Questions tagged [rating]

A number that approximates a player's skill. The actual meaning of the number depends upon the issuing organization. FIDE ratings are used as the international standard for over-the-board chess.

A rating is a number that approximates a player's skill. In general, by comparing two players' ratings, one can gain an appreciation for the likelihood of one player defeating the other. Having ratings allows tournament organizers to group those with close ratings together, affording everyone challenging but winnable games.

Example for USCF rating system

United States Chess Federation (USCF) ratings are common for American players. By participating in USCF-sanctioned events, players earn ratings that are based upon their performance against other rated players.

Remember, ratings approximate skill. They are ever-changing and adapt in accordance to the results of games played. For the USCF rating system, players are playing for some share of the other players' rating points. That is, points won by the winner are taken from the losing player; it's a zero-sum system.

Players with equal ratings have equal chances of winning. A rating difference of 400 points means that the lower-rated player has virtually no chance of winning the game. A difference of perhaps 200 points means the lower-rated player will be challenged but should have a sporting chance.

How are rating points earned? The more the result of a game mirrors the expectation based upon the players' ratings, the fewer points are won or lost. Examples:

  1. If two players are equally rated, the winner will gain 16 points at the expense of the loser.
  2. According to the way the USCF system was designed, an 1800 should defeat a 1400 pretty much every game. If this happens, their ratings are deemed accurate and change very little, perhaps a point or so; the result affirms the relative ratings are accurate.
  3. However, if the 1400-rated player somehow wins, she gains a whopping 32 points at the expense of the 1800 rated player (who now becomes a 1768-player.) Their ratings are out of kilter and need a strong correction.
  4. Finally, an 1800 is expected to defeat a 1600 most of the time, but certainly not all the time. Thus if an 1800 defeats a 1600, the 1800 wins perhaps 10 points.
  5. Conversely, if the 1600 player were to win, the 1600 would gain about 22 points. As before, the players' ratings are adjusted to better reflect reality.

The actual system in use has plenty of tweaks in it - the above simply shows how it works in general.

The average club player is rated at 1500; Club players have ratings from E (low) to A (1800-1999), then labels such as 'Expert' (2000-2199), National Master (2200-2399), and Senior Master (2400+). The USCF does not grant International Master or Grandmaster titles.

Finally, bear in mind there are many rating systems. FIDE and the USCF both use similar systems invented by Arpad Elo). Other systems exist including the Glicko system.

352 questions
25
votes
2 answers

What was Arpad Elo's own Elo rating?

I've found several sources that indicate that Arpad Elo was a "master-level chess player", but it's been very difficult to find any information about what Elo's own rating would have been under his own system - primarily because if you do the…
patbarron
  • 3,609
  • 1
  • 15
  • 37
20
votes
2 answers

Based on Elo rating, what percentile am I in?

How do Elo ratings compare with percentiles, and is there a chart breaking percentiles down by Elo rating ranges?
blunders
  • 4,137
  • 3
  • 26
  • 39
15
votes
6 answers

Going from Expert to Master

What is the difference in mindset or technique that really makes the difference between an expert and a master? I am trying to decide whether or not to start playing again and try to make that leap.
Andrew Latham
  • 3,651
  • 2
  • 23
  • 38
13
votes
3 answers

Chess Hustlers in the Park/Street: Where do they place in terms of skill?

Assuming the species still exists, has anyone researched this enough to provide insight into their approximate rating class/skill level (exclusively at blitz time controls)? How do they typically get into the business or for that matter, demonstrate…
shivsky
  • 3,313
  • 19
  • 28
13
votes
8 answers

How to gauge someone's chess strength quickly?

I plan on teaching a chess class (5 classes in a sequence, each building off the previous, 1 hour each) to high schoolers, with emphasis on applying chess principles to real life. Now, it'll be unproductive to have players of maybe <1200-1300…
chubbycantorset
  • 4,616
  • 7
  • 32
  • 53
12
votes
1 answer

The number of levels in chess

Consider a weak chess player who has just learned how the pieces move and who will lose to a Scholar's Mate. (We can say that he's male and call him "Mr. Newcomer".) Let's define Mr. Newcomer's level to be 0. Moreover, let's say that player A is a…
user181813
  • 223
  • 1
  • 5
10
votes
6 answers

What are some examples of wins by a much weaker player?

I'd like to study examples where a much weaker player succeeded against an IM or GM.
Landei
  • 2,507
  • 17
  • 23
10
votes
4 answers

Is 80 points FIDE equivalent to twice as good a player?

I read somewhere that if you're around 80 points better than another player, statistically you should get around twice as many points against the same opposition. This impressed me because it's not how we generally think about ratings. I worked out…
magd
  • 1,755
  • 13
  • 18
10
votes
4 answers

The strength of blind chess players

Before I begin, I'm not asking about people playing "Blindfold chess", but those who are actually blind. Reading through the FIDE rules of chess like any good player should, I've previously noticed the section containing rules for players who are…
DTR
  • 3,129
  • 1
  • 18
  • 41
8
votes
7 answers

How does the chess24 rating compare to other server ratings?

My question is how the chess24 blitz-rating compares to the rating of other servers, like ICC or the ChessBase server. And whether the chess24-rating inflated over the last few month. The reason for my question is that I used to play a little on…
BlindKungFuMaster
  • 19,216
  • 1
  • 50
  • 79
6
votes
4 answers

Besides rating consistency and rating going up, are there other objective ways to track progress in chess performance?

For example, could one measure ACPL (Average CentiPawn Loss) over time to detect progress? Could masters look at the games of a player and determine they're playing better even if their rating isn't going up, assuming their losses aren't attributed…
moh abdi
  • 91
  • 3
6
votes
1 answer

How do rating systems cope with clusters of players who nearly always only play people within their cluster?

Given the issues with rating people who mostly play within a cluster that contains few people who play outside the cluster, I was expecting rating systems to take more notice of results against another player who was well connected to many different…
Ian Ringrose
  • 695
  • 4
  • 15
6
votes
4 answers

What do you call average?

Is there an average rating? If so, what is it? I saw this in a chess.com forum: There's no such thing as an "average" chess rating, but a solid club-level chess player might be rate somewhere around 1500-1700. In general, a person's rating will…
DialFrost
  • 1,123
  • 1
  • 6
  • 26
6
votes
1 answer

Precise skills connected with rating value

I recall a quote that went something like that: "A 1800 beats a 1500 by tactics, a 2100 a 1800 by foo, a 2400 a 2100 by bar and a 2700 a 2400 by opening(?!)". "Endgame" was in it too, I think. Anyway. Maybe one of you remembers the source and exact…
Hauke Reddmann
  • 16,242
  • 3
  • 27
  • 70
6
votes
2 answers

How do you calculate error bars on elo rating?

In season 14 of TCEC, Stockfish defeated Leela 50.5-49.5 in a 100-game match (the closest possible winning margin). Using an elo calculator such as this one, we can see that the elo difference between the two engines was +3. However it must be that…
Allure
  • 26,534
  • 1
  • 68
  • 146
1
2 3