your on rating level anyway.
3. Icky Ike
Pro-Complainer
27 Sep '06 16:171 edit
Some moves may depend on a psychological factor especially for poorer players like myself where knowing your opponents skill level can greatly help determine lines of play. Knowing what your opponent is used to can help further define their rating. Like I said the information is already there in their game history, I just want it to be indexed.
4. Ragnorak
27 Sep '06 19:24
Originally posted by ItalyBoyBlue
Some moves may depend on a psychological factor especially for poorer players like myself where knowing your opponents skill level can greatly help determine lines of play. Knowing what your opponent is used to can help further define their rating. Like I said the information is already there in their game history, I just want it to be indexed.
If a player is 1500, then he is 1500, whether he only exclusively plays 1200s or not.

D
5. Icky Ike
Pro-Complainer
27 Sep '06 19:39
I don't think you get it. Players may have either inflated their rating by playing very poor players or surpressed their ratings by competing against very difficult players. If two players had a 1400 rating and one got that rating by starting at 1200 and playing numerous 1200 players and the other got that rating by starting at 1200 and playing numerous 1800 players WHICH PLAYER DO YOU BELIEVE WOULD EXERT MORE SOPHISTICATED PLAY??? Obviously they are not equals.
6. Ramned
The Rams
27 Sep '06 19:461 edit
I love this idea, and have already mentioned it. That way it lets people know whether you play easy people or people around your level, etc...

I see ItalyBoy's logic too, mine too. It is could be inflated to a degree. As a matter of fact, I played, on another site, a person rated 1450, and went through his games and saw that he played poeple lower 1000. I creamed him.
7. DoctorScribbles
BWA Soldier
28 Sep '06 00:111 edit
Originally posted by ItalyBoyBlue
I don't think you get it. Players may have either inflated their rating by playing very poor players or surpressed their ratings by competing against very difficult players. If two players had a 1400 rating and one got that rating by starting at 1200 and playing numerous 1200 players and the other got that rating by starting at 1200 and playing numerous ...[text shortened]... ICH PLAYER DO YOU BELIEVE WOULD EXERT MORE SOPHISTICATED PLAY??? Obviously they are not equals.
You are the one who doesn't get it. The very essence of the Elo system is that it doesn't make a damn difference who you play. A 1400 player who has played only 1200 opponents and another 1400 player who has played only 1800 opponents expect to have even equity in any games between each other.

If you think otherwise, you are simply ignorant about how the Elo system works or what its ratings mean, or you are just generally stupid.
8. Icky Ike
Pro-Complainer
28 Sep '06 02:462 edits
Originally posted by DoctorScribbles
You are the one who doesn't get it. The very essence of the Elo system is that it doesn't make a damn difference who you play. A 1400 player who has played only 1200 opponents and another 1400 player who has played only 1800 opponents expect to have even equity in any games between each other.

If you think otherwise, you are simply ignorant about how the Elo system works or what its ratings mean, or you are just generally stupid.
Haha. I know EXACTLY how the ELO system works. It's a statistically derived formula that's based on the current ratings of the players and the outcome of the game. THERE IS ABSOLUTELY STILL A DIFFERENCE BETWEEN THE TWO PLAYERS I DESCRIBED only an idiot wouldn't understand. ELO is one way of describing someone's potential. It's designed to make a person come to a rating that describes their potential only if that person constantly plays opponents near or equal to their rating at every game. The system doesn't work very well when people consistently play other other players far above or below their skill level. It also has no way of showing that a person isn't just on their way to their true rating and is using every game as a stepladder on the way there. Also sometimes people intentionally will lose games to poorer players to surpress their rating and get into lower rated tournaments so that they can beat relatively easy players. Measuring the average rating of a persons last 50-100 opponents makes the ratings more specific and helps mitigate these problems. Obviously you're the ignorant one.
9. DoctorScribbles
BWA Soldier
28 Sep '06 04:082 edits
Originally posted by ItalyBoyBlue
Haha. I know EXACTLY how the ELO system works. It's a statistically derived formula that's based on the current ratings of the players and the outcome of the game. THERE IS ABSOLUTELY STILL A DIFFERENCE BETWEEN THE TWO PLAYERS I DESCRIBED only an idiot wouldn't understand. ELO is one way of describing someone's potential. It's designed to make a person ings more specific and helps mitigate these problems. Obviously you're the ignorant one.
You're wrong. Elo was a sharp enough guy that if the two players I described didn't in fact expect even equity in a way that could be accounted for by their differing histories in the manner that you suggest, then he would have simply included that factor in the rating formula.
10. XanthosNZ
Cancerous Bus Crash
28 Sep '06 05:09
ðŸ™„

Sometimes math is confusing. Give up now ItalyBoyBlue, you're confused.
11. Icky Ike
Pro-Complainer
28 Sep '06 07:45
Originally posted by XanthosNZ
ðŸ™„

Sometimes math is confusing. Give up now ItalyBoyBlue, you're confused.
I'm an engineering student who takes upper division math classes for fun because I love math. I don't need to be lectured by you two. If you don't believe me look at the formula, read about the formula, and try to think about real life - OTHERWISE F@*! OFF - I am right, there is a difference and if you can't figure out why then go ahead and not post on this string. You're useless ignorant feedback isn't helping anyone...
12. 28 Sep '06 07:571 edit
Originally posted by ItalyBoyBlue
I'm an engineering student who takes upper division math classes for fun because I love math. I don't need to be lectured by you two. If you don't believe me look at the formula, read about the formula, and try to think about real life - OTHERWISE F@*! OFF - I am right, there is a difference and if you can't figure out why then go ahead and not post on this string. You're useless ignorant feedback isn't helping anyone...
Sorry, but you're wrong here.
This is kind of embarrassing.
13. Icky Ike
Pro-Complainer
28 Sep '06 09:041 edit
Originally posted by FabianFnas
Sorry, but you're wrong here.
This is kind of embarrassing.
I'm not wrong. You guys are telling me that if two people were to play chess until both of their ratings stabilized at about a 1400 rating but one did it playing 1200's and the other playing 1800's that those two players are of equal strength?!?!? I gaurantee you you're wrong. Can't you see it's not just about the numbers, it's about practicing against advanced openings and tactics... one player would be introduced to them regularly the other would not - this is because of the difference in the skill level of their opponents. To say that both these players are equals contradicts everything that every notable world chess champion has said about becoming better at chess: PLAY AGAINST BETTER PLAYERS THAN YOURSELF!!!

Take for example the world chess champion match taking place right now. ELO's formula is set up so that if someone's rating is 200 points above an opponent's they win approximately 3/4 games with that opponent. Kramnik's rating is almost 100 points lower than Topalov and Kramnik is winning 3 to 1... THE MORAL OF THE STORY IS RATINGS LIE LIKE AMERICAN POLITICIANS FROM TEXAS WHOSE FATHERS USED TO BE THE PRESIDENT DURING THE 80's
14. 28 Sep '06 10:16
Originally posted by ItalyBoyBlue
I gaurantee you you're wrong.
So you say...ðŸ˜´
15. XanthosNZ
Cancerous Bus Crash
28 Sep '06 10:21
Originally posted by ItalyBoyBlue
I'm an engineering student who takes upper division math classes for fun because I love math. I don't need to be lectured by you two. If you don't believe me look at the formula, read about the formula, and try to think about real life - OTHERWISE F@*! OFF - I am right, there is a difference and if you can't figure out why then go ahead and not post on this string. You're useless ignorant feedback isn't helping anyone...
Don't try and pull the "I know more math than you do" card on me. I'm an engineering student as well. I've also spent a decent amount of time modelling behaviour of the ELO formula for various things such as what rating becomes if a player of true rating 1800 plays nothing but players of rating (true and actual) of 1400. The answer? 1800. And what happens if the same player switched and started playing against players of true and actual rating 2000? Still 1800.
The strength of the opponents shouldn't matter (unless you are reaching your floor or ceiling against that opposition [around 700 difference]), against weaker opposition you win more games but get less points per win.
While you may argue that players whose true rating differs from their actual rating skew everyone's ratings the system is based so that any player not playing at their rating level quickly returns to their true play level by taking points from others (who would then take points from others and so on) or giving points to others depending on which way they need to shift. The rating pool changes slightly but order is quickly restored.

You mention a single match as evidence that the ELO system is flawed. But you surely know that the ELO difference will give the average long term behaviour and doesn't claim to be able to predict the results of a single game?