<The fact that Black has a very bad position after <1 d4 d5 2 c4 e6 3 Nc3 Nf6 4 cxd5 exd5> and after <1 d4 Nf6 2 c4 g6 3 Nc3 d5> certainly does not [instill] much faith in Black’s ability to find a satisfactory defensive set-up.”>

Reversing the clauses shows the contradiction more clearly:

<Berliner's rigorous and scientific approach to chess analysis is described in his controversial book <The System> >

A rigorous and scientific white repertoire in less than 200 pages?

A. Players have ignored it since it was published.

Just curious, how many pages do you think would be required?

As for (A), that does make a book controversial. If anything, it might suggest it just the opposite.

As for (B), you've just targeted about 90% of the contemporary published chess literature, haven't you?

To present a winning plan for white (or at least gaining a solid advantage) from move one?

Thousands of pages.

Elite chess players are result-oriented, do we agree? If Berliner had actually presented as good an opening system for white as he claimed to, everyone would be playing it today.

But believe me, whoever reviewed it for "Inside Chess"--it must have been Benjamin or Donaldson--really, really hated it.

From the introduction:

<There seems to be some point where one side is far enough ahead to force his opponent to make concessions as discussed above. We believe that point also exists at the start of the game. The fact that Black has a very bad position after 1 d4 d5 2 c4 e6 3 Nc3 Nf6 4 cxd5 exd5 and after 1 d4 Nf6 2 c4 g6 3 Nc3 d5 certainly does not instil much faith in Black’s ability to find a satisfactory defensive set-up. We have not as yet found any refutation of the Nimzo-Indian Defence, but are close. However, if the analyses of the Queen’s Gambit Declined and the Grünfeld Defence are correct, then the future will certainly bring further refutations, and chess will be a solved game by the year 2030. Many will wring their hands at this, and hope that I am wrong.>

(the book is so good you posted twice.)

Computers via tablebases are solving chess backwards. There maybe a huge leap in technology ( they will call it 'The God Chip') and come 2030 a 32 piece tablebase will be on everyone's mobile phone.

I'm very much in the market for a refutation of the Nimzoindian, given that I seem to be struggling for equality against it at times. So let me know when that gets figured out, <jim>.

But as I've often said, every number number is a lie without an error bar.

One also has to be very suspect of sampling bias for this screwball opening.

And what's <CG> criteria for game selection? (Are we filtering on min ELO, average ELO, recent tournament play, etc. etc.)

<Millbase> gives 1.Na3 a 55% scoring (21 games), dropping down to the expected 42% if Black answers with 1...e5 (6 games).

I'm very much in the market for a refutation of the Nimzoindian, given that I seem to be struggling for equality against it at times. So let me know when that gets figured out, <jim>.>

Berliner was a crackpot, right up there with Weaver <White to Play and Win> Adams. (Weaver in action: W Adams vs G Kramer, 1946) But you knew that. The Nimzo-Indian is Black's best defense to 1.d4, scores 49% for Black, and will never be refuted. btw, I heard Peter Svidler recently say that all mainstream openings will turn out to be drawn with best play, with the possible exception of the King's Indian, which might just lose for Black.

<Speaking seriously, I do think the terrible Black score for the QGD is meaningful. Not that super-GMs, who are generally satisfied to survive with Black, need to choose their defenses that way, but I know that, for example, FSR avoids really low-scoring lines.>

The rest all have White wins of between 37.2% (for 1.Nf3) and 38.5% (1.b3, as you note). Black's winning percentage for these openings varies more widely, from 24.7% (1.Nf3) to 33.4% (1.b3). So 1.b3 is both the winningest and losingest opening move of those five, making it overall the worst-scoring move of them. I am surprised to see that 1.c4 is White's best-scoring move in the database, with a 12.7% gap between wins and losses, while 1.Nf3 (No. 2) is 12.5%.