How?

<Just read off some of the names Petrosian had to climb over to get to the top of the 1961 USSR championship: Spassky, Bronstein, Geller, Korchnoi, Stein, Boleslavesky, Taimanov, Averbakh, Gufeld....

A much, much, stronger field than Ostend 1905.>

The number 10 player in the world in 1858 was Henry Bird, who, frankly, would have been a good opponent for Atkins.

To show the effects of rating inflation in Sonas's calculation, follow the career of Henry Edward Bird. The guy is #10 in the world at 2488 in 1858. By 1901, some 43 years later, Bird still shows up in the top 100. Is his rating reduced to 2200 or something, as common sense would suggest? No, of course not: Sonas rates the 1901 version of Bird at 2479, which is to suggest that this version of Bird was capable of defeating the 1858 version in a match. That's pretty good, considering that Bird would be dead soon.

Suppose that a player like Michael Adams played Anatoly Karpov, in his prime, in a 6 game match, and drew the match with him, 3 to 3.

Great result, isn't it?

Yet, under the Sonas system, Adams would probably lose rating points, or, at best, break even. Adams as 2749 and Karpov at 2833 approximate the following quote from Sonas:

"For example, although a 50% score in a 6-game match against a 2800-rated player would traditionally be called a 2800 performance rating, my formulas (which are sensitive to the number of games played) would conservatively call that a 2728 performance rating, because of the small number of games. If that single six-game match were the only results we had available for you, we would assign you an overall rating of 2728, not 2800. "

OK, that's all I need. Now I can make it simple. At Ostend Geza Maroczy scored 75% (19.5/26) against a field that you now concede included over half of the top 10 players in the world. By contrast at Curacao 1962, Tigran Petrosian scored 65% (16.5/27) against a field that included over half the top 10 players in the world. Both great performances, but Geza Maroczy had a higher rate of success than Petrosian, so he gets a higher performance rating for that event, 2837 to 2829. Now what's the problem?

<Except for the fact that Atkins was only awarded with the IM title after he died. That's a pretty good objective opinion of his strength.>

Really? Why do you think so? Maybe you could explain the criteria FIDE used to me.

From Atkins' bio on this site:

<In the British Championship which he contested 11 times he won the event 9 times viz 1905, 1906, 1907, 1908, 1909, 1910, 1911, 1924 and 1925. In his first attempt in 1904 he finished 1st= with William Ewart Napier but lost the play-off and in his last appearance in 1937 he was 3rd= at 65 years of age!>

Well, everyone knows the English can't play chess.

<His international career was made up by just six events. In the Hastings 1895 minor tournament he finished 2nd= behind Geza Maroczy.>

We know that Maroczy was a fish, so we can discount that.

<In 1899 he won the Amsterdam tourney by a whopping 4 points.>

In fact, he won all his games, but it was a fairly weak tournament.

<In the Hanover 1902 tournament he was 3rd behind David Janowski and Harry Nelson Pillsbury and ahead of Frank James Marshall and Mikhail Chigorin amongst others.>

I'm sure there's a very good explanation for this!

<The upwards trend in ratings in less than a 50 year period appears to be completely unmotivated, and harmful to players like Petrosian, who was the best player in the world for most of the 1960's.>

But...er...if chessmetrics ratings were inflated over time, then Petrosian would be OVERRATED compared to the leading players of 1905. But you seem to think he is UNDERRATED. And Fischer would be UNDERRATED compared to the leading players right now. Instead, because one of Sonas' main goals is to eliminate rating inflation, Fischer still has the highest rating ever, instead of being behind (among others) Topalov, as he is in the Elo ratings.

Here is Sonas' explanation for how he calculated his ratings. It's very user friendly. Read and then maybe you can finally make me understand how Sonas' system is prejudiced against Fischer, matches, Tigran Petrosian, Tarrasch, truth, justice and the American Way. Perhaps you'll even uncover the secret of Sonas' strange partiality for Geza Maroczy.

"For example, although a 50% score in a 6-game match against a 2800-rated player would traditionally be called a 2800 performance rating, my formulas (which are sensitive to the number of games played) would conservatively call that a 2728 performance rating, because of the small number of games. If that single six-game match were the only results we had available for you, we would assign you an overall rating of 2728, not 2800. ">

That's because of the number of games played, not the setting in which they are played. If Adams scored 50% in a 6 game tournament against 2800-rated opposition (instead of in a 6-game match against a single 2800-rated opponent) the results are the same. You see that, surely??

<A more objective, better approach to evaluating results is: every game of chess, played under official tournament conditions, is just as important as every other game played as such. Match, tournament: it makes no difference.>

Half of 10 is 5. That means there were quite a few rabbits out there for Maroczy to bash, in this 14 player tournament. At Curacao 1962, there were no rabbits, every player there was capable of playing a competitive match against the world champion. Not so at Ostend - taking a middle of the road competitor like Burn, you know that he would have slaughtered in a match against the champ.

The top players at Ostend, like Pillsbury, and Schlechter, should bring as much respect as players like Korchnoi or Geller, but at Curacao, there were MORE top players.

Well, everyone knows the English can't play chess. >

I think one of the mistakes of the Sonas system is that it thinks the ages before Botvinnik were like those of Botvinnik, where the world champion was 'first among equals'.

Sorry, but any list that has Janovski rated more highly than Emanuel Lasker has problems with it. The 'World Master' was HEAD AND SHOULDERS above his competition. Players like Janovsky and Marshall probably don't deserve a rating with 200 points of Lasker.

That brings us back to Curacao 1962.
Aside from Petrosian, if you made the claim that Keres, Geller, Fischer, Korchnoi, or Tal could beat the world champion, nobody would laugh at you. You throw Benko and Filip in there to round out the field, but these players should be more highly regarded than Burn, who was the middle place finisher at Ostend.

The system rewards tournament play more than match play. So, for example, Fischer defeats Larsen 6 to 0. How strong a performance does Sonas think this is, 'by today's standards'? Answer: 2750.

That is simply a joke.

E Joppen vs Petrosian, 1954

Stahlberg vs Petrosian, 1955

J H Donner vs Petrosian, 1955

I A Horowitz vs Petrosian, 1955

Y Sakharov vs Petrosian, 1956

Filip vs Petrosian, 1956

Chistiakov vs Petrosian, 1956

E Terpugov vs Petrosian, 1957

Tolush vs Petrosian, 1957

Filip vs Petrosian, 1962

B Wexler vs Petrosian, 1964

Platonov vs Petrosian, 1964

J Kupper vs Petrosian, 1964

Udovcic vs Petrosian, 1965

Larsen vs Petrosian, 1966

Simagin vs Petrosian, 1966

Najdorf vs Petrosian, 1966

Minev vs Petrosian, 1966

E Paoli vs Petrosian, 1967

J Kupper vs Petrosian, 1967

R Byrne vs Petrosian, 1968

Bobotsov vs Petrosian, 1968

J B Bednarski vs Petrosian, 1968

Furman vs Petrosian, 1969

V Kovacevic vs Petrosian, 1970

Z Basagic vs Petrosian, 1972

R Bogdanovic vs Petrosian, 1972