CCRL Discussion Board

Oh it does. The "+" implies there is no bound on the distance and one must accommodate single and double-byte entries. So in that regard the simplicity gained in going from DTM to DTZ50 is lost. True, but that holds for DTZ as well. For the file format itself it shouldn't matter a lot, bu...

I do. Difficulty of writing correct code, complexity of the file format, ease of use, and uncertainty of validation. In the simple-versus-complicated debate I side solidly with Kirill. In my opinion only "difficulty of writing correct code" is a relevant argument. The other aspects you me...

Interpreting probe values of such a DTZ50+ table is quite simple. Let v be the value for a position. If v = 0, the position is a genuine draw. If 0 < v <= 50, the position is won under the 50-move rule. if v > 50, the position is a draw, but won if you ignore the 50-move rule. if - 50 <= v < 0, the ...

As I realize now, identifying cursed wins on 7-men won't need only running recursive algorithm beyond z>50. It will need a second full run (first one for true wins can be stopped at z=50) as cursed 7-men positions can be connected to cursed 6-men, but true wins on 7-men must lead to true wins on 6-...

I can understand that there is much interest in theoretical aspects of EGTB, especially among the theorists gathering here. :-) But please consider that we need something very practical and simple for wide adoption. And we need wide adoption to compute the important 7-men tables sooner. If it turns...

WDL tables give three possible position status. This is to be stored on 2 bits. I'm not sure of this, but a fourth position status can be used for free as it doesn't change the number of bits and therefore won't change final size of compressed data. A fourth position state will affect final size, u...

What kind of info is most relevant for these positions ? DTR info (i.e smallest k which would be required for position to be a win on a k-move rule, with k>50 necessarily), or DTZ info (i.e min number of moves, necessarily >50, required from current position in order to reach a zeroing move which l...

Actually, I believe that this idea can be generalized. Let (k) = 0 < k_0 < k_1 < k_2 < ... < k_n. Now define DTR(k) for position p as the pair: (min{i : p is won under a k_i-move rule}, number of moves to complete current phase) Now: DTR( 1 < 2 < 3 < 4 < ... ) = DTR DTR( inf ) = DTZ DTR( 50 ) = DTZ5...

Of course, there is no reason why the computer should not go down the DTZ-, DTC- or DTM-optimal line, provided there's no way this can land it with a superlong phase. But, as instanced above, there are positions where none of the DTC/M/Z-optimal moves avoid an avoidable 50-move draw. Indeed, so hav...

DTR = the largest move-count that will be seen in a winning line to mate, provided White/Black are minimaxing this quantity. Wherein move-count is the value of the "50-move counter". (Not the number of moves in the winning line, as that would give DTM.) DTZR = the least number of moves to...

I'm not sure if I am missing something, but you need at least 2 bytes to store a DTZ, and only 1 byte to store a DTZ50. It may contribute to the difference? A special purpose compressor will compress symbols corresponding to position values instead of individual bytes. If I am correct in estimating...

There's another way to optimise compression vis-a-vis the 'broken' positions. Since they are never going to be accessed, they can be set to W, D or L depending on what maximises the compression - after you have noted the correct stats. I don't think anyone has done any experiments on this. My compr...

We already have seen some long checkmates reported by Marc Bourzutschky. Yes they attracted a lot of attention. The result is that no one is working in this direction anymore, and we don't have any usable 7-men tablebases. Interesting point of view! You could well be right. I agree that speed and s...

DTZ has the advantage of being immediately connectable to existing 6-men in DTM format (which is not the case of DTZ50) but it will lead to strange knowledge providing DTZ evaluation for 7-men and DTM evaluation for 6-men or less. If we are to reprocess all 3-6men to DTZ in order to avoid this, bet...

I don't think DTZ50 will save much space over DTZ. If it does, the compression algorithm needs fixing. Only very few positions will have DTZ>50, so few that it should not have a noticeable impact on the compressed size. DTZ50 has not only space advantage over DTZ. DTZ50 is also more relevant to the...

I don't think DTZ50 will save much space over DTZ. If it does, the compression algorithm needs fixing. Only very few positions will have DTZ>50, so few that it should not have a noticeable impact on the compressed size. (IMO compression should not be done on a byte-level, but on a position-level bas...