Friday, February 15, 2013

A typical game for the variation

In his comment on last week's piece about the sudden disappearance of Chessville, Mark Weeks was kind enough to draw our attention to the existence of, thus enabling us to show you another curiosity from Ray's Chessville oeuvre.

Just for once, this isn't about plagiarism, but - as far as one can see - about making things up. Small things, but puzzling ones.

Let me explain. This involves Ray's first Chessville column. His columns aren't (and weren't) dated* and the URLs don't carry any dates either, but according to The Closet Grandmaster, writing at the time, this was "his first effort", in July 2006. As you see, it was called An All-Purpose Black Defence and Ray kicked off by looking at a game between Spassky and Larsen from 1979.

It's a long piece and we won't bore you with all of it, but, after the Spassky-Larsen game, which Black won with the Caro-Kann, Ray turned his attention to the Slav. Not just any branch of the Slav, either, but the Chebanenko.

The game he chose to illustrate this variation with was Cebalo-Keene, Student Team Championship 1967. This surprised me a little as, though Ray was a great openings innovator in his day, he never to my knowledge played the Chebanenko, nor indeed do I recall many games in that variation from as far back as 1967.

No matter, because what we seem to have here is a transposition into the Chebanenko, which must have taken place by the next diagram - "the following is a typical game for the variation and we join it after Black's 8th move". In the diagram, after that move, we see our Chebanenko pawn, on a6.

The next diagram, pictured just above, occurred after White's sixteenth. At this point Ray broke with 16...e5! and took over the game.

Having outplayed Cebalo, Ray missed a nice shot...

...but it didn't matter as the game was practically over. "My opponent screwed up his scoresheet and threw it on the floor." What a memory the man has.

Or maybe not, since something is awry with Cebalo-Keene, at least if one is using it to illustrate the possibilities afforded by the Chebanenko Slav, in which, as Ray observes,
"the key move is....4...a6".
To see why, have a look at Cebalo-Keene, with comments by Ray Keene. Not on Chessville, but on Chessgames.

You will notice that it doesn't actually involve a Chebanenko Slav. Not to worry, we already knew it was a transposition. Or so we were given to understand.

In the Chessgames version, it went thus:
1.d4 g6 2.Nf3 Bg7 3.c4 d6 4.Nc3 Bg4 5.e3 Nf6 6.Be2 O-O 7.O-O c6 8.b3 d5
which brings us to the following position.

Cebalo-Keene, position after Black's 8th.

This is the position in which we were supposed to "join" the game from the variation Ray was advocating, which, you will recall, went as follows:
1.d4 d5 2.c4 c6 3.Nc3 Nf6 4.Nf3 a6 5. e3 g6 6.Be2 Bg7 7.O-O O-O 8.b3 Bg4
which, however, reaches not the position above, but instead, the position given in the piece. We give it again, below. Can you spot the difference?

Cebalo-Keene according to Ray on Chessville, position after Black's 8th.

Yes indeed, the pawn in the Chessgames version - let's call it the original version, or better, the real version - is still on a7. Ray has us transposing into a position which didn't actually occur.

This is a passably strange way to proceed, especially if you're illustrating a variation whose whole point is that one plays ...a6.

So when did that move take place? Had it cropped up by the subsequent diagram, showing the position before Ray's 16...e5 ?

It had not. Although there is, of course, a pawn on a6 in the Chessville diagram, in the real game it was still at home on a7.

Cebalo-Keene, actual game, position before Black's 16th.

In fact the pawn didn't move from a7 until move 20 - when it skipped all the way to a5 in one move.

Cebalo-Keene, actual game, position after Black's 20th.

Yes, indeed. The move ...a7-a6, which is what the Chebanenko is all about, never took place.

In the absence of a better explanation, we seem to be drawn to the conclusion that Ray chose to discuss the ...a6 line in the Slav Defence by providing positions, from one of his own games, that never actually existed. This entailed selecting an illustrative game in which the key move never actually occurred.

There's something brilliant about this. Something totally, characteristically Ray. I mean what's the point? What would possess somebody to engage in such a trivial fiddle?

At first, looking at the discussion on Chessgames, I thought I had an explanation: that somehow Ray, notoriously fine memory notwithstanding, had become confused between two different versions of one of his own games that were circulating on databases.

But apparently not, since the exchange above occurred in 2008, two years after Ray's Chessville piece. (The comment from "bettermove" includes Ray's own annotation from that piece. I wonder if they were aware of that, or not?)

So why pretend you played a position you never played? What's the point? I confess I have no idea. I mean, I know Ray is lazy but you'd think it'd be less hassle just to come up with a real game.

But he didn't.

[* Ignore "today is February 8", that's the date I accessed it. Odd, though, that that function still updates although the site has gone offline.]
[Thanks to Angus and Jonathan]

[Ray Keene index]


Anonymous said...

The 1967 game started as one of the Nf6, g6, Bg7, d6, c6 systems, which as we know, Ray popularised in the UK. White made no particular attempt to blow it away, so Black played d5 to transpose into a Slav with loss of tempo. It's more than possible that Ray was familiar with the middle game he was heading for and didn't think a6 was particularly useful in that exact position, so the loss of time was unimportant. There had been a game Uhlmann-Fairhurst from the 1964 Olympiad which had reached the version of the position with a6. That even used a 4 .. a6 move order.

It's an interesting practical idea, being able to transpose from a Modern or Kings Indian into a Schlecter or Chebanenko Slav, even a tempo down. Studying the Slav isn't something that players who play g6 and Bg7 normally bother much with.


Jon H said...

Nothing sums up the Penguin more than this pathetic example...

ejh said...

It's more than possible that Ray was familiar with the middle game he was heading for and didn't think a6 was particularly useful in that exact position

No, it's less than possible, given that (as I point out in the piece) he says

"the key move is....4...a6"

and is discussing the ...a6 Slav.

Whatever explanations and excuses there may be, the idea that he thought ...a6 irrelevant is a total non-starter. If he'd wanted to say "imagine this game, but with ...a6 played" he could have done so.

ejh said...

Incidentally, Chessgames have Uhlmann-Fairhurst deriving from a 4...g6 move order and the a7 pawn never moving (it's captured on move 22). Is that not right?

Anonymous said...

Re Uhlmann-Fairhurst, The game is from 1964 rather than 1958. doesn't have it. for the Scotland matches of 1964

Given the back story of the previous 1958 game, 4 .. a6 may have been a novelty designed to not follow the previous game.



ejh said...

Ah, I see. Now isn't that interesting. In Bologan's book on the variation he cites an article by Olthof in New In Chess Yearbook 81 (2007) which suggests it was first played in Michell - Sultan Khan, British Championship 1929. I wonder if Olthof gives the Fairhurst game?

(Either way, though, it's intrinsically unlikely that anybody, selecting a game to illustrate the a6 Slav, would not notice whether or not that game included the move a6.)

Anonymous said...

The games for the 1929 British appear not to have made it to the databases. Only NIC has the Michell-Sultan Khan game. Fairhurst utilised the variation, as he played it at the 1958 and 1966 Olympiads as well.

NIC have an online database at
This has all three Fairhurst games from the Olympiads.

I agree about the game identification. It would be possible to search for a position with the pawn either on a6 or a7, but slightly complex. If you look at the comments on, the discrepancy had already been noticed.


ejh said...

Indeed, a discussion that you can read in the piece on which we are commenting....

Anonymous said...

The match containing the Cebalo game is reported in Ray's report in a 1967 BCM. The opening is referred to as a Schlecter Grunfeld.

I could believe that Ray's archive of material pre-dates computers. Could it be that the game and notes were wrongly classified when first stored? Still a pawn on a6 isn't the same as one on a7.


ejh said...

I could believe that Ray's archive of material pre-dates computers.

I'd be amazed if it didn't.

Could it be that the game and notes were wrongly classified when first stored?

Quite likely, but it's not likely that it was classified as an a6 Slav. Or that anybody, ever, thought it was one. (At least, not until Ray's Chessville piece.)