Chess rating system

What elle is trying to point out is that most people think they are underrated. It's not about cheating, it's not about statistics, it's about the belief that we're better now than we were. Ratings only show how we have done in the past. It's possible to make predictions on how we will play in the future, but most of us believe our best is still in front of us. 

Reasonable point, that it's a measure of how well you have performed in the past. But if most people are better than their rating, then the person you play with a similar rating is also better than their rating and it evens out.

I had a couple question about ratings in the case that the game has not actually "started," in the sense that either there are 0 moves made, or only white has made the first move, and where one side has timed out.

This could happen for example, when somebody neglects to set their vacation settings, or their vacation has timed out, or in the case that somebody wishes to exit from a tournament, simply resigns without making a move.

My understanding is that under standard USCF rating system, no rating adjustment is given for this circumstance: it is a forfeit. Clearly, a win or a loss under these circumstances is not a reflection of skill in chess, so the reasoning goes that there should be no ratings adjustment.

However, if I understand correctly, the chess.com rating system will actually adjust the ratings of both players as if it were a completed game.

So first question:

Is my understanding correct?

Second question:

Assuming my understanding is correct, is this something that chess.com is going to address?

@pretrip:

"Chess.com once a move has been made it is an game. But why should they adress it? Number of games that go that way is so small that it will affect accuracy of ratings at all. Certainly we do not want engourage dropping connection when u got black pieces"

A few comments here:

First, our you sure about the statement that once a move has been made it is a game? Honestly, my memories of these experiences tend to show differently - at least with respect to speed chess games (I may be wrong here). If you know of a stated policy supporting your assertion, could you point me to it?

Second, the number of games is not really relevant when it creates large drops or gains for a person that are inappropriate. The policy for these situations should not be decided by frequency, but by providing an accurate reflection of chess ratings. I do know of one case in particular (I will leave out names - suffice to say, it was not me) where a 2300 lost 200 rating points for withdrawing from a tournament. For online chess, recovering this many points can take quite a while.

Third: For speed chess games, the color of pieces seems to be assigned stochastically favoring the color for you that you have played less frequently. So it would be a frustrating experience for someone trying to game the system. Frequent connection drops based on color can also be easily detected. Moreover, measurements can be made to see if the connection drops are made "strategically" based on color (e.g., only dropping connection with black pieces against tough opponents, or where opponent strength is an even match). 

Finally: For online games (i.e., not live games, but games with a time control measured as 1 move/x days), there is already motivation to get the best results, either as part of a team event, or part of a tournament. However, even in this case, frequent early forfeits made strategically can be detected.

You can also say that bullet/blitz chess relies on opening knoweledge which is memory and not "thinking". Speed is talent in and of itself, but is not essential for chess mastery. Knowing what to play is crucial for success.

It is foolish to assume a faster player = a better player.

I second this. I really suck at blitz - I still suck at longer time controls but much less so :oP

When I look at my opponents' stats with whom I'm competetive in standard or Online time controls I often find that their blitz ratings are hundreds of points higher than mine. Quite a few of the people I can beat in longer chess could massacre me in blitz. And 'bullet' chess is completely beyond me. Of course, I have a hard enough time keeping up with bingo, so maybe it's just me :oD

can you read? I said exactly that: most of the people's tactic rating is lower than their blitz rating! With me it's the opposite. 

They actually do matter, they make chess more competitive and they give a sense of how strong a player is

Interesting that the discussion led here with some saying ratings don't matter.

They may not matter to some but I find them incredibly accurate predictors of skill.  When i play a 1300 vs a 950 I feel the difference.  A 950 will blunder 3-4x in a game with an en pris piece almost every time.  A 1300 will have to be tricked... beaten... and sometimes kibitzed into submission my my mere 1050ish rated self.

I am actually posting here to ask ... my goodness, they are VERY accurate.  i recently beat a 1300 person.  I went to look at her games and I noticed she has a lot of games against lower rated opponents.  Perhaps that strategy has carried her ot 1300.  But I also noticed she has been at 1300ish for a long time too. 

I took too many Psychology courses in college.  Way too many.  So I'm REALLY in awe at how accurate the ratings are. And we see it in FIDE's antiquated elo rating system too.  Wesley So vs Magnus Carlsen, Pogonina, or Muzychuk would probably be a loss for Wesley (Wesley, I love your games -- this is just a guess based on ratings.)  And my own games are INCREDIBLY predictable against both equal, lower, and higher rated opponents.  Have we somehow mastered this measurement?  I think I fluctate +/- 50 points but wow is the math solid in the rating system.

ευχαριστη!

Duh!

Attackbread & mznor:

My wife has an account here too and she's rated at like 1400.  That is because when you start off your rating is provisional.  There's math behind the ratings that works but your rating really doesn't mean anything until after about 20-50 games depending on who you ask and who you play. 

As far as chess.com problems, I have not seen any and my account has worked well. Your rating will vary wildly for the 1st dozen or so games because the math in the system uses a weighted average of sorts to calculate your rating. So when you first start, you can go up or down by 100 points.  Now, if i play someone 100 points higher I only go up by about 14 points.

Check this out: http://www.chess.com/article/view/chess-ratings---how-they-work

You ask that question about once a week. You're a member of USCF, why not just go to USCF and look it up yourself?

Im a 1525 at standard chess

I play blitz casually, just to play something instead of minesweeper, and haven't been in a tournament since I was a kid, was just wondering how a 1600 to 1700 rated player would do if I was to join a club again. 

My blitz lvl is pretty much on par with my regular chess lvl.

Was just toying with the idea of picking chess up again.

You'll be an average club player according to some of the chess.com statistitcs I've seen.  Keep in mind, average club players are VERY good by any standard.