Comments/Ratings for a Single Item

2 George Duke:

These are other units, if you look closely. And when I sayed "unique", I add "within system", i.e. within these 3 races, with no repeating. For example, the Centaur is not unique outside the system, because is identical to the modern Elephant.

http://hgm.nubati.net/variants/orc-elf/ - you may see all these units how they work.
For example, the Wyvern is a rook always skipping and leaping over the adjacent square. It can't leap over the other squares, only over the adjacent square, then moves as linear. 100% leaping rook is the Pegasus but he is limited to range of 3.

One question, are the point values equal to 31 as they need to be? That leads to the Pawns.

All armies are equals each other, if you summarize (equals to 41 with pawns and without kings). 31 is the only estimated number of which depends on the accuracy of the estimate.

These estimatings are mine, and I can mistake at some of them. But totally, the engine shows that all armies are equal very much (in long series of games).

Are Guard, Pegasus, and Wyvern unique as claimed? Pegasus. Let's find out in follow-up.

I'm thinking about changing names of the units:

Centaur -> Shaman (Bishop-2 leaping); he is a "spellcaster" like a Monk - classic Bishop
Phoenix -> Pegasus (Knight + Bishop)
Pegasus -> Phoenix (Rook-3 leaping)

What about this variant?

But in my opinion, the Phoenix is associated with much power than the Pegasus; and the Centaur (archer) is more warlike than the Shaman.

The logic of the old names was:
- The Pegasus (jR3) is leaper like the Knight, but is flyer/orthogonal like the Griffin/Rook
- The Centaur (jB2) is leaper like the Knight, but is archer/diagonal like the Hunter
- The Phoenix (Elvish Queen = BN) is simply a supreme, divine creature for the elves, like the Angel for Humans or the Dragon for Orcs

Are Guard, Pegasus, and Wyvern unique as claimed? Pegasus. Let's find out in follow-up.

Oh, I remember that:

Guard = Dog in the Space Chess (by Alex Erohno), but this piece is not famous in the Wiki.
The Dog can promote only to light units and can e.p. capture only other Dogs (because of equality with the classic pawn). But the Guard has standard promotion and standard e.p. capturing, because the asymmetry is not equality one to one.

I had updated my estimatings in the main post, with better accuracy, and added the "Unit's comparrison" article.

King: Hero = 2.7

HUMANS
Pawn: Footman = 1.0
Knight: Knight = 3.25
Bishop: Monk = 3.6
Rook: Griffin = 5.0
Queen: Angel = 9.6

ELVES
Pawn: Fairy = 1.2
Knight: Unicorn = 4.25 (heavy)
Bishop: Hunter = 2.8
Rook: Pegasus = 4.75
Queen: Phoenix = 8.1

ORCS
Pawn: Guard = 1.25
Knight: Werewolf = 4.2
Bishop: Centaur = 2.75
Rook: Wyvern = 4.5
Queen: Dragon = 8.4

Well, the naming is not a big deal. I agree with the idea that it is much more important to make the piece images reflect the way the piece moves than reflect the name. I also try to follow this policy in WinBoard. But players of some of the more popular variants, such as Seirawan Chess, in general get very unhappy when you depict the pieces not exactly as they are used to (i.e.when you use something else than Hawk and Elephant for BN and RN).

But using a piece image that suggests a certain move would be very confusing if it clearly depicts something that matches the name of another piece. E.g. if a variant uses pieces that move like a Camel and a Modern Elephant, but calls the Camel-mover an Elephant, this pretty much rules out that I can use the Elephant glyph for this Modern Elephant.

So I agree that it is a lot of merit in the idea that all pieces with Knight moves should look 'knightish'. But I only have a limited number or knightish images available amongst the XBoard built-in pieces, (and even fewer in WinBoard). These are Knight, masked horse (Nightrider), Unicorn, Knight-Rook chimera (Chancellor), Zebra and winged horse. And the Zebra does not seem very suitable. So that leaves the winged horse (unfortunately not available in WinBoard) for the Elvish Queen. But in mythology Pegasus is a winged horse, so having another piece named Pegasus rules this out. And the winged-horse glyph cannot be used for that other piece either, because it would subvert the idea that knightish glyphs should alert people to the fact the piece can move as a Knight. So the original naming really interfered badly with the possibility to choose suitable images. This is why I suggested the name change.

I am not very familiar with the mythology of the Phoenix, but I don't associate it with a very powerful creature, like a dragon). I imagine it as a pieceful, innocent creature, no match for an eagle. I doubt anyway whether the logic of assiging similar functions (such as archers, spell-casters) to corresponding units of the different armies will be very helpful to prospective players (or would even be noticed by them). It just makes it difficult to find names that they would easily remember.

BTW, I started looking into the WinBoard e.p.-capture problem, and it is pretty tough. One unforseen problem is that assignment of e.p. capture rights reflects on detection of 3-fold repetitions. Presence o frights would not make it a repetition, but then the rights should only be granted when there actually is a Pawn that can make the e.p. capture. This means I cannot assign e.p. rights for every double push. But whether there is a Pawn that can make the capture depends on how enemy Pawns move. I guess we don't care much whether WinBoard will fail to detect some repetitions in variants with non-standard Pawns, but I should be careful not to break anything for orthodox Chess.

To H. G. Muller

If we prefer the Pegasus as Elvish Queen then it will be clear inconsistency with the Angel and the Dragon and they also will be removed (but they are very strong fantasy figures). But the Phoenix by many sources is equal to that figures or sometimes even higher.

I think that we can save all the old names because this is not important for 99% players. For example, there is possible to be embarrassed that an elephant/bishop playfully runs along the diagonals through all the board, but because of this, very few people stopped playing classical chess.

About images, my opinion:

Dragon = Chancellor = Knight-Rook chimera (standard, don't change).
Phoenix = Archbishop = Pair of the swords (modern standard, don't change).
Hunter, Centaur = Bishop-types.
Pegasus, Wyvern = Rook-types.
Unicorn, Werewolf, Fairy, Guard = good like now, don't change.

About e.p. capturing, I think that there are 2 main criteria:
1) orthodox pawn could be correct
2) diagonal pawn could be correct (the current problem)
Everything else is unimportant because is extremally rare (and you may slow the engine for nonexistent pawns).

I think I have a satisfactory solution implemented in WinBoard, now. I improved the standard move generation and execution for Pawns to always remember the square skipped by a Pawn or Lance that moves more than 1 step forward, as well as where that Pawn ended up. And I now always allow a diagonal Pawn move to the same file as such a square, and then always remove the Pawn moved on the preceding move. Pawns (or Lances) with redefined moves, however, will consider every move to the skipped square an attempt to e.p. capture, also if that move in the move definition did not have e.p. capture rights (but some other move did), or indeed any capture rights at all.

This fails for Berolina Pawns, which can move through their diagonal non-capture to the square skipped by an anti-diagonal double-push. But there would hardly ever be any need to define Pawns as Berolina Pawns, as Berolina Chess is a standard variant, which has these Pawns and their e.p. habits hard-coded, and other variants could inherit that by defining 'berolina' as parent variant.

The Elvish and Orkish Pawns  do not have this problem:  all their moves can capture, and thus presumably also e.p. capture when they go to a square skipped by a double-push. Only FIDE Pawns have a problem against Elvish Pawns, in that their straight-forward non-capture can reach a skipped square. But this can then be solved by not redefining the Human Pawn, so that the built-in default move generation is used for that. And this does know that only diagonal moves can e.p. capture.

The only problem is that e.p. rights will not always be properly taken into account when comparing positions for the purpose of determining 3-fold-repetition draws. WinBoard might mistakenly think that no e.p. rights exists after a double-push because there are no Pawns next to the double-pushed Pawn that could e.p. capture it. While in fact the Pawns that can e.p. capture it are further away (e.g. after e2-e4 an Orkish Pawn on d3, or after e2-g4 an Elvish Pawn on e4). I suppose we can live with that; Fairy-Max assumes any repetition of a previous position will be a draw anyway.

The fixed WinBoard executable can be dowloaded from http://hgm.nubati.net/winboard.zip .

The only problem is that e.p. rights will not always be properly taken into account when comparing positions for the purpose of determining 3-fold-repetition draws.

It's not important for the most players, if this is no online-tournament engine.

Thank you for supporting new chess variants. Are you interested in programming the online-engine of lichess for this variant? I don't know how difficult is it.

I think the on-line engine used at Lichess is Stockfish (or more accurately, a Stockfish derivative). And I am not interested in working on Stockfish. In the first place it is written in a programming language I do not master (C++). But I also think it would be pretty hard to convert it to support more piece types, and in particular piece types used in Asymmetric Chess, such as the Wolf and Unicorn.

To H. G. Muller:

http://hgm.nubati.net/variants/orc-elf/

Can you put all 9 match-ups for this variant into your online service?

There is all match-ups for Fairy-Max:

// Asymmetric Chess (Orc-Orc)
Game: fairy/Orc-Orc # P..R........NB..Q....Kp..r........nb..q....k
8x8
6 4 5 7 3 5 4 6
8 10 11 9 3 11 10 8
p:125 -16,24 -16,7 -1,5 1,5
p:125 16,24 16,7 -1,5 1,5
k:-1 1,34 -1,34 1,7 16,7 15,7 17,7 -1,7 -16,7 -15,7 -17,7
n:420 1,043,-15 1,070,17 16,043,17 16,070,15 -1,043,15 -1,070,-17 -16,043,-17 -16,070,-15
b:275 17,7 15,7 -17,7 -15,7 34,7 30,7 -34,7 -30,7
R:450 2,3,1 32,3,16 -2,3,-1 -32,3,-16
Q:840 1,3 16,3 -1,3 -16,3 14,7 31,7 33,7 18,7 -14,7 -31,7 -33,7 -18,7
R:450 2,3,1 32,3,16 -2,3,-1 -32,3,-16
Q:840 1,3 16,3 -1,3 -16,3 14,7 31,7 33,7 18,7 -14,7 -31,7 -33,7 -18,7
n:420 1,043,-15 1,070,17 16,043,17 16,070,15 -1,043,15 -1,070,-17 -16,043,-17 -16,070,-15
b:275 17,7 15,7 -17,7 -15,7 34,7 30,7 -34,7 -30,7
# P fWscWifmnD
# p fWscWifmnD
# N WafsW
# n WafsW
# B FA
# b FA
# R yafWgafW
# r yafWgafW
# Q RN
# q RN

// Asymmetric Chess (Human-Human)
Game: fairy/Human-Human # PNBRQ.........Kpnbrq.........k
8x8
6 4 5 7 3 5 4 6
8 10 11 9 3 11 10 8
p:100 -16,24 -16,6 -15,5 -17,5
p:100 16,24 16,6 15,5 17,5
k:-1 1,34 -1,34 1,7 16,7 15,7 17,7 -1,7 -16,7 -15,7 -17,7
n:325 14,7 31,7 33,7 18,7 -14,7 -31,7 -33,7 -18,7
b:360 15,3 17,3 -15,3 -17,3
R:500 16,3 -16,3 -1,3 1,3
Q:960 1,3 16,3 15,3 17,3 -1,3 -16,3 -15,3 -17,3
R:500 16,3 -16,3 -1,3 1,3
Q:960 1,3 16,3 15,3 17,3 -1,3 -16,3 -15,3 -17,3
n:325 14,7 31,7 33,7 18,7 -14,7 -31,7 -33,7 -18,7
b:360 15,3 17,3 -15,3 -17,3
# P fmWfcFifmnD
# p fmWfcFifmnD
# N N
# n N
# B B
# b B
# R R
# r R
# Q RB
# q RB

// Asymmetric Chess (Elf-Elf)
Game: fairy/Elf-Elf # ..B....Q......R...P.NK..b....q......r...p.nk
8x8
6 5 4 7 3 4 5 6
8 11 10 9 3 10 11 8
p:120 -15,24 -17,24 -15,7 -17,7
p:120 15,24 17,24 15,7 17,7
k:-1 1,34 -1,34 1,7 16,7 15,7 17,7 -1,7 -16,7 -15,7 -17,7
n:425 17,043,1 17,070,16 15,043,16 15,070,-1 -17,043,-1 -17,070,-16 -15,043,-16 -15,070,1
b:280 34,3,17 30,3,15 -34,3,-17 -30,3,-15
R:475 1,7 16,7 -1,7 -16,7 2,7 32,7 -2,7 -32,7 3,7 48,7 -3,7 -48,7
Q:810 15,3 17,3 -15,3 -17,3 14,7 31,7 33,7 18,7 -14,7 -31,7 -33,7 -18,7
R:475 1,7 16,7 -1,7 -16,7 2,7 32,7 -2,7 -32,7 3,7 48,7 -3,7 -48,7
Q:810 15,3 17,3 -15,3 -17,3 14,7 31,7 33,7 18,7 -14,7 -31,7 -33,7 -18,7
n:425 17,043,1 17,070,16 15,043,16 15,070,-1 -17,043,-1 -17,070,-16 -15,043,-16 -15,070,1
b:280 34,3,17 30,3,15 -34,3,-17 -30,3,-15
# P fFifmnA
# p fFifmnA
# N FafsF
# n FafsF
# B yafFgafF
# b yafFgafF
# R WDH
# r WDH
# Q BN
# q BN

// Asymmetric Chess (Elf-Human)
Game: fairy/Elf-Human # ..B....Q......R...P.NKpnbrq................k
8x8
6 5 4 7 3 4 5 6
8 10 11 9 3 11 10 8
p:120 -15,24 -17,24 -15,7 -17,7
p:100 16,24 16,6 15,5 17,5
k:-1 1,34 -1,34 1,7 16,7 15,7 17,7 -1,7 -16,7 -15,7 -17,7
n:425 17,043,1 17,070,16 15,043,16 15,070,-1 -17,043,-1 -17,070,-16 -15,043,-16 -15,070,1
b:280 34,3,17 30,3,15 -34,3,-17 -30,3,-15
R:475 1,7 16,7 -1,7 -16,7 2,7 32,7 -2,7 -32,7 3,7 48,7 -3,7 -48,7
Q:810 15,3 17,3 -15,3 -17,3 14,7 31,7 33,7 18,7 -14,7 -31,7 -33,7 -18,7
R:500 16,3 -16,3 -1,3 1,3
Q:960 1,3 16,3 15,3 17,3 -1,3 -16,3 -15,3 -17,3
n:325 14,7 31,7 33,7 18,7 -14,7 -31,7 -33,7 -18,7
b:360 15,3 17,3 -15,3 -17,3
# P fFifmnA
# p fmWfcFifmnD
# N FafsF
# n N
# B yafFgafF
# b B
# R WDH
# r R
# Q BN
# q RB

// Asymmetric Chess (Human-Elf)
Game: fairy/Human-Elf # PNBRQ................K..b....q......r...p.nk
8x8
6 4 5 7 3 5 4 6
8 11 10 9 3 10 11 8
p:100 -16,24 -16,6 -15,5 -17,5
p:120 15,24 17,24 15,7 17,7
k:-1 1,34 -1,34 1,7 16,7 15,7 17,7 -1,7 -16,7 -15,7 -17,7
n:325 14,7 31,7 33,7 18,7 -14,7 -31,7 -33,7 -18,7
b:360 15,3 17,3 -15,3 -17,3
R:500 16,3 -16,3 -1,3 1,3
Q:960 1,3 16,3 15,3 17,3 -1,3 -16,3 -15,3 -17,3
R:475 1,7 16,7 -1,7 -16,7 2,7 32,7 -2,7 -32,7 3,7 48,7 -3,7 -48,7
Q:810 15,3 17,3 -15,3 -17,3 14,7 31,7 33,7 18,7 -14,7 -31,7 -33,7 -18,7
n:425 17,043,1 17,070,16 15,043,16 15,070,-1 -17,043,-1 -17,070,-16 -15,043,-16 -15,070,1
b:280 34,3,17 30,3,15 -34,3,-17 -30,3,-15
# P fmWfcFifmnD
# p fFifmnA
# N N
# n FafsF
# B B
# b yafFgafF
# R R
# r WDH
# Q RB
# q BN

// Asymmetric Chess (Human-Orc)
Game: fairy/Human-Orc # PNBRQ................Kp..r........nb..q....k
8x8
6 4 5 7 3 5 4 6
8 10 11 9 3 11 10 8
p:100 -16,24 -16,6 -15,5 -17,5
p:125 16,24 16,7 -1,5 1,5
k:-1 1,34 -1,34 1,7 16,7 15,7 17,7 -1,7 -16,7 -15,7 -17,7
n:325 14,7 31,7 33,7 18,7 -14,7 -31,7 -33,7 -18,7
b:360 15,3 17,3 -15,3 -17,3
R:500 16,3 -16,3 -1,3 1,3
Q:960 1,3 16,3 15,3 17,3 -1,3 -16,3 -15,3 -17,3
R:450 2,3,1 32,3,16 -2,3,-1 -32,3,-16
Q:840 1,3 16,3 -1,3 -16,3 14,7 31,7 33,7 18,7 -14,7 -31,7 -33,7 -18,7
n:420 1,043,-15 1,070,17 16,043,17 16,070,15 -1,043,15 -1,070,-17 -16,043,-17 -16,070,-15
b:275 17,7 15,7 -17,7 -15,7 34,7 30,7 -34,7 -30,7
# P fmWfcFifmnD
# p fWscWifmnD
# N N
# n WafsW
# B B
# b FA
# R R
# r yafWgafW
# Q RB
# q RN

// Asymmetric Chess (Orc-Human)
Game: fairy/Orc-Human # P..R........NB..Q....Kpnbrq................k
8x8
6 4 5 7 3 5 4 6
8 10 11 9 3 11 10 8
p:125 -16,24 -16,7 -1,5 1,5
p:100 16,24 16,6 15,5 17,5
k:-1 1,34 -1,34 1,7 16,7 15,7 17,7 -1,7 -16,7 -15,7 -17,7
n:420 1,043,-15 1,070,17 16,043,17 16,070,15 -1,043,15 -1,070,-17 -16,043,-17 -16,070,-15
b:275 17,7 15,7 -17,7 -15,7 34,7 30,7 -34,7 -30,7
R:450 2,3,1 32,3,16 -2,3,-1 -32,3,-16
Q:840 1,3 16,3 -1,3 -16,3 14,7 31,7 33,7 18,7 -14,7 -31,7 -33,7 -18,7
R:500 16,3 -16,3 -1,3 1,3
Q:960 1,3 16,3 15,3 17,3 -1,3 -16,3 -15,3 -17,3
n:325 14,7 31,7 33,7 18,7 -14,7 -31,7 -33,7 -18,7
b:360 15,3 17,3 -15,3 -17,3
# P fWscWifmnD
# p fmWfcFifmnD
# N WafsW
# n N
# B FA
# b B
# R yafWgafW
# r R
# Q RN
# q RB

// Asymmetric Chess (Elf-Orc)
Game: fairy/Elf-Orc # ..B....Q......R...P.NKp..r........nb..q....k
8x8
6 5 4 7 3 4 5 6
8 10 11 9 3 11 10 8
p:120 -15,24 -17,24 -15,7 -17,7
p:125 16,24 16,7 -1,5 1,5
k:-1 1,34 -1,34 1,7 16,7 15,7 17,7 -1,7 -16,7 -15,7 -17,7
n:425 17,043,1 17,070,16 15,043,16 15,070,-1 -17,043,-1 -17,070,-16 -15,043,-16 -15,070,1
b:280 34,3,17 30,3,15 -34,3,-17 -30,3,-15
R:475 1,7 16,7 -1,7 -16,7 2,7 32,7 -2,7 -32,7 3,7 48,7 -3,7 -48,7
Q:810 15,3 17,3 -15,3 -17,3 14,7 31,7 33,7 18,7 -14,7 -31,7 -33,7 -18,7
R:450 2,3,1 32,3,16 -2,3,-1 -32,3,-16
Q:840 1,3 16,3 -1,3 -16,3 14,7 31,7 33,7 18,7 -14,7 -31,7 -33,7 -18,7
n:420 1,043,-15 1,070,17 16,043,17 16,070,15 -1,043,15 -1,070,-17 -16,043,-17 -16,070,-15
b:275 17,7 15,7 -17,7 -15,7 34,7 30,7 -34,7 -30,7
# P fFifmnA
# p fWscWifmnD
# N FafsF
# n WafsW
# B yafFgafF
# b FA
# R WDH
# r yafWgafW
# Q BN
# q RN

// Asymmetric Chess (Orc-Elf)
Game: fairy/Orc-Elf # P..R........NB..Q....K..b....q......r...p.nk
8x8
6 4 5 7 3 5 4 6
8 11 10 9 3 10 11 8
p:125 -16,24 -16,7 -1,5 1,5
p:120 15,24 17,24 15,7 17,7
k:-1 1,34 -1,34 1,7 16,7 15,7 17,7 -1,7 -16,7 -15,7 -17,7
n:420 1,043,-15 1,070,17 16,043,17 16,070,15 -1,043,15 -1,070,-17 -16,043,-17 -16,070,-15
b:275 17,7 15,7 -17,7 -15,7 34,7 30,7 -34,7 -30,7
R:450 2,3,1 32,3,16 -2,3,-1 -32,3,-16
Q:840 1,3 16,3 -1,3 -16,3 14,7 31,7 33,7 18,7 -14,7 -31,7 -33,7 -18,7
R:475 1,7 16,7 -1,7 -16,7 2,7 32,7 -2,7 -32,7 3,7 48,7 -3,7 -48,7
Q:810 15,3 17,3 -15,3 -17,3 14,7 31,7 33,7 18,7 -14,7 -31,7 -33,7 -18,7
n:425 17,043,1 17,070,16 15,043,16 15,070,-1 -17,043,-1 -17,070,-16 -15,043,-16 -15,070,1
b:280 34,3,17 30,3,15 -34,3,-17 -30,3,-15
# P fWscWifmnD
# p fFifmnA
# N WafsW
# n FafsF
# B FA
# b yafFgafF
# R yafWgafW
# r WDH
# Q RN
# q BN
 

OK, done. Except for human-human, which seems just orthodox Chess.

As for the Fairy-Max game definitions: it is better to leave out the Betza move definition of the Human pawn (the line strating with "# P" or "# p") where the Humans play the Elves. These lines are only for sending to WinBoard to reconfigure its move generator, and it is better to rely on the built-in generator for FIDE Pawns, because this makes proper destinction between capture and non-capture moves when deciding if it should e.p. capture. (Against Human or Orc Pawns this does not matter, because the FIDE Pawn's only non-capture move could never reach the skipped square. But against Elvish Pawns it can.)

About Fairy-Max:

I have tested 200 games for each non-mirror match-up (100 white + 100 black) and have got this statistics:
Elf < Human 46,25%
Elf < Orc 40,25%
Human < Orc 38,25%
It seems to be: Orc > Human > Elf, but the game analysis indicates that Elf > Human > Orc at the openings. And it is logically, because Elves have more active Pawns and the Orcs have less. That's why the Elves and Humans must be initiative but they don't, prefer to passive openings. They don't get any advantages in space, and Orcs, after the opening without any problems, punish them for their passivity. The Orcs have equal statistics playing white and black, it only confirms their weakness in the openings (potential but not real in auto games).

And this is a problem (and the feature) of asymmetry, different sides mean different opening's speed. If we will skip openings playing passive style and our side have an advantage at openings, then we will lose this advantage.

Without strong openings I can't get an answer to the question: are the Orcs imba, or not?

I have tested games at the time control of 1 sec per turn because each game need to be manually shown and I have no time to watch a lot of games in a long time. The WinBoard have a serious bug to incorrectly counting the wins and loses in the match of the same engines. It may show 20-20-10, but really there was 30-10-10 (if manually watching the games).

For example of opening's advantages:
- Elves have active Bishops from the very beginning, such moves are available and not bad (vs Human): 1.e4 - Bf4!
- Elves have active Rooks from the very beginning, move like 1.Ra4 is available (vs Human), the Orcish Wyverns may also leaping there, but only the Pegasus' move is good (because creates a real threats of Rxa7 or Ra5 - Rxa8)
- Elvish Knights (Unicorns) are centralized and can quickly enter the game through the squares of e2 and d2 (although not quickly as Human Knights), but Orcish Knights need to free squares f1 and c1 first (need to starts with their Bishops).
- Elves don't need to activate their Bishops, they are already active! Elves even don't need to move them for castling, because even simple Kf1 and then, if necessary, Re1 (leaping over the King and the Bishop) do the same, but faster.
- Orcs have problems with their short Bishops, because Be3 or Bd3 may block central Pawns (much more important than f и с after Human Knights moves Nf3 or Nc3).
- Elvish Pawns have many variants to enter the game, but Orcish Pawns lose all their defensive bonus when moving. Elves can checkmate Orcs at the first move: 1.g3?? Bxg3#

What do you mean by "game analysis indicates"? The result of the game is an objective measure, but any reported score merely reflects what you put in, in terms of piece values.

Note that WinBoard reports the score per player, and that in a match the players alternate color. So you cannot associate the reported number of wins with a particular color. There are two ways to get what you want: one is make sure the games of a particular test are automatically saved in a separate file, not containing anything else, and then opening that game file with WinBoard (Load Game). In the header of the game list WinBoard then reports the statistics of the game sets, wins, losses and draws for white. (As in general multi-game files can have many players.) The second method would be to start WinBoard with the 'Additional option' -sameColorGames 100 (or any other number of games you desire). This suppresses the color alteration, so that you can associate the scoreof the first player with white. (It also overrules the games per pairing).

For example how quickly elves may evolve an initiative, the game vs Humans:

1.ec4 d5 (with favorable exchange: Footman is weaker than Fairy)
2.cd Qxd5
3.ce4 Qg5 (blocking Ne2 due to Qxg2)
4.Be3 Qg6
5.Qf3 Nd7
6.d5 Qd6
7.Ra4 b6
8.Be4 Rb8
9.Bc6+ Kd8
10.Rxa7 Ngf6
11.Ra8+ Rxa8
12.Bxa8 Bb7
13.c6 Bxa8
14.b7 Bb7
15.Qxb7+ Ke8
16.Qxd6 ed
17.Ne2 with a significant advantage

Or another variant:
10....Bb7
11.Rxb7 Rxb7
12.Ba4 Nc5
13.c6 Ra7
14.Bxc5 Qxc5
15.Ne2 Rxa4
16.b7 Nf6
17.Nd4 Qxc1+
18.Ke2 Qc4+
19.Kd1 Qd5
20.Nc6+ Ke8
21.Qxd5 Nxd5
22.c8Q Nf6
23.Nb5.

I don't check all variants, its a simple illustration of elvish style how it can be at the openings. You may see these variants:
first demo
second demo

H. G. Muller

What do you mean by "game analysis indicates"? The result of the game is an objective measure, but any reported score merely reflects what you put in, in terms of piece values.

An objective measure if the openings are right. Game analysis means that value indicated by game engine if switch to analysis mode. Maybe it's a problem of fast timing control but at the 10 minutes to 40 turns engine plays closed and passive openings too.

All pieces values summary are equals, scores differ because of the position advantage, activity or abilities to get material advantage, which are maximum for elves and minimum for orcs at the openings.

As I had read, the first 3 turns are "random" but this random can't be objective because players play openings to get an advantage and to use the strong abilities of their pieces. Openings like 1.c3, 2.d3 are not that openings which is actual.

There will be not important if all 3 races have equal opening's abilities, but they haven't.

Originally, I had think that such passive style means a bad design of new units, but when I had launch Human-Human mirror (or Orthodox chess), I saw the same style.

About saving results into the file - thanks, I will try it.

There is an example of auto game Human-Human at 10 minutes:

1.Nf3 d5 2.c3 c6 3.d3 Be6 4.Nd4 Bd7 5.Bf4 Qb6 6.Qb3 f6 7.Nf3 Na6.

Can white realize their tempo advantage playing this style? No. And elves and humans can't also (versus orcs).

I think that elves have advantage at openings, orcs - at midgames, and humans - at endgames. Both elves and humans can't get enough advantage from openings, but orcs get full advantage at midgame, because the engine is much stronger at midgame.

Maybe, adding openings base will fix this situtation, but it's not objective. I think the program shows that the maximum advantage of the Orcs is about 0.5 pawn (if all openings are passive), and only multiplayer statistics can get more detailed estimatings, including full value of the openings.

You should not take the scores reported by Fairy-Max that seriously. For one, the positional part of its score is relative to the current position. Its own position can be far better, and it can still report a negative score (with equal material). Because the opponent can improve his position a bit more (or faster) than the side to move. This is for instance clearly to see directly after it castles, when the opponent has not castled yet. Castling gets a large bous (or it would never do it), and of course it is usually not possible to prevent the opponent from doing it. So even though this only equalizes, and the opponent could have other disadvantages too (like more-poorly developed pieces), Fairy-Max will still report a negative score, just because it can never match the castling bonus that it already collected.

Maybe, it will be better, if the Fairy-Max will have not static, but dynamic evaluating of pieces, according the same criterias, that I use for balancing new units (you may read it at the end of main post). It is very experimental way, but if all weights are correct, an evaluation of positions will no depends on user's subjective values and mistakes. The basic idea is that the 60-80% of the piece's value is determined by its average attack/speed and can be pre-calculated at the beginning of game, but then add a positional part of values to each piece depending on it's current position, attacks and moves limiting by obstacles, activity (value of attacking squares), acceleration, agility and potential. There is no fact that it will work as universal, but if there is successful, it will be great.

Now, if any chess variant (with different armies) is imbalanced, there is no assurance that it is objective not only because of poor openings but because the engine makes decisions depending on evaluations of positions, which depend on the user's subjective values of his own pieces.If the user is wrong by his values, then the engine is wrong by its evaluations. And I believe that adding right positional criterias make the openings much better without dependings on the openings book. Because any passive openings like c3 or d3 will impair the positional power of own pieces, blocking their current moves and abilities. For example, move c3 kills orthodox Knight's potential to attacking squares d5 and a4, limits its speed and agility (moves x threats) to attacking b5 and e4 and commonly is the very bad move, except cases that square of d4 is under opponent's attack and it is important to defend this square or to prepare active move of d4 (not d3). There is no need to calculate any deep variants to understand for the engine that this move is terrible.

Maybe, it will be better to create position's evaluating special config, in which user can define values of many position criterias (weights) if he want to correct the engine's playing style.

And the next idea is to improve endgames by learning the engine any type positions. For example, user edit the position, place some pieces and then launch the analysis and the program automately generate "Nalimov's tables" for the endgames with these peaces. I find that now the engine have several problems to checkmating a bare king with the Wyvern or even with the several Queen-type units (sometimes make stalemating instead checkmating), although this is a problem only if there is very fast timing control. But as for me, it's very interesting to explore the endgames of Dragon (Chancellor) versus Griffin (Rook) and many others with new units, because these endgames haven't any existing Nalimov's tables yet.

It doesn't work (5.0b3).

Update: I add the string /sameColorGames 100 to the end of file winboard.ini and it seems to work correctly, thanks.

You left out the leading hyphen when you typed the option in the startup dialog. All option names (as opposed to the option values) should start with a '-' or '/'. (These are treated as equivalent; one is Linux style, the other is Windows style. The '=' between an option name and value is also optional, and can be a space instead.) So in the ini file it worked because you did write the '/'.

I had saw that pawns' moves puts to hash as "c3" but with elvish diagonal pawns there are some different pawns able to move here.

Update: sorry, it was my mistake - I use the same file for different match-ups, they were not elves in that game. As I understand, I need to use different hash-file for each match-up.

H. G. Muller:
I'm interested in your statistical method of evaluating pieces with Fairy-Max. But I don't know details and need to some advices.

For example, I want to evaluate Pegasus, then change 2 Rooks by 2 Pegasus and launch long series. If I get a result of winrate of 60% (for example), how to convert it to centipawns? And how long match need to played for accuracy of 0.25 pawn? 0.1 pawn? 0.05 pawn? Is it works if I set a fast timing limits of 1 sec per turn, or maybe better to use another timing limits for such tests?

How is better to test pawns if they have different options to promote (different versions of queens)? First test Queens, get values, exchange them and then test pawns?

Do you have results for famous peaces, like Archbishop, Chancellor and Elephant (jB2) on the 8x8 board? What about base (orthodox) pieces: Knight, Bishop, Rook, Queen, King (as Man)?

"For example, I want to evaluate Pegasus, then change 2 Rooks by 2 Pegasus and launch long series. If I get a result of winrate of 60% (for example), how to convert it to centipawns?"

You can then delete the f-Pawn of the side that scored 60%, and repeat the test. If it now scores 45%, you know the Pawn was worth 15%, so the 60% translates to an advantage of 2/3 Pawn, which would be 33cP for a sigle Rook-Pegasus difference. If you do this for many piece comparisons,and each time deleting a Pawn makes the score drop ~15%, you can expect that to be always the case, and skip the Pawn-odds test if the score was close to 50%.

"And how long match need to played for accuracy of 0.25 pawn? 0.1 pawn? 0.05 pawn?"

The standard deviation of the score percentage in a match of N games is approximately 40%/sqrt(N). For N=100 that would be 4%. If a Pawn is indeed worth ~16%, that would be 25cP. But in 32% of the cases the error would be larger than that. To be within error bars 95% of the time you should count on an error of 2 standard deviations. To get that to 25cP you would need 4 times as many game, i.e. 400. To get the standard deviation to 0.1 Pawn (2.5x smaller) you would need 6.25 times as many games, i.e. 2500. For 5cp it would be 10,000. But I doubt that piece values can be defined with this precision anyway; the whole idea that pieces have aconstant value is only an approximation, and in practice the value could differ dependent on what the opponent has, or where they start in the opening. (As is most dramatically demonstrated by the fact that in 'Charge of the Light Brigade' 7 Knights crush 3 Queens.)

"Is it works if I set a fast timing limits of 1 sec per turn, or maybe better to use another timing limits for such tests?"

It is better to use time cotrolslike 40 or 60 moves per minute, than a fixed time per move, so that the engine can allocate the time where its needs it, and is not forced to abort a depth iteration that took unexpectedly long, and let the effort goto waste. But 40 moves/min should be fine. I never tried faster games.

More later.

I think that I need 10k games per unit.

My aim is the balance not far than 55%, such as 52 +/- 2% in each match-up. There are 5 different unit's types: queens, knights, rooks, bishops and pawns, so for each type accuracy must be +/- 0,8%. And it equals 0.05 pawn per type (if pawn = 16%). "Per type" means the full complect (for example, 2 bishops or 8 pawns).

Have you tested this values for orthodox pieces? This is no important for balancing (because of relative difference), but good to publicating absolute values. And I know that this is only statistics, real values are dynamical and positional.

I think that if the Orcs will be the imba, the best way to balance them is limiting promotion of their Guards, for example don't promote to a Dragon. But the next 2 candidates, a Wyvern and a Werewolf are very closed, so Fairy-Max "promoting to only Queen" (where the "Queen" is one of that pieces #7 or #9) will not be as good as for natural Queen.

Hello Dmitry,

As a fellow inventor I'd like to congratulate you for taking the daunting endevour of creating a different armies game. I'm, personally on the fence for that as balance is hard to obtain. But in the end "Nothing worth doing is easy" (I don't remember who said that first).

" My aim is the balance not far than 55%, such as 52 +/- 2% in each match-up. There are 5 different unit's types: queens, knights, rooks, bishops and pawns, so for each type accuracy must be +/- 0,8%. And it equals 0.05 pawn per type (if pawn = 16%). "Per type" means the full complect (for example, 2 bishops or 8 pawns). "

That is not an efficient approach. Whether the armies are balanced follows from having the complete armies play against each other. That is only a single measurement, which would need a much lower accuracy (2% ~ 1/8 Pawn) than each of the individual piece values would need to make the sum accurate enough. Adding piece values is an approximation anyway. Some pieces cooperate better than others, some pieces combat some opponent pieces better than others. You would never see that when you measure the pieces in isolation. Of course you need to approximately know the piece values to make realistic measurements, though. (E.g. whether a player should seek a certain trade or avoid it could influence the valuesof the involved pieces.)