FIDE versus USCF pairings

Running Chess Tournaments
1

Something interesting came up recently in a tournament I was playing in. I was part of a 4 person group of players at 1 point after 1 round [small event]. The tournament was FIDE rated as well as USCF rated. You have in order a 2400, a 2250, a 2210, & me a 2013. Without any colors taken into consideration the natural pairings would be 2400 vs 2210 & 2250 vs 2013. However, as it turns out that would have two players due White set to play & two players due Black set to play. The USCF pairing program [apparently using the 80 point limit] made the pairings of 2400 vs 2250 & 2210 vs 2013. Now we are talking about a 4 player group where most TDs doing hand pairings would have 2400 vs 2013 & 2210 vs 2250 [i.e. 1-4 & 2-3]. The TD at the tournament checked online and found that FIDE would also have done the same. The TD asked the players [1 who had complained about the original pairings] if they would be willing to go with the FIDE pairing solution, which is what happened.

The question needs to be asked if a FIDE event needs to adhere to FIDE pairing rules if there is any discrepancy? Any advise or suggestions about this?

Larry S. Cohen

2

I was not aware US Chess had a pairing program. It really depends on the settings that the TD was using.

Alex Relyea

3

FIDE rating does not require FIDE pairings.

4

USCF-wise, the pairing program was simply following rule 29E5e, Comparing transpositions to interchanges: “If pairing a round in which 29E5b, The 200-point rule, is used” [ e.g. when pairing round 2 ] “an interchange involving a smaller rating switch than a transposition should be preferred to the latter unless the transposition satisfies the 80-point rule.” In this case, please note that the 200-point rule, not the 80-point rule, applies since color equalization, rather than mere color alternation, is at stake.

The pairing program preferred a 40-point interchange (2250 minus 2210) over a 150-point transposition (the smaller of 2400 minus 2250 and 2210 minus 2013), in accordance with the rule.

FIDE now insists (I believe) that the FIDE Laws of Chess be used in FIDE-rated events. But the Laws of Chess do not include the pairing rules. Whether FIDE pairings must be used may depend on the scope of the event. For example, FIDE may require FIDE pairings in norm events but not in lower-level events, or something like that. And, of course, this requirement may change from year to year.

Bill Smythe

5

Yes, Most important issue is color balance; the risk of multiple
+2’s in round 3 that can’t be addressed. Interest is the first round complaint; first round pairings are set in stone

6

I have no idea what you are trying to say.

Alex Relyea

7

I think Birdsman is trying to say that FIDE considers color equalization (and even color alternation) to be more important than rating differences. In fact, the FIDE pairing algorithms do not take rating differences into account at all.

But they do take rating order into account. If four players are rated 2400, 1900, 1500, 1100, or if they are rated 1704, 1703, 1702, 1701, it’s all the same – in both cases they are ranked 1st, 2nd, 3rd, 4th.

Bill Smythe

8

The quick rule of thumb for FIDE is that a transposition is always preferred to an exchange/interchange. If you can get a valid pairing that has top-half players paired against bottom-half players and the minimum number of color issues in that scoregroup then that is what FIDE will go with.

9

I think in this particular case the FIDE pairing was superior to the US Chess version. There could easily be some counterexamples.

As to the original question only norm events require strict adherence to FIDE pairings.

10

A 40-point interchange seems more natural to me than a 150-point transposition. Then again, I also feel that a Harkness odd man is more natural that dropping the bottom (dropping the bottom is more likely to reduce perfect scores while using the Harkness odd man/drop the middle is more likely to avoid a larger discontinuity in the pairings and reduces the value of the “Swiss gambit”).

Skipping interchanges means that with 1st and 3rd due white and 2nd and 4th due black, FIDE prefers pairing 1-4 & 3-2 even if the ratings are 2700, 2150, 2149, 1599 (i.e. 2700-1599, 2149-2150 instead of 2700-2150, 2149-1599).

11

I’m guess I am curious. How many would keep the original pairing (1-2, 3-4)? How many would switch it to 1-4, 2-3? Once again the ratings were 2400, 2250, 2210, 2013. At what point would the rating difference become too great to switch? Or would you switch it regardless? For instance, take Jeff’s last example. The rating differences are more extreme.

12

As with many of the differences between FIDE and USCF treatments, the Dutch system was developed for “professional” chess. Color trumps ratings, as color is much more important at the highest levels, and the rating range in a high-end Swiss would likely be quite a bit smaller than you would see in a typical US Swiss. Also, the fact that they attempt to make the pairings completely deterministic, so there’s no TD judgment involved. (The full set of pairing rules is probably only slightly less complicated than rocket science—fortunately, most of the really messy stuff probably only comes up if you try to run 9 rounds with 18 players or something like that).

13

Let’s say the raw pairings (I hate to say “natural”) were* 1901 WB vs 1500 WB

  • 1501 BW vs 1100 BW
    In this case it would seem much more natural to switch 1501 with 1500, giving* 1901 WB vs 1501 BW
  • 1500 WB vs 1100 BW
    than to switch 1500 with 1100, giving* 1901 WB vs 1100 BW
  • 1501 BW vs 1500 WB

But, again, it would violate top-half vs bottom-half. Lots of players seem to think that, just because they’re in the top half, they have a God-given right to play somebody in the bottom half. What these players don’t realize is that an 80-point switch (for color alternation) is OK, which means that if somebody in the top half is within 80 points of somebody in the bottom half, each should be considered to be in that gray wishy-washy area between the two halves, each available to be placed in either half, if necessary or desirable.

The same logic would apply when color equalization, rather than mere color alternation, is at stake, except the permissible difference now becomes 200 points instead of 80.

So what can we do to overcome this player ignorance? The only answer seems to be the education of thousands of players who have only “top half vs bottom half” in their repertoire of rules. And we all know that education of such a large number to such a degree is an impossible task, don’t we? So instead we go with convenience to avoid arguments.

The following US Chess rule complicates matters further:

29E5e. Comparing transpositions to interchanges: If pairing a round in which 29E5b, The 200-point rule, is used … an interchange involving a smaller rating switch than a transposition should be preferred to the latter unless the transposition satisfies the 80-point rule.

This wording muddies the waters a bit, partly because it mentions the 200-point rule at the beginning and the 80-point rule at the end. Apparently it is designed for use when the 200-point rule applies, but in some cases replaces it (somewhat arbitrarily) with an 80-point rule. A slightly better choice of words (at the end) might have been:

29E5e. Comparing transpositions to interchanges: If pairing a round in which 29E5b, The 200-point rule, is used … an interchange involving a smaller rating switch than a transposition should be preferred to the latter unless the transposition is within 80 points.

The above rule is somewhat awkward in any case. It’s one of those “wink wink” rules, where the writer talks about “concerns” of the players, as if he secretly wants to say, “Don’t tell anybody I said this, but players sometimes don’t like this rule so you might want to consider not following it.” A bunch of those popped up in the 4th edition, and have been carried over to the 5th and 6th.

Bill Smythe

14

I agree. Actually made such a switch, 3rd round of a 16 player swiss. I transposed pairings in top group. Natural pairing was 2100 v 1100 on bd 1.

I am most interested in the complaints about first round pairings. Aren’t they pro forma, set in stone??

15

Why would they be?

Alex Relyea

16

Well, they are set by rule. In this case, swiss top half vs bottom half

17

Swiss top half vs bottom half still leaves board one colors with two possibilities.

An 8-player round robin had 40,320 possible valid round one pairings (prior to the determination of the pairing numbers).

USChess round one pairings can regularly have complaints about intra-team pairings (usually based on either teams not being used, or on teammates having been entered with different team codes).
They can regularly have complaints about which rating was used (supplement for the current month, supplement for the previous month if the tournament is early enough in the month to opt to use that instead, highest of the last six supplements, latest rating after the most recent tournament which was rated after the latest supplement was created, higher of regular or quick for a quick tournament, etc.)
They can regularly have complaints about which section a player was assigned to.
They can have complaints regarding whether or not accelerated pairings were used.

18

well, although I can understand that, the rules for such things are straightforward. I’d be willing to adjust ratings if
the FIDE rating is significantly higher than the USCF rating

I’d listen but mostly rules for these issues do not give me much flexibility

19

Not often. For example, occasionally tournaments switch me with the next player higher or lower when I’d be due to play my wife. There are also tournaments with players with the same rating, and frequently multiple unrateds. Sometimes it happens the way you suggest. Not often, and certainly not by rule.

Alex Relyea

20

Agreed…the same application with siblings; I was making the point they are few.