Standard Ratings Boost

Sort:
Avatar of chessredpanda

cool!

Avatar of Wind

I kinda like it.

Avatar of geennaam

Can anyone explain exactly why the ratings are so much lower then the FIDE ratings? Just interested in the mathemathical side of the story, thanks in advance

Avatar of theVchip
geennaam wrote:

Can anyone explain exactly why the ratings are so much lower then the FIDE ratings? Just interested in the mathemathical side of the story, thanks in advance

I find the ratings here are actually notably higher, not lower.  If you seek an explanation for differences, you could start with organized offline play is generally rated in section, by strength, with measures to minimize sandbagging or inflated ratings against weaker opponents. Organized offline play is also rated against human opponents - no computer-easy, medium, etc.  OTB is real ratings on real performance, with measures in place to try to keep things accurate.  Online chess, not so much.

Avatar of VLaurenT
geennaam wrote:

Can anyone explain exactly why the ratings are so much lower then the FIDE ratings? Just interested in the mathemathical side of the story, thanks in advance

I see two main reasons :

- starting rating average on chess.com is 1200 while FIDE ratings were historically above 2000,

- large number of cheaters in standard on chess.com, sucking away points from the pool before being banned.

Avatar of zborg

The possible large size of the cheat pool, and the experience of NM @FLChessplayer (who played 86 games while only achieving a 1550 rating), both dovetail with @Hicetnunc comments above.  Thank you.

Avatar of Paul_A_88

THANK YOU

Avatar of fershenko7

Are the ratings on chess.com intended to emulate ELO rantings?? If it is like that I think thay are right doing the adjustment because my Standart ranking is now 2055 which very similar to my ELO rating. 

For blitz my real strenght is above 1900 ELO but here is about 1700. There should be an adjustment there as well. IMO Wink.

Avatar of theVchip
fershenko7 wrote:

Are the ratings on chess.com intended to emulate ELO rantings??

No.  The chess.com rating method makes that impossible.  For every person here close to their real ELO, their are a ton who are way too high.  I'll give an example of a member here who is rated 1800 here on standard time.  He plays his rated games exclusively against computer-medium, which is rated 500-600 points too high, hands him 3 pieces by move 15 without fail, and he beats it every time.  For a glimpse of reality, in his last 5 games against people, he lost to players rated 1371, 1367, 1436, 1455, and 1371.

If this guy is given the rating boost mentioned here, he will go to 2200.  So you'll have a person rated '2200' on chess.com, despite the fact he has a losing record to under 1400 players over the course of hundreds of games against these players.  Can that happen in real ELO?

Avatar of the_archbishop

I don't like this. It just tells us these ratings are meaningless. What we used to fight for no longer makes sense. Just as if someone told you: today you get one dollar for any dollar you own. You lose any sense of value.

Moreover, this creates a huge gap with blitz ratings, which most members don't understand, even though it may suit titled players.

I don't want to discuss the details, but this change should have been much more progressive.

Avatar of fershenko7

Well it seems this just doesnt work (after reading your explanations). Anyway i'll keep on playing no matter the rankings Wink.

Avatar of NewYorkRod

The rating jump was especially important since I am working with a teacher. They suggested I not worry about it, but I couldn;t get any games at my previous strength. One day, I just accepted whatever I got. Suddenly I was beating players ranked much higher than I. It gave me a lot of confidence and my teacher said my games were improving. Even now finding the truth, I see that I had always been stronger. I had only been stuck in the 1200+ class because of an error. In USCF, I had been 1455. I was re-rated at...you guessed it...1455. Thanks for the update. I appreciate it.

Avatar of small_potato
fershenko7 wrote:

Are the ratings on chess.com intended to emulate ELO rantings?? If it is like that I think thay are right doing the adjustment because my Standart ranking is now 2055 which very similar to my ELO rating. 

For blitz my real strenght is above 1900 ELO but here is about 1700. There should be an adjustment there as well. IMO .

Trying to emulate real FIDE ratings is dumb because internet chess is a different skill to real OTB chess, and people may be relatively better at one or the other. Sure the skillsets are more similar than they are different, but it's still not the same thing. I make mistakes online that I'd never make OTB and vice versa. This just seems an exercise in giving out free magic beans to gullible people to make them feel better about their chess ability.

Your chess.com live standard rating signifies your ability at chess.com live standard chess relative to other people who play chess.com live standard chess.

This change maybe doesn't make a lot of difference (other than make it awkward to look at your historical rating to see how you're improving or not), but it always worries me when I see an organisation messing with perfectly fine systems to appeal to the lowest common denominator.

Avatar of OlafWillnecker

My old standard rating was almost the same as my actual FIDE rating. I don't understand this artificial change.

Avatar of vmsfinale
theVchip wrote:
fershenko7 wrote:

Are the ratings on chess.com intended to emulate ELO rantings??

No.  The chess.com rating method makes that impossible.  For every person here close to their real ELO, their are a ton who are way too high.  I'll give an example of a member here who is rated 1800 here on standard time.  He plays his rated games exclusively against computer-medium, which is rated 500-600 points too high, hands him 3 pieces by move 15 without fail, and he beats it every time.  For a glimpse of reality, in his last 5 games against people, he lost to players rated 1371, 1367, 1436, 1455, and 1371.

If this guy is given the rating boost mentioned here, he will go to 2200.  So you'll have a person rated '2200' on chess.com, despite the fact he has a losing record to under 1400 players over the course of hundreds of games against these players.  Can that happen in real ELO?

refer no: 240 pg 12

Avatar of erik
Optimissed wrote:

It may be to stimulate interest, pure and simple.

nope. this was done 100% in an effort to make all of our ratings more accurate. it obviously did not work for everyone. very sorry. 

Avatar of GreenCastleBlock
erik wrote:

Before doing this, we took a hard look at the numbers, and on average, saw a HUGE disparity between standard, blitz, and bullet. We had some theories on why, but the truth was they were there. Not for every single person, but it was significant and across the board. I mean, people rated 2500 in blitz were 1900 in standard. That's not right.

Because the 2500s quit playing Live Standard after they kept getting beat by 1800s using computers.  That'd be my guess.

Avatar of Benzodiazepine

My theory on why I, personally, saw no standard rating boost is that my blitz rating is/was lower than my standard rating.

So you guys conditionally boosted standard ratings for those with standard ratings lower than blitz ratings!?

Avatar of duniel

My rating dropped 200 points! What is going on?

Avatar of Benzodiazepine

Lol, looks like someone failed at doing teh SQL.