Ratings & Comments

I did once measure the value of the Kneen and the Quight. This confirmed what I had also observed in divegring short-range leapers: roughly 2/3 of the value of a piece is determined by its captures, 1/3 by its non-captures. The value of the Quight was thus about 5, and that of the Kneen about 7 Pawns, in a FIDE context.

I did observe an interesting irregularity, though: testing divergent compounds of Knight and Commoner, which themselves test as nearly equal in value, the combination mNcK (Kning?) proved to be about half a Pawn stronger than the other combinations. I suspect that there are some penalties and bonuses on the move pattern as a whole, which happen to cancel each other, so that its value is in the end not much different from that of other 8-target leapers. The K move set is very poor on 'speed', but very good on 'concentration' (i.e. lots of orthogonal contacts between the target squares). Speed seems to be mainly important on non-captures, determining how many turns you will need to reach the location where you are needed (such as a promotion square). Concentration seems to be mainly important for captures, determining whether you can attack Pawns in such a way that they cannot simply move away, and whether the piece has mating potential. The Kning thus has the best of both worlds: the Knight move provides the speed, and the King captures the concentration.

I think you can draw Allfarie pieces, with something like mspaint. Once drawn I think you need to work with the admins at this site to get the graphics installed into a particular web directory. Fergus has written a How-to on this. (I think). Your icons above look like Knight, Camel, Zebra, rook, bishop, ???, Gryphon. They are well done, but have a look/feel that is not as satisfying as an Allfarie piece. What got me interested in chess as a child was the pieces them selves, and the wood carvings. For your game, I love the introduction of the joker. It is a very interesting piece. If your game is showing legal moves I personally feel the arrow design is not needed. If you are making a commercial board game that is not a computer program, I admit your arrow piece set would be very useful.

This is just my opinion, and your piece set is well designed.

Commoner, Knight and Kight all tested as equally valuable, from an opening position. (E.g. when I replaced the Knights of one player by Commoners it did not affect the win rates.)

They're harder to quickly tell apart, and they can be hard to read. For example, I thought the third piece in your lineup was a Knight until I noticed the first piece, which is more clearly a Knight in a comparison between the two.

I thought so, too, but I cannot add a grid as there is not enought room on the 50x50 picture. Also it seems to me difficult to imagine pictoreals. For this game a rook+knightrider was already drawn by Greg. It has to be here somewhere. This will be called varan. Then I need a gnu, this is also on the website. And for now I'm using cobra for a piece that moves and captures like a rook and ferz and moreover captures like a cannon. Someone called it a mortar which sonds good enough, although I prefer something more mithical. These should go with both gamecode and interactive diagram, so I'll have to do some drawing, and it is easier to do these simbolic pieces.

Also I do not know how to make the difference between white and black pieces.

If all your pieces have symmetrical moves, you could just include one direction in the image. Then you wouldn't need so many squares on a grid.

I think this way things will look more ugly. The bishop for example will have the arrows finished at half board!

I think movement-diagram-inspired piece symbols can work, when you don't try to reproduce the movement diagram too literally. After all, the purpose of a piece representation in a game position is not to teach a completely unintiated player how the piece moves. Perhaps only to give him a hint for that. But he can be expected to know how these hints are encoded in the shape. The most important is that the shapes are strikingly different. Then the palyer will identify them at first glance, just like he would be able to distinguis the picture of a camel and a horse in a pictorial representation, or the various kanji in an oriental representation. In the end these are all just shapes that have to be distinguished from each other; camel or horse heads mean just as little to a person who has never seen an animal in his life as kanji do to a westerner.

Your Knight, Camel and Zebra representations fail the criterion of easy distinction. They are 'topologically similar', and the distinction has to come from carefully examining the distances. It would be much better to indicate smaller distances by making those touch, as touching / non-touching is an absolute difference. Like:

People that would no longer be able to recognize it as the move diagram can simply memorize "this shape is a Camel, and that shape is a Zebra", which still doesn't make them off any worse then whether they had to memorize which kanji stands for Gold and which for Silver.

I also would not make the arow you use to indicate sliding too long; it causes too much disparity in the size of symbols, and also makes the symbols for Rooks and Bishops have too little 'body'. Much better to indicate infinite-range sliding by something in the width of the lines. E.g. like making it extra thick.

I think off-topic discussions in game logs are fine. They are basically discussions between two people which are generally not seen by others and other people cannot easily participate in. Third parties could place Kibitz comments but that doesn't really happen and the no politics policy should apply there. But game log comments between players I think are fine and individual logs could always be made private if it became an issue.

It's a slippery slope to censor all talk of politics (or religion, or the lack thereof, for that matter).

I have changed my wording on that and moved it to the conduct section.