micawber wrote on 08/21/08 at 17:30:28:
Hi Markovich,
In general I can agree with your criteria. But just a few remarks to get the discussion started.
1. The criteria can conflict.
It's white's move and there are many plausible continuations can conflict.
There are quite a few positions where white has an obvious best continuation, and black has many plausible replies (choices of plans) that determine the characteristics of the position.
2. The condition occurs frequently in practice is a bit subjective.
There are some positions that occur frequently in my practice,
that are perhaps relatively rare in terms of general statistics.
Perhaps these could be termed 'personal tabia'.
For me it is natural to study these positions with more care than others, based on my opening repertoire, style and taste.
I find the concept of tabia specially usefull, where a position can
be reached by many move-orders/transpositions.
Hey.
Broadly I agree. You're probably right that 'White to move' is a mistake. As I said, I think there are degrees and degrees of 'tabiya-ness.' You could call that 'subjectivity,' I suppose.
I do think that almost all positions that would qualify under my proposed criteria could be reached by various plausible move orders (for example, the Normal Position of the Evans).
I think it would be possible to write a computer program, which would have to be able to evaluate chess positions for relative chances, that would take any given collection of games and identify tabiyas based on some solidification of these criteria. That program thrown at a major game collection could be used to produce a new openings classification scheme.
There is a hierarchy among tabias; just to take an example from my recent studies, 1.e4 Nf6 2.e5 Nd5 3.d4 d6 4.c4 Nb6 5.f4 dxe5 6.fxe5 Nc6 7.Be3 Bf5 8.Nc3 e6 9.Nf3 is a tabia (a black-to-move one, proving your point about that); so, continuing, is 9...Be7 10.Be2 0-0 11.0-0 f6 12.exf6 Bxf6; so, continuing, is 13.Qd2 Qe7 14.Rad1 Rad8 15.Qc1 h6 16.Kh1 Kh8 17.h3 Bh2 18.Bg1. There are in fact several plausible move orders that produce each of these, so even in these cases clearly identified with a single system, there is some value in classifying the positions and not the move sequences.