You think? If the person's rating was inflated, then you really haven't beat somebody with a "2100".
It's probably best to provide a "better" measure of your opponent's rating, such that you could better assess your own performance.
Ratings..
Sort:
Jul 1, 2008
0
#3
the most accurate rating is the most RECENT rating. that said, in the BEST WIN stat we can probably change that at some point :)
Jul 1, 2008
0
#4
erik wrote: the most accurate rating is the most RECENT rating. that said, in the BEST WIN stat we can probably change that at some point :)
Okay, I can accept that as reasonable. Thanks!
Jul 1, 2008
0
#5
gdadson wrote:
You think? If the person's rating was inflated, then you really haven't beat somebody with a "2100".
It's probably best to provide a "better" measure of your opponent's rating, such that you could better assess your own performance.
Well, one can't be sure all the time if a rating is inflated, so who knows really what the "better" measure is, it can't be determined just by opinion or guesswork. So I guess we'er stuck with the posted rating, inflated or not.
Jul 1, 2008
0
#6
Magna_Canticle_XII wrote: gdadson wrote:
You think? If the person's rating was inflated, then you really haven't beat somebody with a "2100".
It's probably best to provide a "better" measure of your opponent's rating, such that you could better assess your own performance.
Well, one can't be sure all the time if a rating is inflated, so who knows really what the "better" measure is, it can't be determined just by opinion or guesswork. So I guess we'er stuck with the posted rating, inflated or not.
The problem is that you're not stuck with the posted rating...
Your opponent might have gained a few ELO since you played...
Jul 1, 2008
0
#7
I brought this up awhile back, especially with regards to "best beaten". It just seemed counterintuitive to put the post-rating value there. One poster in that thread tried devil's advocate, and came up with some reasonable ideas, but ultimately I am still of the opinion that if I play a 2100 and beat them, and their rating goes down to 2050, then I have beaten a 2100, not a 2050.
Yes, Erik, of course though 2050 represents the best estimate of that player's rating, of course.
Jul 1, 2008
0
#8
Exactly right, Ozzie and torqamada...
The rating after the game may very well be a good estimate of his or her rating, but it wasn't that estimated rating I was playing. So therefore, I would like it reflected, the "estimated" rating of my opponent the second before I beat them, in my stats.
Why don't store both ratings, after and before the game?. I know this require a redesign of the database but if we are in a beta version this is the right moment, at least I think so ;).
Jul 2, 2008
0
#11
tkanzakic wrote:
Why don't store both ratings, after and before the game?. I know this require a redesign of the database but if we are in a beta version this is the right moment, at least I think so ;).
because every time you add a column, you make the DB that much slower :)
Yes, that's true, but how this fields only will be used when a new game start and when you are consulting your stats, you can put it in a different table into the database, this way you don't have a significant change in the performace and give this new functionality. that's just a suggestion, I don't how is the actual database design.
Jul 2, 2008
0
#13
Erik makes a good point about speed of database operation, and none of us would want a slow down. I think if we just give them time to reprogram the one stat, it will get down and satisfy most all of us.
I tell you before that I have no idea about the actual database design, I give a suggestion to satisfy your demand, and don't think that this change to the database will affect the performance of the web site.
i've "drawn" a game against a lower rated opponent in a tourny and it so frustrating to see my pts.and rating affected though i was up in material, and he was the one who initiated the draw. even before the start of the tournament he was already out of our rating range and i think it's unfair for us 'cause we can't choose our opponents.im sure this is one of the reasons why most games take too long to finish, "finding a way to go through the flaws of our rating system". i don't mind losing points on a draw but not when i'm up in material & my opponent initiated the draw, or maybe implement "draw is draw" rule no rating and pts. diduction.had i known; i should have joined the higher rating range insteed(when i was within its rating).if i'd lose,i'd lose fewer pts.playing higher rated opponents and i'd gain pts. if i draw against higher rated..
Did you accept the draw proposition?, if you did it and your oponent has a lower rating than you it's right that you lose some points on your rating, no matter how bigger could be your advantage.
im confused ???
if you were up in material then why did you accept the draw?
and what do you mean by "he was the one who initiated the draw", do you mean he offerred the draw? in which case you could have declined if you were winning
also what does material have to do with rating changes?
Beating somebody 2100 once in a flash is not sufficient. Consistency over a period will stabilise the rating Thats how we should take it while off course looking for improvement in the database system
Jul 2, 2008
0
#19
vijaykulkarni wrote: Beating somebody 2100 once in a flash is not sufficient. Consistency over a period will stabilise the rating Thats how we should take it while off course looking for improvement in the database system
Another important thing I look for in an opponent to judge their real ability, is their average opponent rating....You can have someone that is 2300 with an average 1700 opponent rating, that is not as good as a 1950 with a 2000 average opponent rating.
Jul 2, 2008
0
#20
Magna_Canticle_XII wrote: vijaykulkarni wrote: Beating somebody 2100 once in a flash is not sufficient. Consistency over a period will stabilise the rating Thats how we should take it while off course looking for improvement in the database system
Another important thing I look for in an opponent to judge their real ability, is their average opponent rating....You can have someone that is 2300 with an average 1700 opponent rating, that is not as good as a 1950 with a 2000 average opponent rating.
sorry but in the famous term of NM Reb,
hogwash.
In your example if you're 2300 it means you've beaten pretty much all those 1700 guys, 1950 against 2000 means you haven't even drawn 50% of the 2000 guys. Take it from me winning almost every game with no draws is very difficult indeed. It would be easier to get to 1950 playing 2000 rated than 2300 against 1700 rated, you only need to hold on for draws, grinding out a win in each and every game is a real challenge, one draw drops your rating a lot.
As for your *real* ability versus rating that's a different matter, but if you play enough games the two will equalise over time.
I know I can't be the first one to bring this up.
When you win a chess game here, your own game record should reflect what that opponents rating was before they lost to you, not the new rating they got because of losing to you. If a player is 2100 when you beat them, then 2100 should be their rating reflected in your stats.
Alexndre