Do you think that rating really shows skill of player?

atotalscrub wrote:  'I'd also like to add that the same player may play at a different strength on way day vs another. They could be tired, hungry, or just not focused. That could make them appear to be playing worse than their rating would suggest. It could also mimic sandbagging to an extent if they stubbornly continue to play too many games while tired or hungry and lose a lot of rating points. We're not machines.'

Indeed. I think I need to look at overplay, and playing when Hungry.Angry.L.onely.Tired. And just like in poker going on tilt and making stupid moves on the back of a bad run can be an Achilles heel. 

After a player has played a bunch of games, ratings are extremely accurate in measuring what they are intended to measure--your chess performance. Your rating doesn't measure your chess "understanding" or your chess "knowledge," it measures your results. You won, you lost, or you drew. 

I find the idea that there is a chess playing "skill" that isn't reflected in one's rating unconvincing. 

n

Definite correlation between rating and still no question about it play for a while and you will see there's absolute correlation no doubt about it it's the only way to go really even though it would be nice if it weren't related but it is absolutely related the rating and the ability no question about it forget about it that's the way it is

Rating is not a measure of skill, its a measure of recent performance.

bhnrfemhrjfmb

I've been hovering between 1650 and 1700 on this site.  I don't have an ELO.  This site says I'm in the  top 5% to 6% of their players.  The ELO distribution would put me around 1900 for the same percentile.  Is it the case that the ELO pool of players is stronger than those found here on Chess.com?  How could we tell if that's true or not?  I would expect that below 1000, there would be a larger number of Chess.com accounts than under 1000 ELO - so certainly not a normal distribution, but my intuition is that in the upper range the distribution would be more or less the same.  Is there a distribution graph available for Chess.com?    I'd like to compare it to ELO's.    

BTW, duh!  Of course rating is an indicator of ability -- by definition.    

No. My rating is 1917 but I have only ever defeated an 1839 player once. The average rating of players is 1745 when winning and the average is 1762 when losing. So, I would guess my "real" rating is in the area of 1750, not 1917. It's just that every win adds a few points and I have crept up to 1900+.

Some things can skew your rating and Chess.com not willing to fix. For example my son played a few games on my account here [I did not know he was doing this]--they were all fast play games and while he won all of his games it gave me a rating in the 1600s.

I am probably better than "1600" as my records against masters has been about 68% for the past 35 years and my record vs GMs has been 100%  [4 wins out of 4 games with 4 different GMs]

1. Ratings and Skill level (or knowledge about chess) remind me alot of the difference in how people do in college. 

there are plenty of people that aced all of their tests and placed at the top of their class. Some of those also went out and did very well in business. But others kind of waffled. Why is that? One student knows the formulas and book answers, and the other student knows how to connect better with people, maybe realizes more intuitively how to work on teams, has learned how to become more valuable to clients of firm - even though they scored lower on the "book" exams in college. 

There is a difference between book smarts and street smarts. It is the same in chess (i am not talking about book opening knowledge).

One player can know all about many diverse openings and variation, many middle game ideas (outposts, open files, semi-open files, weak pawns, backward pawns, weak squares, color weaknesses, space, pawn structures, endgame strategy, triangulation, passed pawns, outside passed pawns, outflanking, etc. etc. another player might only know two-thirds of all that.

but let's say the first player sometimes gets very nervous, does not manage their clock time well, sometimes gets so caught up in a strategic concept and figuring out how to gain a large advantage, but they miss a relatively basic tactic and then toss the game. 

Let's say the second player knows much of what the first player knows, but does not get quite as nervous as often, has a tough attitude about playing, does not get "lost in space" endlessly analyzing a position, and is reasonably adept at tactics and endgames. 

The first player arguably knows more about chess. you could say they have more skill in a sense. But they are likely to not perform as well as the second player who is more practical and does a better job of getting maximum results out of what they do know, even thought it is certainly less than the first player. 

I am not saying this is a stereo type that plays out in every scenario, but I have certainly seen and experienced it many times from both sides. 

therefore, while ratings and skill are correlated, one is not the final indicator of the other. 

Sand bagging is a major thing in chess. For those unaware sandbagging is when a player plays below their actual strength. They do this to enter lower-level tourneys and win prize money. I knew a guy in LA that played at an 1800-1900 level normally. He sandbagged the few times he played in Swiss tourneys at Chess Palace so he could play in the 1200-1400 section. The top prize for the section was around $2500 or so as I recall (it was a pretty big tournament, grandmasters could play for free). 

beacause chess.com allow player choose start rating, which lead to this mess

 I'm a low rating [apparently 1500 isn't good ,but i dunno anything about chess ratings and crap],and with the power of dumb luck i can beat a higher rated-person.

we should just show how many times they've won, resigns from the other person/draws not included.