07 August 2017

USCF Rating Lists in the 1950s

Two recent posts covered the introduction of USCF ratings in 1950:-

The second USCF rating list was published in the 5 March 1951 issue of Chess Life (CL). It was introduced by an editorial titled 'National Ratings', signed Montgomery Major, the editor of CL.

In this issue we publish the second list of National Performance Ratings, as of December 31, 1950. Before we comment upon certain of the changes in rating of individual players, it may be well to state that for the second and last time, we publish the complete list of available ratings of U.S. chess players, regardless of membership in the USCF. It was not originally our intention to do this, but the National Rating System has aroused so much interest throughout the country that we feel it is a very definite service to chess to publish once again the whole list for comparison and study.

Hereafter, however, below the rank of master we well publish only the names of USCF members. This is not an attempt at dictatorship as we have been accused illogically by certain readers, but merely the recognition of a basic principle in America that the average American expects and is willing to pay for what he receives, provided he gets value received for his money. The cost of maintaining a rating system is considerable, for it demands careful statistical computations by a trained statistician. Such expert services cannot be obtained without charge, even if the charge is a nominal one in comparison with the work accomplished. Since the members of the Federation are footing this bill with their dues, it is only just that the benefits should be exclusively theirs. And since any chess player can become a member of the Federation tor the nominal dues of $3 per year, it is obviously ridiculous to claim that this restriction of published ratings to the USCF membership works any hardship on any player. Any player, who cannot persuade himself to part with $3 for the many benefits to chess provided by the Federation, cannot be very much interested in the rating system, however much he may profess to be.

Comparison between the first list of ratings and the second list provides some interesting studies as well as a very definite illustration of the effect of the "lag" in computation of ratings. But quite possibly some readers may be confused by the apparent discrepancies, and tberefore the subject demands a few illustrative comments.

It will be noted, undoubtedly, that Isaac Kashdan has dropped from the list of senior masters to the list of masters, and that he has done this without competing in any rated event since the list as of July 31 was published. This may at first glimpse seem illogical, but it actually is a very clear example of the principle of the "lag" in computing ratings. The ratings (as of July 31, 1950) covered each player's career from 1947 through the first half of 1950, and the published rating was his highest rating in any one of these four periods.

In the case of Kashdan, 1947 was a gala year. He won the U.S. Open Championship at Corpus Christi. This bolstered up an already high past record of performance, including his second to Reshevsky in the 1946 U.S. Biennial Championship. But 1948 told a somewhat different story. Kashdan only placed second in the 1948 US. Open Championship at Baltimore, and again was second in the 1948 U.S. Biennial Championship at South Fallsburg. So, when the performances in 1947 were removed from the current computation In the listing as of December 31, Kashdan's rating then was determined by his highest scoring in 1948, 1949 or 1950 and the resultant drop in his performance rating reflected his less successful appearances in recent tournaments. It is noteworthy that if Kashdan had followed his poorer year in 1948 with a more successful performance in 1949 or 1950, due to the "lag" procedure in rating, his one bad year would not have made any appearance in the ratings. Thus the "lag" tends to protect a player against one bad season, but cannot continue to bolster up his ratings over a period at years.

As Kashdan through a series of circumstances, including illness, has not competed in any rated event since 1948, he has not had an opportunity to reestablish a senior master rating performance.

An illustration of the reverse principle in the "lag" comes from the advent into the master class from the expert group of Eliot Hearst, F. S. Howard and Walter Shipman in the ratings as of December 31, 1951. These younger players began to be felt in chess as far back as 1946 and 1947, but their climb into the master class was slightly delayed by the drag effected by the lower performance points of their earlier chess career. To overcome the effect of this "lag" it was necessary for each of them by consistent performance to prove that their successes were not merely a flash in the pan. For Eliot Hearst it was the New York State Championship in September, 1950 that provided the ultimate boost into master class. for F. S. Howard it was the New Jersey State Championship.

In many cases, it will be noted that there has been no change in the rating. These players have not competed in rated events in the last half of 1950, while their peak period of performance has been since the year 1947, so no change is effected by removing the choice of 1947 from their performance rating basis. Their standings will only be effected by their performances in 1951 in tournaments yet to be played and rated.

The composite image below shows some of the highlights of the second list.

A total of 16 rating lists were published in the 1950s. Some years, e.g. 1954 through 1956, saw only a single list.

CL 1950-11-20; as of 31 July 1950
CL 1951-03-05; as of 31 December 1950
CL 1951-10-05; as of 31 July 1951
CL 1952-03-05; as of 31 December 1951
CL 1952-10-05; as of 31 July 1952
CL 1953-05-20; Spring 1953
CL 1953-12-20; Fall 1953
CL 1954-06-05; Spring 1954
CL 1955-05-05; Spring 1955
CL 1956-05-20; as of 31 December 1955 (A)
CL 1957-05-20; Spring 1957
CL 1957-08-20; as of 31 March 1957 (B)
CL 1958-03-05; as of 30 September 1957 (C)
CL 1959-02-05; as of 30 September 1958
CL 1959-08-20; as of 31 May 1959
CL 1959-12-05; as of 30 June 1959 (?; D)

(A) List No.10; 'New Classifications Adopted'
(B) 1st 1957 Supplementary List
(C) 2nd 1957 Supplementary List
(D) Supplement No.1; signed Frank R. Brady, USCF Rating Statistician

Kenneth Harkness retired mid-1959. The 1950s were the 'Harkness' decade for U.S chess ratings.

No comments: