CCRL Discussion Board

Hi

I analyzed some chess endings and I rechead this position:
8/8/8/2p5/8/8/1kP5/3K4 w - - 0 1

The Tablebase 3.2 engine says that it i lost for white. The engine says that 1. c4 will loose in 13 moves.

But the position is a draw with 1. c4

Where can i find the Tablebase 3.3 engine and the TBAccess.dll (the new one)?

Have a good day

Black Bubble

Hi again

I am analyzing a lot of other positions and it seems when white (or black) makes a double pawn move the analysis are completely wrong. (must be a draw but he marks lost)

4k3/4p3/4K3/8/8/8/4P3/8 w - - 0 1
8/8/8/3p4/8/5K2/k2P4/8 w - - 0 1
8/7p/8/8/6K1/8/1k5P/8 w - - 0 1

Strange behavier of the engine.
I downloaded the whole 3 and 4 pieces again, checked md5 sum, and used only them.

I didn't use 5 and 6 man to be sure there is no problem with them.

Can someone confirm this behavier of the engine tablebase.eng (3.2)?

Thanks

Black Bubble

Can you try some different engine? I don't know what is "Tablebase 3.2" but all is working fine for me with engines like Crafty or Spike. Also you can try to analyze using "Wilhelm" software.

Hi

I tried with other engines, there seems to be no problem.

The engine i have is the tablebase.eng (i think it was on the CDs Endgame Turbo 1, I don't remember).

I only wanted to know from where this behavior is coming.

I wanted to use Fritz 10 to analyze my endgames but on the Chessbase Site they say that for use with 6 man I have to use the TBAccess.dll (seems to be Tablebase engine 3.3), but I don't know where to get it without bying Endgame Turbo 3. Is it freeware this file or not?

http://www.chessbase.de/support/support.asp?pid=645 (in german)

But tablebase.eng is a lot older. It exist since the 5 men tablebases, and I thought his analysis would be perfect, but this is not true in some cases.

If someone has tablebase.eng, can he try if this behavior is the same.

Thanks

Black Bubble

The Tablebase 3.2 engine works fine for me under Chessbase9. In your first position it shows the draw after 1.c4.