Lone Pine 1975 Swiss pairing controversy

Running Chess Tournaments
1

Reading through old issues of Chess Life and Review, in the September 1975 issue I saw a complaint by Larry Evans about the last round pairings at Lone Pine 1975, and chief TD Isaac Kashdan’s response. Here were the standings and the colors needed for equalization in the top score groups:

Liberzon 7 W

Evans 6.5 B

Gligoric 6 B
Gheorghiu 6 W
Quinteros 6 B
Weinstein 6 B

Liberzon had played Evans and Quinteros
Evans had played Liberzon and Quinteros
Quinteros had played Liberzon, Evans and Weinstein
Weinstein had played Quinteros

Ratings aren’t listed, but apparently Weinstein was rated substantially below the other players, and was the only player without a FIDE title.

Kashdan’s initial pairing, as he told Evans the night before the last round, was:

Liberzon - Weinstein
Gheorghiu - Evans
Quinteros - Gligoric

Evans objected to this loudly and at length, disturbing games that were in progress, saying that the following pairing was clearly preferable:

Liberzon - Gligoric
Evans - Weinstein
Gheorghiu - Quinteros

After thinking about it overnight, Kashdan decided on this:

Liberzon - Gligoric
Gheorghiu - Evans
Quinteros and Weinstein paired in the 5.5 point group.

It’s interesting reading Kashdan’s reasoning for his pairings. This was two years before the publication of the 2nd edition rulebook, and pairing procedures weren’t spelled out in as much detail as they are today. Kashdan’s treatment of odd men was to pair them normally in the next score group, so in this case Kashdan considered the natural pairings to be:

Liberzon - Gheorghiu
Evans - Quinteros
Gligoric - Weinstein

Evans and Quinteros had already played. Transposing Quinteros and Weinstein would fix that, but colors would be wrong on all three boards. So instead he paired Weinstein against Liberzon, Gheorghiu against Evans, Quinteros against Gligoric.

Evans said that since Liberzon was in clear first place he should be paired with the highest player that he hadn’t already played. Kashdan decided to follow this principle both for Liberzon and Evans:

Liberzon - Gligoric
Gheorghiu - Evans
Weinstein - Quinteros

However, Weinstein and Quinteros had already played.

That problem can be fixed by pairing Evans against Weinstein and pairing Gheorghiu against Quinteros, which is the pairing that Evans demanded. Kashdan considered this to be illogical. “How does Pairing B [the one with Evans playing Weinstein] come about? The only way is to find the strongest opponent for Liberzon, as above, but then, instead of following the same procedure for Evans, merge the latter with the other cards, when he would play Weinstein. It was “impossible” for Liberzon to play Weinstein but highly desirable for Evans to do so. This hybrid idea seemed quite illogical to me, and I did not consider it seriously.”

As mentioned earlier, Kashdan decided to pair Evans against Gheorghiu and drop Weinstein and Quinteros to the 5.5 score group.

I think the 6th edition rulebook supports the pairing that Evans favored. Rule 29D1 “a. In the case of an odd number of players, the lowest-rated player, but not an unrated player, is ordinarily treated as the odd player and is paired with the highest-rated player he or she can play in the next lower group. Care must be taken in doing this that the odd player can be paired in the next score group, that the remaining members of both affected score groups can be paired with each other [my emphasis], that the odd player has not played all the members of the next lower group, and that the color consequences are acceptable. … b. If the conditions in (a) cannot be met, then try treating the next lowest-rated player as the odd player, or pairing the odd player with a lower-ranking player in the next score group. … There is no rating limit on the permitted switch if it is needed to keep the score groups intact. …”

Following this rule, and starting with the highest score group, first Liberzon is paired with Gligoric, then Evans would normally be paired with Gheorghiu, but since Quinteros and Weinstein have already played, Evans is paired with Weinstein and Gheorghiu is paired with Quinteros.

Kashdan’s final pairing does have the advantage of pairing Evans against the strongest remaining opponent, but at the cost of breaking up the 6 point score group and allowing multiple players to score 7 points. Are there any NTDs today who would make the pairings the way Kashdan did?

Edit: It was the September 1975 issue of CL&R, not August 1975.

2

I would not pair any of the top 6 ranked players down to the 5.5 group. There are 6 players with 6+ points and there are legal pairings among them, many which were mentioned. The score group has priority over ratings. Pairing 2 sixes with 2 5.5s would guarantee 4 players to be unnecessarily be paired out of their score group. It would have been helpful to see the ratings and color history before deciding which pairings were best.

3

I would want answers to a few questions before attempting to give a definitive answer.

(1) Were USCF Swiss pairing rules even in effect for Lone Pine 1975? (If memory serves, norms were available that year, after not having been previously available.)
(2) Were there any material differences in USCF’s (or FIDE’s) Swiss pairing rules circa 1975 versus today?
(3) Is there any additional information about color history available? (Maybe someone happens to have a crosstable from an old state publication handy…)

I believe the 1975 tournament was 10 rounds, which at least helps understand the equalization options Kashdan had.

Having said all that, I would be very loath to break up the leading score groups unless absolutely necessary. Among the top six players Mr. Messenger listed, I’d probably make any legal pairing I could to avoid dropping anyone out of that group.

4

A crosstable was published in the July 1975 issue of Chess Life and Review, but unfortunately without colors. It does show the ratings of the players:

Liberzon 2485
Evans 2516
Gligoric 2593
Gheorghiu 2521
Quinteros 2505
Weinstein 2372

In the last round Quinteros was paired with Shamkovich (2505) and Weinstein was paired with Benko (2492). All six players drew their games. Browne (2554) and Panno (2545) joined the players tied for third by winning their last round games. Weinstein earned an IM norm, and would have earned a GM norm if he’d already been an IM.

Yes, it was a 10 round tournament.

5

There appear to be only four pairings which are “legal” under current (and I assume even historical) practice:

  1. if Liberzon is paired with Gligo, then Evans-Weinstein is forced to avoid Weinstein-Quinteros (Evans’ recommendation)
  2. if Liberzon is paired with Gheorghiu, then ditto
  3. if Liberzon is paired with Weinstein, then Evans can be paired with Gligo or you could have
  4. Liberzon-Weinstein, Evans-Gheorghiu (Kashdan’s original pairing)

#2 is out immediately, since it gets all colors wrong and is inferior in pairing sequence to #1, which maximizes correct colors. #4 is inferior to #3, since both get two colors correct, but there’s no reason to upfloat Gheorghiu rather than Gligo since it doesn’t improve colors (overall). That may be a more modern interpretation, since Evans-Gheorghiu gets the colors correct for that pairing but pushes the bad color down a score group.

I honestly don’t see any reasonable sequence which produces #4 rather than #1.

6

Even in those days, it would have been poor practice to pair the 6’s into a lower score group unless no other pairing was possible because of players having already played others in the score group. There were also no rating limits for switches. While color was a factor, it was not at the level it is today. It was a common practice to adjust pairings to aid a player to get a FIDE norm; FIDE was not as fussy about that then.

I would want to see color and other pairing history for these players. For example, had a player been dropped into another score group for pairings in any other round? He would not be dropped down again.

Of course, Evans wanted to play Weinstein. Gheorghiu was known to be particularly good with White, and Gligoric, whether with White or Black, would have been hard on Evans’ King’s Indian Defense. I have often been offered “suggestions” by players for pairings that would favor them. Better to go by the numbers and ignore the names. They don’t realize that sometimes it is worse to get what they want.

7

Apparently it was a consequence of the way Kashdan handled odd men: pairing them within the next lower score group. As Kashdan initially saw it, Gligoric couldn’t be paired against Liberzon or Evans because all three players were in the top half of the combined 7/6.5/6 point score group. In his CL&R response he didn’t say anything about the possibility of interchanges, just transpositions.

I have a copy of the 2nd edition of the USCF rulebook but not the 1st edition so I don’t know what was considered “legal” in those days. I also don’t know if they were using USCF or FIDE pairing rules, assuming FIDE even had rules for Swiss pairings.

8

Frankly, unless there was some pairing rule that forced Kashdan’s odd-man handling or some other heretofore unrevealed circumstance, I’d have to say that Evans’s pairings would probably have been my choice.

9

Of course, Evans’ pairings are great for Evan. One thing not considered is whether Kashdan liked to take suggestions or whether he liked/disliked Evans. It would not have been the first tournament I saw in that era where personality played a part in pairings. All perfectly legal, of course, but also interesting to see how different TDs would pair.

10

That appears to have been Kashdan’s take on it. He wrote in CL&R, “It is a matter of judgment, and I still feel that I acted in the best interest of the tournament and all the players. I am not certain that Evans was equally well motivated.”

Even though the Evans - Weinstein pairing would have been favorable to Evans, from the vantage point of 2014 it looks like the right pairing, at least according to our present rules. Maybe the controversy in 1975 was a factor in making the pairing rules clearer now than they were then.

By the way, I’m not sure why the pairing would have been Evans - Weinstein and not Weinstein - Evans. Both Evans and Kashdan assumed that Evans would be White, but both players were due for Black and Evans was higher ranked, which according to the 2nd edition rulebook would mean that Evans should get Black unless that meant giving Weinstein the same color three times in a row.

Edit: “…unless that meant giving Weinstein” instead of “unless that meant giving him [Evans]”.

11

I have the 1974 1st edition. Relevant is:
14. The odd man is paired with the highest-ranked player he has not met in the next-lower group.

It doesn’t say anything about trying to preserve the integrity of the lower-score group, so Kashdan’s pairings appear to be correct by the rules of the time.

12

And all of this matters why exactly? A pairing controversy from 39 years ago?

13

I had a feeling there was some missing evidence. Gracias.

14

I see it as an interesting piece of history. If you disagree feel free to ignore.

15

As I remember, at about that time, the idea of pairing the lowest in the score group as though he were the highest in the next group (Harkness method), was going out of fashion, gradually yielding to the current philosophy of pairing the lowest in the score group against the highest in the next group.

Bill Smythe

16

Chess players enjoy post mortems. This is post mortem in multiple ways.

17

Except isn’t Harkness based upon upfloating rather than downfloating? If there are seven players in a score group, 4 plays the top of the next score group, as if that person becomes number 8 to fill the higher group. If you do the pairings by upfloating, you take Gligo and then (by process of elimination) Weinstein and get Evans’ preferred pairings. It’s not clear there was ever any support in the rules for Kashdan’s original idea.

18

Martinak’s post indicates that Harkness was not an issue. Instead it was the first edition wording (having come out a year earlier).

"I have the 1974 1st edition. Relevant is:
14. The odd man is paired with the highest-ranked player he has not met in the next-lower group.

It doesn’t say anything about trying to preserve the integrity of the lower-score group, so Kashdan’s pairings appear to be correct by the rules of the time."

Based on that you start by finding the board one opponent (can’t be Evans because that would be a rematch). Then you find the board two opponent, which left two 6-point players that have already played each other, making them both the odd man. You next find the boards 3 and 4 opponents.

It would not be done that way today by people following the current pairing rules, but it sounds valid based on the pairing rules then in effect. It is also the ultimate in top-down logic.

19

Kashdan’s original pairings (with Liberzon playing Weinstein) are the ones for which I can’t find any justification. The final ones seem to be correct given the wording at the time.

20

Here’s how Kashdan explained his original pairings, with Liberzon playing Weinstein. He doesn’t say anything about Harkness.

"Since the top two had already played, my normal system is to drop both to the next group, and then pair that as a six-player group. The cards were in the order Liberzon, Evans, Gligoric, Gheorghiu, Quinteros, Weinstein. Move the top three in one pile, and the bottom three in another. The first try is Liberzon-Gheorghiu, good except for color; Evans-Quinteros, already played; Evans-Weinstein, good except for color; Gligoric-Quinteros, good except for color. This is Pairing C, legal, but all three wrong in color. Since four players needed black and two white, it was possible for two of the pairings to have the correct colors, and I should certainly try for that.

"Let us try again, with the cards back in the two piles: Liberzon, Evans, Gligoric, and Gheorghiu, Quinteros, Weinstein. Liberzon has played Quinteros, so Liberzon-Weinstein, good including color; Gheorghiu-Evans, good including color; Quinteros-Gligoric, good except for color. This is Pairing A. It is legal, with colors correct in two cases, the best that can be done. This was my original tentative pairing, which Evans had objected to so vehemently.

“Had Evans not objected I might have stopped here, and perhaps that is what I should have done. However, he had a point. This was a very unusual situation. The leader was a full point ahead of any possible opponent. Perhaps he should play the strongest man available. At least it was worth considering. …”

It sounds like Kashdan had his own system of handling odd men independently of what was in the USCF rulebook, but he had second thoughts about it based on Evans’s argument that Liberzon should face the strongest possible opponent. As Kashdan said in his letter: “I have frequently stated that the rules must not be made so inflexible that the director can exercise no judgment whatever. Leeway for the director could of course lead to abuses, but really unusual situations may call for special action.”