CCRL Discussion Board

I agree it will never blunder, but it will make moves that no human would play. E.g. queen sacs just to convert to another won position as fast as possible. I'm sure that if you gave a human a choice between say KQRRKQR or KRK, they would pick the latter every time. Most humans would trade or sacri...

I don't understand what you mean by a DTZ50 sometimes leading to 'unwise' moves It will never throw away a win or a draw. If you set it to minimize DTZ50, it will also avoid making silly moves, in other words moves that are obviously inefficient. It is true that DTZ50 isn't sufficient to construct t...

Unfortunately, there are no hard and fast rules on how to attempt to swindle against a fallible opponent. There are various hereustics you could use, for example, make the move that gives the most opportunities for the opponent to blunder, make the move that maximizes DTC/DTZ, try to convert into en...

So what would it be used for? Evaluating possible future positions in the computers search tree?

As I understand the rules, checkmate ends the game immediately and hence, a draw cannot be claimed.

By the way, what exactly, does a WDL50 table store?

syzygy wrote:Perform 1 ply and probe the table again. By definition, the position after 1 ply is d-1 plies from a zeroing move. The value returned by the table for that position is either k or k-1.

This isn't true if the ply was a zeroing move

So how would your system handle the positions posted by kronsteen?

I was talking about the whole only store every other ply thing
If you only store every second ply, the computer will have to search an extra ply deep at runtime.

Bside, if you were going to do that, it would be more efficient to just store all of the WTM and omit the BTM tables

So basically, you are sayign to only store half the table and make up for it by having the GUI search two plies in depth at run time?

Why not go farther and only store every fourth or every eigth ply?

Ok, I'm convinced now that full moves aren't sufficient

While either side can make a zeroing move, only one side can mate.

I thought we decided that DTM50 was pretty much useless? Is it really not possible to map plies to moves after generation without messing up the 50-move rule in border cases? You could be right, but until I've verified this for myself I hope you're wrong. I think koistinen is wrong as well. Since in...

Which still comes to a max of only 7.66 bits per position (for practical purposes, you'd use a byte) Thats why DTZ50 is so useful: you only need a byte per position no matter how complicated a position it is, and it is still sufficient to win a won position or draw a drawn position. The only reason ...

The main reason that I said 100 instead of 50 numbers was because in most complicated endgames, there are at least a few positions where either side could mate. Now that I think about it however, you could actually do with just 50 numbers as you said, all you would need is an extra bit to show who w...

I guess ultimately, it comes down to storing what you can and then searching the rest

I guess it all comes down to what you want to use the tablebase for. In an engine game, a mate is a mate no matter how long; all you are about is whether you can force mate or not. In this case, all you need is a single bit per position and movecounter, which is equivalent to the DTZ50 mentioned abo...

Are you saying that you only need to store the value of each position when the move counter is at 50? That would be sufficient to tell you whether a position was won, but it would not be sufficient to actually tell you what the winning moves are. Consider three positions, A, B, and C Position A is a...

I am really confused by all this talk of DTR, DTZ, DTZ50, DTS- + etc It seems to me like there is only one relevant metric under the fifty move rule: the minimum number of moves required to mate from a given position with a given move counter state. Any one position is actually 100 positions. For ex...