Check out Grant Acedrex, our featured variant for April, 2024.


[ Help | Earliest Comments | Latest Comments ]
[ List All Subjects of Discussion | Create New Subject of Discussion ]
[ List Latest Comments Only For Pages | Games | Rated Pages | Rated Games | Subjects of Discussion ]

Comments/Ratings for a Single Item

LatestLater Reverse Order Earlier
Who is Who on Eight by Eight. A compilation of Zillions-estimated piece values on an 8x8 board.[All Comments] [Add Comment or Rating]
Andrew Hudson wrote on Fri, Oct 12, 2018 12:29 PM UTC:

I really love the work he's done making this list, but I have to ask: Why is the Cannon rated higher than the Leon? If I understand the Leon right then it's strictly better.

Also, from what I've heard, the Cannon belongs way lower down than just beneath Rook. Maybe Zillions has some glitch when it comes to screening pieces, or maybe it's all in the starting location implemented.


H. G. Muller wrote on Thu, Sep 6, 2018 02:26 PM UTC:

Interesting for sure, but to be taken with a rather large grain of salt. As the author points out, this is just how Zillions of Games values the pieces. Zillions is not always right, and not very strong. E.g. it has been shown by extensive computer self-play with far stronger programs that BN and RN are worth 8.75 and 9 (on a scale where Q = 9.5) on 8x8, which deviates a lot from the ratio 109 / 134 that Zillions get through some (propriatry) point-counting algorithm.

That being said, the Lion is indeed very strong, because of its ability to make hit-and-run captures. In Chu Shogi (12 x 12 board) it is considered about 1.66 times as valuable as a Queen, in the middle game. (In the end-game, where the board empties, the Lion becomes less valuable because it cannot be shielded well from slider attacks.) On 8x8 this ratio should go up significantly, as in Chu the Queen benefits from the larger board, while the Lion hardly does. An Amazon (QN) appears to be worth exactly as much as a separate Queen and Knight on 8x8, i.e. about 1.33 Queens.

Also note that the value of pieces can be strongly affected by what the opponent has. E.g. three Queens lose to 7 Knights (and only Pawns and King in addition) on 8x8, which violates all our ideas about the relative value of Q and N.


Anthony Viens wrote on Thu, Sep 6, 2018 06:23 AM UTC:Good ★★★★

I just stumbled on this page, and it's quite interesting.  (Very old, but still good.)  I am, however, very surprised to see the Lion listed as the most powerful piece.  If I needed a snap decition between Amazon/Lion I would have said Amazon on reflex....very interesting.


George Duke wrote on Fri, Feb 5, 2010 05:58 PM UTC:
This is a great 9-year-old list with faults. Derzhanski felt compelled, out of courtesy to C.V.Page, to use modified Betza notation. Nobody is going to coalesce around opaque Betza notation, which takes moments longer to interpret than the one or few regular sentences to describe the same phenomenon. Jeliss has 200 piece-types in ''All the King's Men,'' and Derzhanski deliberately happens to have 200, the two lists having less than 1/2 overlap. Truelove's summary of Pritchard has about 800 piece-types, but signals multiple uses when relevant in different CVs, meaning more like 2000 Truelove-Pritchard piece-types in context of working CVs. I happen to like Truelove's because of having become familiar with the 'ECV' CVs. As in problemists' journal Die Schwalbe, vibrant German CV community always seems to have been co-equal with the British 1900-1990, and presumably duplicative overlap may never have fully achieved 50%. I previously estimated Gilman's number of piece-types in Man&BeastsXX, 2007 to present, as 2000. I imagine Gilman may set on a course to incorporate many other, new recognizable as well as obscure piece-types from Gruber into Man&BeastsXX over ensuing years. Derzhanski's spacing makes it the easiest list to use, just for the names, so far as it goes. Yet Derzhanski's here is not without uninterpretables and not without many errors. Error, for example: Quang Trung SkiRook must mean Rook^Dabbabah not ''D^R,'' and either or both would be only to capture. And undeciperable: Scirocco ''Camel plus'' could mean more power to the piece or to the promotee than or other than plain Camel. The reviewer has to go to Scirocco anyway to find out what gives. Many of these markings are plain and simply Derzhanski's personal notes.

George Duke wrote on Thu, Feb 4, 2010 07:02 PM UTC:Good ★★★★
This is an interesting compilation all but forgotten that it is good Knappen revives. I have ignored Derzhanski's since the month it went up in favour of using Truelove's list, and the organizational article, like the 'Man&BeastsXX' and those recently by Carillo and Knappen's own 'Nachtmahr', is always worth more by far than the individual CV, or even fifty new-combination artworks.

David Paulowich wrote on Tue, Nov 7, 2006 02:44 PM UTC:Good ★★★★

Ivan Derzhanski has compiled an impressive list of chess variant pieces. I wish that he had included more detailed notes, for the benefit of newcomers to this site. Here are some brief comments.

Nightrider combination pieces are discussed in the 'Notes' section of my variant Unicorn Great Chess.

I believe the 'short queen' (moves one or two squares like a queen) has been around for decades, under a variety of names. Peter S. Hatch calls this piece a Seeress in the ELVES army of his Fantasy Grand Chess (1999 and 2000). Tony Mez calls it a Guard in his Combo Modern Day Chess (2006). Peter Aronson and Ben Good use an unusual piece with the same two-square move in their variant Golem Chess. (2002)


6 comments displayed

LatestLater Reverse Order Earlier

Permalink to the exact comments currently displayed.