Wednesday, March 09, 2016

On FIDE Ratings and Tournaments

Since the Boylston is going to host a FIDE rated tournament in two weeks, I thought it would be worthwhile to give a quick overview of some of the differences between the USCF and FIDE when it comes to ratings and tournament rules.  For the most part, you can show up, find your pairing, and play your game without thinking about any of the subtleties, but if there's a dispute or if you're just curious about the minutia, keep reading.  There's an awful lot of information here, and I tried to explain everything from the perspective of a player who has played in both USCF and FIDE rated tournaments.

FIDE Laws of Chess

The vast majority of rules for a FIDE tournament are the same as USCF tournaments.  The pieces all move exactly the same way, touch move still applies, and players still have to write down all of their moves.  The time control for the vast majority of FIDE events has changed to have a 30 second increment.  The most common time controls are currently G/90 + 30 or 40/90 +30 SD/30 + 30.  To translate - either the whole game in 90 minutes with a 30 second increment from move 1, or 40 moves in 90 minutes followed by sudden death in 30 minutes with a 30 second increment from move 1.  

Playing with a 30 second increment is quite different, especially if you're used to 40/90 SD/20 from the Thursday Night Swiss.  The extra 30 seconds every move means that it's very hard to actually put your opponent in time pressure in the first time control.  Mentally, you should add a minute to your opponent's time for every two moves before time control.  So if someone's clock says 5 minutes at move 30, treat it like 15 minutes to make 10 moves instead of 5 minutes to make 10 moves.  As someone who benefits from putting opponents in time pressure - that's a huge practical difference!

On the other hand, in a sudden death time control, the time pressure can be a lot worse.  It's one thing to mentally make the change to playing on the delay (either 5 or 10 seconds depending on the tournament).  That's just enough time to check for blunders, make the move, and hit the clock.  30 seconds is a little bit sneaky - it feels like it's long enough to actually consider the merits of the position and play a good move, but by the time you write down your opponent's move (yes, you have to notate to the bitter end with the 30 second increment), you barely have enough time to decide what your opponent was trying to do.  With practice, the time control is fine, but G/90 + 30 is definitely a culture shock the first time you try it.

There are a few other differences in the rules that are important enough (and likely enough to occur) to mention here.  The fundamental difference between FIDE and USCF is that in FIDE rated tournaments, the assumption is that there will be an arbiter watching every single game for just about every move.  Under USCF rules, TD's are often treated as additional witnesses in the case of an illegal move, touch move, flag falls, etc.  For example, in USCF rules, if both flags are down, the game is a draw.  In FIDE rules, if both flags are down, the arbiter must decide which flag fell first.  Going back to that assumption that there is an arbiter watching, it will always be the case that the arbiter saw which side actually ran out of time first.  To this end, clocks that stop when a player runs out of time aren't discouraged the same way that USCF rules discourage such clocks.  

When notating in USCF tournaments, it's fine to either write the move down on your scoresheet first or to play the move and then notate (assuming that you're using a paper scoresheet instead of an electronic one).  In all FIDE tournaments, the move must first be played on the board and only then written down.  The single exception to this rule is that if you are trying to claim a draw by triple repetition or via the 50 move rule, then the intended move must first be written down, then the clock should be stopped before you execute the move on the board.  

Illegal moves are also different between the two organizations.  In normal USCF play, if the TD sees an illegal move made when there's no time pressure, the TD may correct the move.  In practice, the TD's at the Boylston will correct illegal moves when they see them.  Since we don't watch every game, it's entirely possible that nobody will notice.  If more than 10 moves have passed since an illegal move was made, the current position stands and play continues.  Under the auspices of FIDE, arbiters should correct all illegal moves and touch move violations, even if neither player makes a claim.  Additionally, there's no limit to how far back the game may go in order to correct an illegal move.  (Did you play Bc1-f4 over the d2 pawn on move 3?  Then your game is going all the way back to move 3!)  The penalty for the first illegal move is the same - two minutes are added to the opponent's time.  For the second illegal move, FIDE rules call for the offending player to be forfeitted while USCF rules simply add two more minutes to the opponent's time.  

One last rule which is interesting, but not applicable to this tournament deals with calling your opponent's flag in a non-sudden death time control.  In USCF tournaments, in order to have a successful time forfeit claim, the claimant must have a "reasonably complete scoresheet".  This means that there can be no more than three missing or incomplete move pairs.  If your flag has fallen and your opponent has an incomplete scoresheet, you can actually call your own flag, and your opponent is not allowed to fill in any more moves after that.  Someone who stops the clock and claims a time forfeit and turns out to be wrong is penalized by having two minutes added to their opponent's clock (although the TD at a recent CCA tournament in Stamford didn't know that penalty and I was out of luck when it came to my extra two minutes...).  

In FIDE tournaments, the entire above paragraph goes out the window.  Since the arbiter is watching the game, the arbiter will determine whether or not 40 moves have been made, and the scoresheets, while useful for this purpose, will not disqualify either player from claiming a time forfeit.  Once again, since the time control for the New England Masters is G/90 + 30, this won't apply and is merely a USCF rule that many players aren't familiar with.

Ken Ballou has put together a comprehensive comparison of the differences between FIDE and USCF rules.  There are a few more rule differences that I didn't mention above simply because they are less likely to come up.  If you have the time, the following pdf is a great thing to read, and you might even learn a new USCF rule or two.

FIDE Ratings

FIDE ratings are similar to USCF ratings in that you start with a provisional rating, and after a certain number of games you receive an established rating.  However, that's pretty much the end of the similarities.  In order to receive an initial FIDE rating, a player must complete five games against a FIDE rated opponent.  Additionally, the player must score at least half a point in those five games.  The five games don't have to be in a single tournament, and can span more than two years (26 months to be precise).

After a player has completed at least five games, his or her new rating is calculated by using all games completed so far.  To find an initial rating the following steps are performed:

  1. Find the average rating of the opponents (Rc = Rating of competition)
  2. If the player scored exactly 50%, this is the player's new rating
  3. If the player scored more than 50%, the new rating is equal to the avg. opponent rating + 20 points for each half point above 50%
  4. If the player scored less than 50%...
    1. Determine the player's percentage (eg. 1.5 / 8 = 0.1875)
    2. Lookup the percentage in FIDE's table 8.1
    3. Subtract the value from the table from the avg. opponent rating to get the initial rating
FIDE actually rounds a fractional score to the closest hundredth (so 0.1875 is 0.19 on the table) and then rounds the calculated rating to the nearest integer.  If you don't want to find that table every time,
this formula gives an approximation: P = 1 / (1 + 10 ^ (-D / 400)).
P is the expected score (as a percent) and D is the rating difference between the two players.

If you're still with me, you'll notice that effectively FIDE "punishes" players who have a good tournament by capping their initial rating to the average rating of their opponents plus 40 points for each win.  In USCF ratings, if a player beats an opponent, the initial rating has 400 points added to the opponent's rating.

To make up for these lower initial ratings, FIDE has a very high K-factor.  For anybody who's not up to date with the math behind chess ratings, a K-factor is essentially the multiplier that is applied to the normalized rating change after a game or tournament.  A higher K-factor implies that the rating is more volatile and will make the rating change a larger amount.  All FIDE players start with a K-factor of 40 until they have completed 30 games.  After the first 30 games, the new K-factor is 20 (i.e. a player's rating will change 50% as much as it would for the first 30 games).  If a player ever goes above 2400, then the K-factor is further reduced to 10.  Lastly, in order to recognize that junior players often improve very quickly, the K-factor for FIDE ratings is left at 40 until a player turns 18 or goes over 2400.

If all of this sounds interesting, and you want to get a FIDE rating, consider playing in either section of the New England Masters.  Both sections are FIDE rated, and registration is open to anyone, regardless of rating.


Boylston Chess Club said...

Interesting. I assume at elite tournmts arbiters can review the video resolve issues.

Robert Oresick said...

Just to clarify the situation I assume that play in New England Masters will also be rated by USCF. So playing games in the New England Masters will have an effect on one's FIDE rating and also one's USCF rating.

Andrew said...

Yes, any FIDE tournament held in the US must also be USCF rated. FIDE rated tournaments have to be submitted via the USCF, and part of that process is basically the requirement that the tournament is first rated by the USCF.