I modified the sentnences

I think you can omit "the type and number of pieces captured in each other", as in a game without drops this must be the same if the position on the board is the same.

Ah, I fixed them. Thanks !

// Even if there are only two Kings left on the board, the King closer to the enemy camp wins in the end, as the same situation cannot be created more than three times.

// About threefold repetition :

If the same situation occurs 3 times with the number, type, and location of each other's pieces remaining on the board, the player who created the situation loses. Simply put, no player can create the same situation 3 times in one game.

 - Even if it does not appear in a row, if the same state occurs 3 times in a game, threefold repetition is established.

 (Of course, it is also impossible to repeat the check consecutively.)

You write the variant does not have piece drops, but then you write: " The player who can no longer move or drop pieces loses".

It is not clear what you mean by 'situation' in the description of the repetition rule. Is that just the same board position, or does it also mean the same player has to be on move? When you write the 'after it occurred twice no one can repeat the position again' you suggest that it does not matter who has the move. Because one it does, only the player that created the position for the first time could recreate it.

And about the absence of a 50-move rule: it it really your intention that when an endgame of King versus King results the players should go on until one of them gets stalemated by the ban on repetition? That seems awful.

The ascii diagram of the initial position seems redundant. And wrong too, because there is a Templar in there.

Can this be republished ?

Yes nice! I suggest to add a definition for "cardinal". Is that synonym of Orthogonal in this context? Maybe some diagrams would help understanding the moves. I have seen a typo : heaens. Why these name of Imp and Fiend? I have to check, I don t know these words in English.

I've fixed the file names, alongside a few formatting bugs and making sure pawns don't share a letter with any other piece.

While I understand why Earth got the a-h files, I would suggest giving them to the skies, while Earth gets the middle ones (j-q). This makes the files be in alphabetical order when the board is setup, and as an added nice bonus, the "missing" files i and r keep the color alternation. vg a1 white, b1 black, ..., h1 black, (i1 white), j1 black, ..., q1 white, (r1 black), s1 white, ..., y1 white, z1 black.

Nice!

It's particularly interesting how some of the 1-piece endgames (rook, angel) end up looking like a classical one to force the king out of Earth and then having the friendly king deliver the final attack. (The angel's isn't completely clear to me, since the lone king may move to the underworld early to avoid the angel's attacks, but I guess that's a significant enough detriment that then the friendly king can manage on his own.)

That helps, yes. To be explicit about the "last" question in each case then, the answer is "not"? My piece on a4 (1) cannot be captured by and also (2) blocks movement (to a1...3) of an opponent rook on a8?

Thanks Ben for the questions, here it should be clearer in the description (although it is already described between the lines, but that is probably not enough).

When capturing on a switch, first assume that there can be only one piece on a switch (here I differ from Fergus Duniho's Reroute66, a variant of my idea). So, for example, if a rook or queen starts from a1...a3 or from rank 4 and the switch is not occupied, then either square 4 or square a4 can be occupied in the switch.

But there is no choice if the switch is occupied by a piece. If a rook or a queen moves from a1...a3 or from rank 4 into the switch, then the piece in the switch must be captured (because two pieces on the switch are not possible). If the piece was on square 4, then the opponent's piece is on square 4 after the move has been executed (applies to a4 in the same way).

Furthermore, an occupied switch cannot be jumped over and a direct change from 4 to a4 (vice versa a4 to 4) is not possible - differently in Reroute66.

Have I understood the question correctly and hopefully answered it correctly? I would be glad.

I didn't really follow the discussion on this variant earlier, but gave the page a fresh read. I think I mostly get it now, except how capturing on a switch works. It might be clarified in the comments, but it should be made plain in the page text as well. (Perhaps it is there too and I have missed it and/or it wasn't quite clear enough.)

Let's say I have a piece on 4, and the opponent has a piece that can otherwise move to a4 but not 4. Can they capture (and if so where do they end their move) or not (if not, could they potentially continue moving through a4 past my piece? I'm not sure that makes sense for any piece at the edge of the board like this, so probably moot, but I'll ask in case)?

Same question reversing 4 and a4. (Now the last question certainly can apply, e.g. the opponent piece is a rook on a8.)

Interesting! Creative! Insightful! I do have one small thought: would you consider transposing the courier and the knight? The knight would then be in its usual position, where expert players use it to contest the center and protect the king and queen. The bishop, a long-distance runner, would probably do just as well near the edge.

I just noticed the question (only 6 years later). I think the stronger player will win. Here is a game with notes:

1. d4 d5 mandatory first move

2. g3 Nc6

3. Bg2 e6
   Pawns only move 1 space, except for the first required move (1. d4 d5)

4. e3 b6

5. f3 g6

6. Kf2 / Hia e1 . . . White Hia goes to e1 (mandatory) 6.. . . . Bg7

7. b3 Ba6

8. Bb2 Nge7

9. c3 Kd7 / Hia e8

10. Nd2 Kc8

11. Nh3 Kb7 As there is no castling in Shatar, Tony moves his King manually to reach something like a Queenside castle. He moves his King to b7 to free the Rook on a8 for his next move.

12. a3 Rc8

13. a4 Hia d7 Neither player is used to the Hia but they know it is a great defensive aide to the King. Tony moves his towards the Black monarch.

14. b4 b5 Timmy wanted to pawn-fork the Bishop and Knight. Tony played b5 to stop it.

15. axb Bxb5

16. Qc2 Ra8

17. Hia e2 . . . Black can’t capture the Hia with his Bishop as he’d have to stop at d3 due to the Hia’s protective field. . . . a6

18. Ra2 Ra7

19. R(h)a1 Hia d6

20. e4 dxe

21. Nxe4 . . . attacking the Hia . . .Hia d7

22. Nc5+ Kb6

23. Nxd7+ exchanging Knight for Hia . . . Qxd7

24. Hia d3 . . . Annoying, the Black Bishop can’t take the Hia . . .Rha8

25. Qb3 e5 Struggling for counterplay

26. d5 Nxd5

27. c4 Bxc4

28. Qxc4 N(d)xb4

29. Qb3 a5

30. Hia c4 Bf8

31. f4 Bc5 (This is not a check due to the Hia)

32. Bxc6 Kxc6

33. Ra4 Nd3+
    Remember that Knights are immune from the Hia forces but Hias can still capture Knights. Here the Hia can’t capture the Knight because that would activate the Bishop’s check (from c5)). White playing 34. QxN would be a blunder because Black would play 34….QxQ and the Hia could not recapture due to the Bishop check factor.

34. Kg2 Nxb2

35. Qxb2 e4 (perhaps dreaming of a Queen)

36. Ng5 Qe7

37. Rb1 e3

38. Qb5+ Kd5 (Kings are immune from Hias)

39. Rd1 #

The Japanese Jostlers pawn is a regular Shatranj pawn (though with a different promotion). The image editor doesn't allow me to mix piece sets, unfortunately.

The piece images are not showing up on the diagram for this game