Comments/Ratings for a Single Item

I also made a rule-enforcing preset for Nutty Shogi.

You can find it here: https://www.chessvariants.com/play/pbm/play.php?game%3DNutty+Shogi%26settings%3Dnutty

The kanji for 'Hawk' (鷹) in the piece Lion Hawk is the same as used for 'Falcon' in the Horned Falcon; while the 'Eagle' (鷲) appears in both the Soaring Eagle and the Free Eagle.

I am curious if the historical moves of the Lion Hawk and Free Eagle are not just Lion + Bishop and Queen + double-move Ferz respectively.

Could the Lion Hawk and Free Eagle instead have moved as Lion + Horned Falcon and Queen + Soaring Eagle respectively?

The Lion Hawk would be more powerful, able to move as a Lion and slide as a Queen except vertically forwards.

The Free Eagle would be slightly less powerful, moving as a Queen with the added Lion power covering two spaces each on the forward diagonals.

This could be the reference that the Free Eagle could move twice as a 'Cat Sword' (Ferz) in the Shōgi Zushiki and Sho Shōgi Zushiki, perhaps created from ambiguity in how the move is described.

I think this is unlikely. I have never seen the Edo-era description of Tenjiku Shogi, but I do know how such descriptions looked in general. They basically showed a drawing of the initial position, where each square contained a drawing of the way the piece moved, with the kanji of the piece name in the center. Sliding moves would just be radial lines, and it would be immediately clear whether the Lion Hawk also had orthogonal slides.

Only complex moves, such as the Lion, would not be obvious from the drawing, and would be described by an additional line of text. I once have seen the picture for the Teaching King (a piece from Maka Dai Dai Shogi); it had lines in 8 directions, each with 3 perpendicular line segments crossing it for indicating the multi-captures of teh Lion Dog. (And the text line then said it would move like Queen or Lion Dog.)

I wonder how the game would be different if the jumping Generals instead had range capturing moves like their Taikyoku Shogi cousins?

The Taikyoku page on CVP states thee Generals move the same there as in Tenjiku. Wikipedia mentions the range-capture, though.

I think it would be dramatically different. Because then you don't need to move Pawns to develop the Generals; they would be fully developed in the initial position. So instead of the usual 1. k6 to threaten mate through 2. Vo10# you could immediatel play VGo10 (capturing your own k Pawn) to check. This would not be mate in that case, because black can still interpose his VG (capturing his own Lion Hawk and Kirin; blocking with his GG (and thus pinning it)  would get him checkmated immediately through 2. GGxi5#). This inflicts heavy losses on gote. And sente could play 1. BGp9 first to make it even worse.

Yeah, that doesn't seem like a good idea.

Do you want to update this page to remove the rule that the jumping Generals cannot jump-capture royalty?

Note: My presets for Tenjiku Shogi and Nutty Shogi follow the rules described in their respective Rules pages, so they will not be changed unless you update them.

Adam DeWitt wrote on 2023-01-01 UTC

Do you want to update this page to remove the rule that the jumping Generals cannot jump-capture royalty?

Note: My presets for Tenjiku Shogi and Nutty Shogi follow the rules described in their respective Rules pages, so they will not be changed unless you update them.

Well, I don't want to have any discrepancy between what is in Wikipedia and on CVP, so I would then also have to change it in Wikipedia. To not run into an edit battle there, this would have to be done with care.

Well, I don't want to have any discrepancy between what is in Wikipedia and on CVP, so I would then also have to change it in Wikipedia. To not run into an edit battle there, this would have to be done with care.

Understandable. I don't think this would be too hard, provided that the forbidding jump-capturing royalty rule is mentioned in the Disputed Moves section along with a reason for the change (perhaps your comment from our conversation in the Betza notation (extended) page?). But then again, maybe the Wikipedia page is far more sensitive to rule changes than I think it is.

All other rules equal, this would make for a very interesting game.

Playing about with this diagram I did seem to have run into some strange bugs:

• Passive burning works correctly, but not when the promotion dialogue turns up. E.g. Rook General g4xg13 from the opening setup survives even though it lands next to the Fire Demon. If I promote it then move it next to the FD again it's burned as expected
• Under certain conditions (it seems to be contingent on whether it's a piece's first move), promotion seems to become corrupted: it gives me the promotion dialogue, but doesn't promote it regardless of which option I select. Subsequent moves allowing promotion (into/within/out of the promo zone) likewise ask but don't react, and the move seems to permute surrounding pieces in an unpredictable way, often with the same highlighting as for burn victims, though sometimes on squares with pieces still (or newly) on them. I think this is contingent on whether it's a piece's first move:
  • 1. RGg4xg13=GG works (besides the above bug allowing it to survive)
  • 1. Pg5–4 Pg12–g11 2. RGg4xg13 similarly works
  • 1. Pg5–4 Pg12–g11 2. RGg4–g5 Pg11–g10 3. RGg4xg13 triggers the bug

The two bugs together also mean that a piece subject to the second bug can never be passively burned within the promotion zone

Indeed, when WeirdPromotion() is used to pieces that would also promote otherwise it involves more than just returning the desired promotion piece. The Diagram makes a special call to it (recognizable by promo = 1021) to inquire whether there might be promotion choice, and in the passive burning case this should be used to suppress the choice that maxPromote and promoZone would suggest. I don't think the current version does this entirely correct for Shogi promotions, though. I will have a look at it tomorrow.

I also experienced this bug with contagious promotions. It seems the shogi promotion prompt is suppressing the WeirdPromotion result.

Indeed, a user promotion choice does overrule everything else. If a fixed promotion is required for pieces that would offer choice, the promotion query should be suppressed. Not only because it would prevail, but because it makes no sense at all to let the user choose and then do something else.

The mechanism the Diagram uses for this is to call WeirdPromotion() with the promo parameter set to 1021 (which is not a valid piece type), to inquire whether there is promotion choice. When this call returns 1001 the query will be suppressed, when it returns 1000 a query is forced, and in any other case it decides itself (based on maxPromote, promoZone and promoChoice).

In the AI it works slightly different; there (for a piece that could normally promote) WeirdPromotion() is called with promo equal to the promoted piece (as per promoOffset), and to suppress the promotion this would have to return 1022. It is the deferral (promo=0) that must be turned into a kamikaze move by returning 251, so that moves that would not naturally promote also get burned.

So when a passive burn is detected, WeirdPromotion() should return 1001 if promo == 1021, 251 if promo == 0 and 1022 otherwise.

I am tempted to submit the following variant ('Minjiku Shogi'):

If I can prove that it is not a forced win for white, that is.

Note that it is entirely implemented without user-supplied JavaScript, using the captureMatrix to define the jumper ranking and Dragon burning, and using the trackPieces with spell=burn to cause the passive burning. The shape of the burn zone is controlled through blastZone=F and spellZone=F.

I am tempted to submit the following variant ('Minjiku Shogi'):

Is it intended that the Fire Dragon doesn't burn when capturing another Fire Dragon?

Replacement capture should work normally. If Fire Dragons try to burn each other, the moving one is destroyed, and won't burn anything else. Just like the Fire Demon in Tenjiku Shogi.

A few minutes ago I was editing the Diagram, slipping in two extra pieces. There might have been a short transient period when the wrong piece had the burning properties. And you might have been so unlucky to try it at exactly that moment. Now it should work.

I was not really happy with the fact that there were two Kirins and Phoenixes on each side; that is very un-shogi-like. The Kirins on b2 / i9 are essential for preventing the initial position to be a forced win. (These play the role of the Eagles in Tenjiku Shogi.) But those on i2 / b9 weren't. So I now replaced those by Phoenixes. (Sorry Jean-Louis, an off-center asymmetry. ;-) ) And replaced the Phoenixes on a2,j2 / a9/j9 by a Side Mover.

This raised the problem of how to make that Side Mover promote. It would be logical to let it promote to Ninja, as these have the same footprint, and the Ninja is stronger. But I had used the Ninja as promoted Minister. So I now changed the Minister's promotion to Rook General (which was not participating before. In Tenjiku Shogi the RG aren't very useful; their fate is to be traded for each other.) That at least makes the promoted Minister stronger than a promoted Gold, which I think it should be, because the Minister is stronger than a Gold. Not that the jumping would be very important in the end-game. But Ministers and Side Movers are very difficult to promote anyway, only stepping forward.

This raises the number of participating types to 27 (20 different moves, of which 3 only obtainable through promotion).

Replacement capture should work normally

Can confirm that it does for most pieces, and indeed FD's burn after most captures, but FDxFD seems to suppress burning fsr

Ah OK, I see it now. Burning is configured through the captureMatrix, which makes it piece-type selective, both w.r.t. attacker and captured piece. But the size of that matrix is of course dependent on the number of participating piece types. I had added a few types since the first version of the Diagram, but had forgotten to extend the row for the Fire Dragon with more burns, so that capturing the few pieces latest in the list did no longer trigger burning.

I am still not completely happy how this works, because the way the moves are encoded as burns now excludes they could also promote. So it would be impossible to burn on the move where the Queen promotes to Fire Dragon, even if I wanted that. Pehaps I should change the standard script such that when a promotion occurs (which could also have be specified by the capture matrix) that it consults the capture matrix (again) to see if burning was specified for the promoted piece making that same move.

BTW, I now altered the promotion of the Diagonal Jumper (= Bishop General) from Minister to a 'Lion'. Which in this case does not have the power of the Chu-Shogi Lion, but only a (two-step) area move. It seemed more logical to use that move than to give it the usual KNAD. The Minister was an embarrasingly weak promotion for the Diagonal Jumper, as it is worth less. (But this promotion could still be useful, as a Minister has mating potential, which the color-bound Diagonal Jumper lacks.) Now it promotes to a strong piece (the Diagram values it as nearly a Queen), which could contibute to its value.

Here's an Interactive Diagram for Tenjiku Shogi (following Wikipedia/CVP ruleset).

March 1 the next addition yearly correspondence championship for Tenjiku Shogi will start.

You can still register by sending a mail to [email protected] .

Perhaps that would be a good opportunity to 'feature' this variant.

This is a copy of that same Diagram, adapted to play according to the 'modern' rules. (I.e. where generals can jump-capture the King, and the FD has a vertical move.) Just to test how long it needs to seet that after 1.j6 there is a mate threat.

[Edit] Not bad! After fixing a bug in scoring the loss of a piece making a kamikaze move, it plays 1... SEn11 in reply to 1. j6 even at the default setting of 2.5 ply. And it takes 47 sec on my PC to play that move.

I have been streamlining the AI of the Diagram a bit, to make it possible to play variants like Tenjiku with a (more) reasonable thinking time. One source of trouble was that the jumping generals always have many captures, which tends to blow up Quiescence Search, even though they are all bad. To combat this I implemented more aggressive pruning of high x low captures of protected pieces. Such moves were already pruned, but only 'after the fact', as it was not known in advance whether a piece was protected. So it such moves would still be tried, and all replies generated. It would then abort when on of these reply moves could recapture. I now use information from the preceding move, recording all friendly pieces that blocked moves that had capture ability. These were protected before the move. I now prune H x L capture of these pieces in the next ply. This is not 100% reliable, as the preceding move could have abandoned a piece it was protecting, or block a protection by another piece. But it is a very fast method, and the idea is that the speedup, through which you could afford larger search depth, would more than compensate for the occasional inaccuracy.

Another source of inefficiency was duplicate moves. Since the AI does not use a transposition table, it would not recognize that a move leading to a position it already searched was a duplicat, and it would search it again. Ideally duplicat moves should not occur, but it is often cumbersome to make an XBetza description that avoids them. (E.g. for BA the A leap is a duplicat if the square it leaps over is empty.) E.g. B(paf)0cB includes a plain cB  next to B, duplicating all non-hopping captures. This could be avoded by writing the non-hopping part as mB, but that gives misleadin highlighting. Particularly disastrous were the area moves: aaK has 8x7x7 = 392 paths when centralized on an empty board (and together with aK and K it makes 356). But there are of course only 49 possible destinations, so each of those can be reached through about 9 paths on average. I now let the move generator suppress repeated destinations for the same piece, if there are no locust squares.

I also hand-tuned the value of the Great General upward; based on its move the Vice General would be worth more, because the value estimates ignore the blocking power. This also distracted (and thus slowed down) the search, by thinking that GG x VG* (kamikaze) was a good move, worth searching.

As a result the time it needs to play 1... SEn11 after 1.j6 at 2.5 ply went down from 47 sec to about 7.

In the process of debugging this I also stumbled on a bug in the piece tracking; at game level this works fine, but in the AI it did not. (Because at some point moves for an enemy piece are generated to test if it delivers check, and this made it track pieces of the opposit color.) This is now fixed too.

Of course Modern Tenjiku Shogi is a game that is impossible to play without opening knowledge, which the Diagram doesn't have at all. So it will still lose in a few moves against a player who knows what he is doing. But at least it plays reasonable moves.

I read on the wiki talk page for Tenjiku about the idea that jumping generals could have been intended to do more than just jump-capture.

The start position shows the generals in front of the Fire Demons, both straight ahead and diagonally. There is no opportunity for jumping generals to capture Fire Demons in the start position, which is intentional.

It doesn't seem right that there is so much protection for Fire Demons, yet in some versions of the game, the King could be threatened and captured without even being able to evade the attack, as it's boxed in.

(In fact, the Bishop General could just mate the opponent's King on the first move if there were no restrictions on jumping.)

If the generals could jump whenever they wanted as far as they liked, the game would become even more tactically sharp than it already is. However, it doesn't break the game.

The Great General can't jump two squares diagonally to threaten one of the Fire Demons as the Rook General would capture it.

The best the Bishop Generals can do is to manoeuvre and attack a Horned Falcon or Soaring Eagle.

The idea that jumping generals could capture all of the opposing pieces they jumped over in one turn is plausible. I don't know if the game would break, but the inventor(s) of Tenjiku weren't sentimental about pieces — allowing Fire Demons to punch large holes into positions — so having more pieces do the same seems logical.

The Great General could just capture the Vice General; Free Eagle; Queen; Drunk Elephant; and a Pawn on the first move (with check). This doesn't seem intentional.

One of the Bishop Generals could move to the edge of the board and threaten to capture a Soaring Eagle; Water Buffalo; Phoenix; Drunk Elephant; and a Pawn. The Soaring Eagle can leap out of the way, though.

Otherwise, everything is sufficiently well defended that jumping generals couldn't capture a total of anything worth more than themselves, as they are even stronger than their current form.

That could be why Fire Demons are so powerful — with burning and sliding and an area move — because they would be ecliped otherwise by the jumping generals.

The board size (16×16) and large number of pieces that do very little makes more sense if there were even more crazy pieces. The game might devolve into a capture-fest, but that might have been the intention.

Whoever had the imagination to create this game has to be admired. I consider this to be the most ambitious of all the historical variants as its construction is so delicately balanced, even with the ambiguity in the rules themselves.

Here's an Interactive Diagram for Tenjiku Shogi (following Wikipedia/CVP ruleset) with the new ruleset used on Wikipedia, which was edited to remove the rule against capturing royalty via range-jump. The GC preset has been updated as well.