Fairy-Max 4.8 released

[hgm]

Also note that I really had no idea on the piece value of the Archbishop when I slammed together the fmax.ini piece description file. So it is quite possible that Fairy-Max voluntarily engages in losing trades.

As a wild guess I put A=18, on a scale R=12, C=22, Q=23. On a centiPawn scale this would read approximately as R=490, A=740, C=905, Q=950. And the other piece values are inherited from 8x8 Chess, which could also make a differece.

[GothicChessInventor]

Hello again,

You might want to take a look at this link:

http://www.gothicchess.com/piece_values.html

That discusses a starting point for deriving piece values for Gothic Chess on the 10x8 board.

Also, here is a link to many Gothic Chess games played on the old BrainKing site:

http://www.gothicchess.com/bk_archive.html

I have some of the top players' games there.

And will any more programs wish to enter the 2007 Gothic Chess Computer Championship? You can see we have 8 programs this year so far.

Please let me know!

[GothicChessInventor]

One that i designed myself and one that i designed looking at the pieces from the program Gothic Vortex of Ed Trice. Perhaps Ed may confirm here that i did not violate any copyrights here if i allow this font for downloading.

George,

Feel free to distribute your font with the Gothic Vortex pieces for Chancellor and Archbishop. I never claimed any design rights for these pieces. You will notice Fencer still uses them on BrainKing despite the disagreements I have had with him in the past. They are public domain and free to use.

[hgm]

That discusses a starting point for deriving piece values for Gothic Chess on the 10x8 board.

OK, thanks. It seems I overestimated the Archbishop aby about half a Pawn, compared to Queen. So I gusess I should reduce its value from 18 to 17, in the fmax.ini file.

Its end-game value should be higher, though, if it can beat B+N single-handedly.

[smrf]

Let me add a table on pieve value models:



and here is SMIRF's proposal for C / A piece icons:



Reinhard.

[George Tsavdaris]

Let me add a table on pieve value models:

On your old page you had the following method for calculating piece values:

The strength of any piece has a linear relation with the sum of: the number of all affected squares which could be covered at all by the chess piece on an empty board, calculated for every possible square it can go(this concerns all ever possible targets of defense or attack) and the number of all affected directions into which a chess piece can radiate its effect on an empty board, calculated for every possible square it can go (this concerns the maximum number of figures targetable at the same time), divided by the number of all possible squares the piece can legally go on an empty board.


That gave different values from what you have on this table now. For example you gave back then Queen=9.03, but now you have 9.6.

Have you changed your method? If so then can you tell what you are doing now?


Also in my opinion after enough experience with Gothic Chess(but i guess it would be valid for other 10x8 variants), i believe that both you and Ed Trice underestimate the value of Archbishop and that a much higher value- a bit less than Chancellor(at least in the middlegame)- would increase the strength of your programs....

After combining some methods of calculation i give the values:
Pawn=1
Knight=2.62
Bishop=3.02
Rook=4.86
Archbishop=9.12
Chancellor=9.28
Queen=9.54

If i read and understand FMax.ini file about giving the values of pieces to FMax i will try and play with my values against default values.....

[GothicChessInventor]

I put together an Excel spreadsheet where you can examine piece trading scenarios:

http://www.gothicchess.com/10x8_values.xls

This compares values for all authors who published values.

[GothicChessInventor]

[GothicChessInventor]

Its end-game value should be higher, though, if it can beat B+N single-handedly.

I computed the 5-piece tablebases for Gothic Chess. This gives a very clear indication of how the pieces do against one another.

http://www.gothicchess.com/db_5_stats.txt

Also, Archbishop vs. Bishop + Knight has a max win of 202 moves!

http://www.gothicchess.com/endings/kakbn/game.htm

Other Gothic Chess endgames:

http://www.gothicchess.com/javascript_endings.html

And I think I am one of the first people to show the 7-piece tablebase results for 8x8 chess with a replay utility:

http://www.gothicchess.com/javascript_8 ... dings.html

Feel free to share the links.

[smrf]

Well, I have added a mobility component c) since long:

a) Wf : the average of all affected squares which could be covered at all by a chess piece (this concerns all ever possible targets of defense or attack)

b) Wr : the average of all affected directions into which a chess piece can radiate its effect (this concerns the maximum number of pieces targetable at the same time)

c) Wm : the mobility value: √(Wf³ - Wr³) * boardusage / (width - 1) / (height - 1)

Reinhard.