The Harkness Rating System

One of the first posts on this series covering the introduction of chess ratings in the U.S. was The First USCF Rating System (July 2017). The main article in the post, 'National Rating System by William M. Byland, USCF Vice President in Charge of Rating Statistics', ended with

For the long labor of compilation and computation involved in these listings. which furnish an invaluable base for future ratings, we are deeply indebted to Rating Statistician Kenneth Harkness.

The Wikipedia page, Kenneth Harkness, credits him with having 'introduced the Harkness rating system, which was a precursor to the Elo rating system'. Harkness wrote his first feature article on ratings for the 5 March 1952 edition of Chess Life (CL), the same issue where the fourth National Rating List appeared; see USCF Rating Lists in the 1950s (August 2017), for a summary of all early rating lists. Titled 'Picking the Winner at Havana', the Harkness article started,

As this is written the big international tournament at Havana is getting under way. Although the final line-up has not yet been announced. the list of probable competitors includes some top-flight masters from Europe, South America and the United States. This country is represented by U.S. Champion Larry Evans, Grandmaster Samuel Reshevsky, former U.S. Champion Herman Steiner, Senior Master Israel A. Horowitz, veteran U.S. Master Edward Lasker.

Naturally, we all hope that one of our boys will bring home the bacon. And even with the price of meat these days, you can buy a lot of bacon with the $2500.00 first prize being offered by our good friends in Cuba. If our brilliant young champion Larry Evans brings home that much dough he can pay his taxes and still play bridge at the Marshall Chess Club for a fifth.

Good chessplayers being even more consistent than racehorses, it is no trouble at all for your Rating Statistician to lay down his copy of Racing Form for a few moments and give you the probable order of finish at Hialeah -- I mean Havana. Judging by their past performances, as measured by the rating system, the boys will pass the line in the following order:

1. Samuel Reshevsky, USA...2704
2. Miguel N. Najdorf, Argentina...2714

It should be a photo-finish between these two Grandmasters. They ended up one-two at Amsterdam, 1950, and New York, 1951, alternating for first prize. We give Reshevsky the edge because he has made higher ratings than Najdorf in the past and because he is out to avenge the loss of the U.S. Championship to Larry Evans last year. Sammy will play harder than ever to recover his prestige. [...]

A few months later, in the 20 May issue of CL, Harkness wrote a follow-up article. It appeared under the following banner.

'How the Rating System Works'

The article started,

Many readers of CHESS LIFE were favorably impressed by our recent forecast of the results of the international tournament at Havana. With one or two exceptions, which we will hasten to explain now that the race is over, the predictions were about as near as you can come without the use of a crystal ball.

To get some idea of how closely the national rating system measures tournament playing strength, let us compare the ratings earned at Havana with the last averages of the contestants:

['Player, Last Average, Havana Rating' for Najdorf, Reshevsky]

We predicted a photo-finish between these two grandmasters, giving the edge to our ex-champion. An unexpected draw with one of the tailenders cost Sammy the first prize, so be tied with Najdorf.

Note how the ratings earned at Havana confirm the correctness of the previous ratings -- and vice versa. A difference of less than 50 points is negligible.

[Ditto for Gligoric, Eliskases, Evans]

We claimed that any one of these three could take third prize. It was Gligoric who came in third. with Eliskases and Evans tied for fourth and fifth. [...]

This was the first in a series of eight articles under the title 'How the Rating System Works'. I'll look at the following articles in the next post on early U.S. chess ratings.