"They say you always move the wrong rook. This is the exception, but White could never have dreamt of the game continuation where his choice of rook becomes decisive."

(forgive me if there's a slight change in wording, it is more than 30 years ago I last saw the book...)

After 26.-,Rxb3, if the rook on a1 would have been on d1 instead, then 27.Qc1 would not be possible, and on 27.Rd-c1 black calmly answers -,Rb-b8.

I don't think Larsen implied anything more than he said, but your statement that the rook would 'do nothing' on d1 seems to me to indicate a limited understanding of how rooks are used.

There are several good arguments for keeping the rook on the d-file. That they may not be good _enough_ is another matter.

E.g. by playing 24. Rac1, White keeps one rook at the c-file, while still keeping a rook to support the central files d, e and f. Which, in general, are more important than the a and b files.

Additionally, a rook at d1 has the option of moving to the 3rd rank, where it can participate in attack/defense on the queens wing.

It is very common that rooks 'do nothing', as you call it, else than e.g. backing pawns (in the centre). This 'doing nothing' for a long time is one reason why deciding on moving the right rook to the right square can be very hard, even for very strong players.

For these reasons the tongue-in-cheek expression "You always move the wrong rook" was coined, and Larsen's comment hits the nail on the head.

In this position, if White plays 24.Rac1, there is 24.-,Qa4 to consider.

And, as we know, 24.Rdc1 prevents the obvious -,b3 move.

<...and it is grossly underestimating Korchnoi to suggest he may not have seen the three-move sequence leading to 27.Qc1, or the general character of the game continuation.>

I also didn't see Larsen suggest that?

What he said was, that Korchnoi couldn't have dreamt about what happened, which in plain English means that he would never have expected it to happen.

That strongly suggests that Larsen would have expected both Karpov and Korchnoi to see it.

Well, no, it wouldn't be considered by either of these players, or even a decent club player, since it pretty much loses immediately to the (not that hard to spot) 25.Ba6, which wins the exchange by force.

After your position 24.Rac1 Qa4??

click for larger view

And now after 25.Ba6

click for larger view

I don't think Larsen implied anything more than he said, but your statement that the rook would 'do nothing' on d1 seems to me to indicate a limited understanding of how rooks are used.>

I looked at the position again and I fully stand by my original statement. Rdc1 is far superior to Rac1 and the rook would do nothing on d1 in this case. Of course it is often a good idea to have a rook supporting a central pawn, but this d-pawn is not a passed pawn, and has no chance of becoming a passed pawn any time soon, and it is not even being threatened in any way, which means that the rook on d1 is just passively overprotecting an already solid pawn.