Help! Why don't fairy chess pieces have values?

I just did wrong.

There’s a website with the best explanation of fairy chess piece values I’ve ever seen at

nobody can agree on the values

Plz show me

The chancellor is regarded at minimum 8 + smidgen.

The archbishop is an intermediate piece between the rook and the queen, slightly closer to the queens value. Estimated at 7.75 face value while other experts alludes even 8 blank.

Estimating piece values is bound to give totally wrong results. The only reliable way is to measure the performance in games. On 8x8 (or 10x8) Chancellor and Archbishop are both very close in value to a Queen, and even closer to each other. On a scale where Q=9.5, we would have C=9, A=8.75. When you rapce Q by A in the FIDE start position, and give the Q player Pawn odds (so an imbalance of Q vs A+P), the Archbishop wins the game more often than not.

I don't doubt your materialistic findings, but it may be hard for many to accept that an A is about a Q in basic value.

Well, it is also hard for some to accept that the Earth isn't flat. That doesn't alter the fact.

In general people will suffer by acting on their misconceptions. In this case that is pretty harmless; they will just lose games and be considered weak players. Much better than refusing to accept covid is a life-treatening disease...

It’s at TricesChess dot com on the left menu is an item with complete maths

But that's all plain nonsense, right? This whole idea that safe checking would be important is a bust. It predicts that a non-royal King, which cannot safely check at all, has no value at all. While it is obvious even without elaborate testing that it should have some. (In fact its value is similar to that of a Knight, and K+Q vs K + non-royal K is a general draw, showing its defensive value.)

A divergent piece, moving as Q but only capturing (and checking) as K is even worth as much as a Rook, despite having no safe checks. And a Knight that would not be able to capture a King (i.e. cannot deliver check), but can capture any other piece in the normal way, is hardly worth less than an ordinary Knight.

The value for the Archbishop given on that website are totally off (about 2 Pawns too low).

It is far more important how effective a piece is with regards to Pawns than how it performs against King. As they say, the Pawn is the soul of Chess.

Taylor published his paper in 1876 using the same idea, and it was championed as a great paper. Trice went one giant leap further. He generalized the board from square to rectangular, and extended the computation to any type of piece, not just 8x8 chess pieces. His paper was peer reviewed and published. And he crushed variant programs like yours and ChessV and those games are shown online. You’ve got zero credibility.

According to the 4 player chess/variants server, fairy pieces do have values (which you can see when playing a "Game of Points" match with fairy pieces). I believe that some of these values are inaccurate, but even my system would be different from someone else's system. So I can't say for certain what the true values of pieces should be, but I can assure you that Chess.com at least has a solid points system in place.

Then again, depending on the position or variant at hand, some pieces are worth much more or much less. For example, in a labyrinth-like variant, a knight and a queen could be valued almost equally, while a rook be nearly useless. So it all depends

Black to move above. Who’s winning?

I worked as a research scientist all my professional life, and believe me, that you manage to get something published in a prestigeous peer-reviewed journal like Phys. Rev. Lett. doesn't mean it's not plain nonsense. Let alone the ICGA Journal, which is so desperate for copy they take almost anything.

What you seem to be missing is that scientific or mathematical issues (and Chess can be viewed as a branch of mathematics) are not decided by the 'credibility' of the person who points out a fact, but by verifiability or deducibility of the fact itself. So if you want to convince the readers that what I said is wrong, then show us why a typical position where one player has an extra non-royal King, but otherwise equal material (or better yet, 8 extra non-royal Kings) does not put his opponent at any disadvantage. (Which should be the case if the claim of that paper that non-royal Kings have zero value is true...) To make it even more lifely, we could play a game here where I take the odds of replacing my Queen by 8 non-royal Kings (placed on 6th rank, playing black). You think you can win that?

I am not sure what on-line games you talk about, and attach little value to games that were not played under supervised conditions. People can publish anything on line; it is very easy to have a program play itself until the result is not a draw, and then claim you played the winning side yourself. My program Spartacus, which won the 10x8 variant tourney a decade ago, is actually a private engine, and I don't recall I had it play Ed ever.

That being said, I should point out that 10x8 Chess programs are considered weak, even when the corresponding 8x8 engine is of GM level. The point seems to be that because of the wider board it is very important how you distribute your power (which resides mostly in the 3 super-pieces) over the board, and can exploit a poor distribution by blocking the center to 'freeze' it. When this gives you a super-piece majority on the Kings' side, it is often decisive before the opponent can redistribute his pieces. This is of course easy to repair by a dedicated evaluation term. But in orthodox Chess you don't really have this problem, so converted normal Chess engines would not contain it.

Tell me the value of the pieces as I let them drop during this game I won.

Even without looking at the game I can tell the piece values are the same as always, (say Q=950, C=900, A=875, R=500, B=350 (BB=750), N=300, P=80-100), since piece values are by definition not dependent on the actual position, but an average over all positions that plausibly can occur in a game.