Talk:Maka dai dai shogi

Early talk
I just checked two pieces with Japanese wikipedia, the teaching king and the lion dog, and they were both wrong. J-wiki even gave the refs for one of them, the powers of which have been debated in Western descriptions. All the pieces need to be reviewed. kwami (talk) 18:45, 6 August 2008 (UTC)

Sourcing notes copied from J-wiki; moves confirmed with J-wiki diagrams. However, these do not always match the J-wiki verbal descriptions, and these have not all been checked. kwami (talk) 22:39, 6 August 2008 (UTC)

Neat deva-capturing rule. Never noticed it :( kwami (talk) 08:33, 13 March 2009 (UTC)
 * When a piece captures a Deva/Dark Spirit, then that piece promotes to a Teaching King/Buddhist Spirit (the promoted form of those pieces) immediately. OosakaNoOusama (talk) 20:01, 2 June 2013 (UTC)
 * It's under the "Promotion" section already, but I would agree that it's a good idea to place it here. Double sharp (talk) 03:51, 3 June 2013 (UTC)
 * I wonder where these Deva/Dark-Spirit promotion rules come from. I did not find them in the TSA description of this game. (But perhaps I overlooked it.) I agree they are good rules: they effictively prevent the two strongest pieces (Teaching King and Buddhist Spirit) to be traded away, and thus serve a similar function as the Lion-trading rulse in Chu. I have one concern about them, though: what happens when a royal captures TK or BS? If my sole King would be forced to promote to BS when it captures one, rather than to Emperor, I could not capture at all. (Or could I? The opponent would not have captured my King, and would not get any chance to do it at all now it did a disappearing act...) I was contacted by someone from Japan who seems to be influential in the Japanese Chu Shogi Association, asking me to create an AI for Maka Dai Dai, and sending me rule details. According to him Elephant, Prince/King and Emperor would keep their usual promotion (or non-promotion) when capturing TK or BS. He also insisted that capturing Dv or DS would not cause promotion of the capturer to TK or BS. H.G.Muller (talk) 10:06, 10 March 2015 (UTC)
 * I don't know: they're not covered in the footnotes, which represent the only things I know about the content of the primary sources. It is a nice rule, but if it's not historical, we should not present it as primary IMHO. My (very bad) comprehension of Japanese tells me that this rule appears on Japanese Wikipedia (see their articles on those two pieces): the Chinese Wikipedia presents this rule too.
 * Well, in the smaller variants, you could just as well phrase the rule as "you lose if you have no royal". Only from maka-dai-dai on up would there be a difference. But given that you can (rather suicidally) move your king into check in any of the variants, I suppose you would be allowed to make that capture, although it would not be a very good move. Of course, this all assumes the Wikipedia ruleset. Double sharp (talk) 13:31, 12 March 2015 (UTC)

Diagrams?
Any advice on how to construct the diagrams for the pieces' movement? OneWeirdDude (talk) 18:04, 5 June 2013 (UTC)
 * See User:Double sharp/Maka dai dai shogi and my sandbox. I started it last year, but never got around to finishing. Double sharp (talk) 08:34, 6 June 2013 (UTC)
 * Well, I restarted work on it on 21 December this year. At this rate it should be finished quite soon. Double sharp (talk) 05:04, 24 December 2013 (UTC)
 * ✅ The huge table containing diagrams is now in the article! Now to work on tai shogi. Double sharp (talk) 15:22, 24 December 2013 (UTC)
 * Or you could just hit yourself on the head with a brick. (I think I'd take the brick.) — kwami (talk) 22:04, 24 December 2013 (UTC)
 * Heh. That's not going to stop me from trying! :-) (Although I would agree that the brick would be preferable to doing taikyoku, whose inconsistencies with the smaller games would be immensely annoying.) Double sharp (talk) 03:33, 25 December 2013 (UTC)

After eight years, I eventually did it anyway. See tai shogi. (It is not as bad as it looks, because all but nine pieces appear in one of the smaller variants already, usually dai dai or maka dai dai. So it is just a copy-paste effort. Incidentally, the fact that it's such a combination and thus doesn't require much of a learning curve makes me suspect that whether or not dai dai, maka dai dai, and tai were actually played, they likely were intended to be played. Taikyoku with its widely divergent moves makes me more sceptical, but maybe those moves are modern inventions anyway from what I vaguely recall hearing.)

For this variant the Japanese Wikipedia has very different promotions from the English sources; usually they agree. I mostly followed the English sources, except when they must be wrong (e.g. lion dog as three-square limited range mover, which must be wrong because then "teaching king = queen + lion dog" is pleonastic), because in dai dai shogi and maka dai dai shogi you can see from the footnotes that the original Japanese sources disagree a lot, so while the English sources sometimes made arbitrary decisions on which move to pick, it's probably better that way. (So that prospective players have one set to follow.) I also suspect that the English sources' promotions make a more playable game, because then not all the strong pieces are forced to be shy of capturing for fear of demotion, but I admit that I haven't ever played a game of tai shogi and don't really intend to try. :) Double sharp (talk) 09:03, 3 January 2022 (UTC)
 * Not so big a brick, then.
 * I wonder if the Japanese promotion rules aren't the point of the game: you have some powerful pieces, but only get to use them once; others may become powerful, but it's difficult to get them into position to promote. I imagine that might create a very different feel to the game, one where you need to carefully consider the expenditure of your resources, or where you try to force your opponent to capture your pieces and so dissipate their forces. And the pieces that can promote to power are a defensive threat rather than an offensive one: if you invade that area, you risk giving your opponent more power if they can capture your piece, one they don't have to worry about wasting, so sacrificing a pawn could be prove rather costly.
 * A rather long-haul game. But then, we're talking about monks who probably had a lot of time on their hands. And I wonder if there might've been a different philosophic point to this game -- a capture may weaken the attacker and strengthen the defender. With the English rules, it's just a big game of chess. — kwami (talk) 09:18, 3 January 2022 (UTC)
 * The English rules have demotions too. They just don't make all the powerful pieces demote. According to English rules, Hook Mover and Capricorn demote to Gold, but Long-Nosed Goblin and Peacock do not promote. According to Japanese Wikipedia, all four demote to Gold. In Japanese Wikipedia you can never get another Hook Mover, because the Poisonous Snake is said to demote to Gold in Tai; in English rules, you can get another one by promoting the Poisonous Snake. So the difference rather seems to be: in the Japanese Wikipedia rules all powerful pieces can only be used once, whereas in the English sources' rules some can only be used once, but some are permanent things. In fact, English rules have it so that among the hook movers:
 * You have one that you can't keep, but can try to get a non-demoting copy of (Hook Mover);
 * You have one that you can't keep and can't ever get back (Capricorn);
 * You have one that you can keep, and can try to get another copy of (Long-Nosed Goblin);
 * And you have one that you can keep, but can't ever get another copy of (Peacock).
 * Since Tai appears to have the point of including every piece, having all four possibilities actually realised also seems to make some sense, and it keeps your philosophical point in the first case. :)
 * Also, the fact that you can never get another Hook Mover under Japanese Wikipedia rules is a bit weird for the power piece. It reminds me of the reasoning behind the rule The Chess Variant Pages advocates for Maka Dai Dai: there you only are forced to promote when you capture an already promoted piece. Because otherwise, the Hook Mover and Capricorn demote on the first capture and you can never get another one via promotion, so there's no possible fun to be had with them. (See the section below.) That seens strange for the most powerful pieces, considering the rules in Chu specifically aimed at keeping the Lion alive despite the fact that you can get another one by promoting the Kirin.
 * The last reason is that the royal piece in Tai is an Emperor, and I can't imagine how he's ever going to be checkmated if all the hook-moving pieces are gone by the time we reach the final assault, but I admit I haven't played it to see what'd happen. :)
 * But of course we cannot be sure. To make the table, I had to pick one version to be primary, and the confusions and inconsistencies in the Japanese Wikipedia articles meant I didn't want to pick it to be so. I left their promotions in the footnotes, so anyone who wants them can still have them. Double sharp (talk) 09:37, 3 January 2022 (UTC)

Promotion
The May 25, 2013 patch from OneWeirdDude removed the mentioning of the = sign for deferral from the notation section, with the remark that promotion is always compulsory. The Promotion section still mentions that promotion is optional when capturing unpromoted pieces, however. So I think that patch should be reverted. I don't know if it is truly controversial whether promotion indeed can be optional in Maka Dai Dai Shogi. The game seems designed for optional promotion: many of the important pieces in the initial setup (e.g. hook movers) cannot be obtained through promotion, and the game would quickly degrade into a rather boring battle between vast armies of Golds when every capture forced them to promote. In Dai Dai and Tai, which are reported to have mandatory promotion on capture, the important pieces either do not promote at all (Dai Dai), or non-promotable versions of them can be obtained by promoting lesser-valued pieces. So for those games compulsary promotion works fine. But for Maka Dai Dai it would be a disaster. H.G.Muller (talk) 10:25, 3 July 2014 (UTC)
 * I reverted their edit. I'm guessing it was based on a misreading of the "Promotion" section: I reworded it slightly in an attempt to make it clear that you do have a choice if the enemy piece you captured was unpromoted. Now it reads "Pieces that can promote can choose whether or not they will promote when they capture an unpromoted enemy piece; however, if a piece captures a promoted piece, it must promote if able." Double sharp (talk) 11:55, 3 July 2014 (UTC)

Capricorn and hook mover
Hi, I saw in the notes that the capricorn and the hook mover are said in some western sources to be able to move only to make a capture (contrary to dai dai, wherein the long-nosed goblin and the hook mover are not compelled to capture to move). It has been quite a while since I last looked at it, so maybe my question is pointless, but wouldn't the initial position of these pieces in maka dai dai (and hishigata) confer them too much power, were they be able to move without capturing? – UseresuUK (talk) 10:51, 7 October 2014 (UTC)
 * Maybe. Big disclaimer: I have not played maka-dai-dai at all, but the frontal positions of these piece do make me worried that this allows an instant win via blitzkrieg. However, see note 23:
 * "The Edo-era rules appear to say only that the capricorn and hook mover move two times as a bishop (mcaB) or rook (mcaR), implying that they may capture twice. There also appears to be no mention of needing to capture in order to move, which occurs in some Western sources. However, making these pieces able to capture twice makes the hook mover move differently in dai-dai and maka-dai-dai shogi, and hence is probably an error."
 * I would personally go with allowing them to move without capturing, but not giving them lion power. In the first place, this keeps the hook mover the same in both dai-dai and maka-dai-dai: it would otherwise be the only piece that appears in both games but does not keep the same movement pattern. Secondly, allowing them to capture twice make the blitzkrieg even more likely (because it allows igui from a great range, defense seems almost futile).
 * Now, you may have noticed that my arguments are mostly based on the original sources. This is all fine and well, but it may well be that this gives Black an instant win out of the opening. To find out for sure, though, analysis is required, and I do not think I can do that well. Double sharp (talk) 15:21, 30 November 2014 (UTC)

I finally figured out the origi of the misconception that the hook movers could only capture. The text in the 'Rules Help' of Sean Evans' ShogiVar program says: "The 'Hook Mover' is not a jumping piece, and must end its move on a capture". This was intended to mean that it had no Lion power (as the implementation of Maka Dai Dai Shogi in ShogiVar shows). But some people apparently took it to mean that the second leg was only allowed on captures! H.G.Muller (talk) 10:36, 28 July 2017 (UTC)
 * It looks like sanity does not always prevail, even if it quickly gets us to the correct conclusion. ^_^ Double sharp (talk) 14:04, 28 July 2017 (UTC)

Emperor
Can it pass a turn by jumping to the same square it already is on? Double sharp (talk) 06:19, 2 December 2014 (UTC)
 * I would say 'no'. In general pieces cannot move to where they already are, even though that square is the only exception in the set of squares with a certain property that they can go to (e.g. a file or a rank for Rooks). Pieces like Lion that can pass a turn, can only do so because they move multiple times, and happen to move back to the starting square after having left it. They cannot pass a turn if there isn't an empty square next to them to move to first. H.G.Muller (talk) 19:47, 9 February 2016 (UTC)
 * Perhaps this is of interest too: I asked the guy that sent me the rule file for the 'Modern Maka Dai Dai Shogi' about a situation where an enemy Gold is protected only by its Emperor, but attacked by my own Emperor plus a Rook that is pinned against my Emperor. So RxG would lose me my Emperor to the pinning Bishop, B x +K. But capturing +K x G is then allowed, because his Emperor cannot recapture mine, as I am protected by the Rook. So a piece is only considered protected by its Emperor if it is not attacked by another piece. This is the opposite logic as used in Chu-Shogi Lion capture, where if an opponent Lion is attacked by my Rook and Lion (but the Rook is a 'false attacker', because it is pinned against my King), and protected by another Lion (+Ky), I cannot capture Ln x Ln even though the recapture +Ky x Ln would in itself violate the Lion-capture rules, and could thus be argued to not count as protection. H.G.Muller (talk) 20:03, 9 February 2016 (UTC)
 * This seems to imply that emperors cannot move into check. This would be very interesting given that the rules on check and checkmate in most shogi variants for kings seems to be that you can move into check freely – though it is not a very good thing to do. (Well, a counterexample: on an 8×8 chessboard, let's have a black queen on b7, a white king on c2, a white rook on g1, and white crown prince on f6, and a black king on h7. Here 1.Kb3 is not a bad move as 1...Qxb3 simply gets answered by 2.Kg7#.) So the rules appear to be different between kings and emperors. I would much prefer the latter be the case, as I find stalemate as a win rather distasteful. (To me, a chess variant should end when one player has no moves, and then we look at the status of his or her king, the objective piece.) I don't know how much of this is because FIDE chess was my first variant: I felt the same annoyance about the treatment of stalemate and repetition in xiangqi. As for the effects of the byzantine rules against lion-trading, I doubt this will have much practical significance as 1.RxLn would obviously be a much better move, gaining a substantial material advantage. However, rule confusions may be annoying. Double sharp (talk) 13:43, 10 February 2016 (UTC)
 * Oh, it's pinned. Hmm. This is actually an interesting situation that should be raised on Talk:Chu shogi. The chu shogi article actually says that such "false attackers" do count as protectors, which is logical since you can legally (if stupidly) expose your king to check in chu shogi. Is this specific case ever stated anywhere? Double sharp (talk) 05:49, 2 March 2016 (UTC)

I see the current version on the Maka Dai Dai and Tai pages is that the emperor can expose himself to check making normal moves or capturing non-royals, and that it is only the capture of royals that is restricted. Honestly, I think this makes more sense, and have adopted it throughout. It is more consistent with the usual idea in shogi and its variants that exposing your king to check is just a bad idea, and not actually illegal. The restriction on capturing royals is the only really necessary addendum, without which the game breaks: if you are allowed to capture royals while exposing yourself to check, then games of Maka Dai Dai would end almost as soon as somebody got an Emperor. The Crown Prince is unlikely to be created and there would be no way to retaliate before game over. Restrictions on capturing royals fixes this. Double sharp (talk) 06:42, 20 October 2023 (UTC)

Modernisation efforts
The idea to make this game, surely more worth a revitalisation than the boring bibliographical nightmares that are dai-dai and tai, more familiar to players of chu and dai shogi is laudable. (Dai and tenjiku probably deserve a revitalisation as well. Ko just strains the memory too much IMHO.) But why stop there? Here, some ideas:

The following promotions changed:
 * Free dragon → RbB (to match the reclining dragon)
 * Free wolf → frlQ (to match the evil wolf)
 * Free bear → BrlR (really, how much is the jump worth?)
 * Blind bear → FrlW (to match the free bear)
 * Old rat → fWbF (to match the bat, which maybe should become "free rat". Its move is not stated anyway in the sources, so it does not have to match dai-dai anymore.)

I leave the "free" mechanic alone, as that is not hard to learn. We can keep the "free boar" as a pun, or simply make it a rook (like free cat → bishop). I prefer the rook.

The phoenix and kirin can change their promotions to queen and lion, respectively, to avoid learning yet another pair of new pieces.

The lion dog can use the Japanese-Wikipedia move. For symmetry with the lion, though, we can have the furious fiend (promoted lion) become lion + Q3 (gaining a partial lion dog move), and the lion dog now promote to a 聖像 (seizō; icon, sacred image, statue) that moves as lion dog + two-step area move (cf. tenjiku shogi)! This idea was given by H.G.Muller.
 * Or perhaps we could have both promote to the furious fiend, now defined as lion + lion dog. It's not unheard of that pieces share promotions: it's just unheard of that that shared promotion is not the gold general. But hey, it's a modernisation: we can do what we like! Double sharp (talk) 19:46, 12 March 2016 (UTC)

If I were trying to make it slightly more familiar for chessplayers, BTW, I'd use FIDE pawns, FIDE knights and FIDE-like go-betweens that can move or capture as backward pawns, and become free on promotion; but I would not change the earth general. (For people liking shogi pawns, I'd still recommend using FIDE knights, so that it becomes more on a par with the donkey.)

Promotion should either continue to not be mandatory, or demotion should be abolished. Capturing a deva or dark spirit should still promote you to TK or BS respectively. Double sharp (talk) 05:52, 9 February 2016 (UTC)

Tallying all of this up, and assuming that the "free" mechanic is known, and a knowledge of dai shogi (itself easily obtained from a knowledge of chu shogi), here is all one needs to learn:
 * The Deva and dark spirit and their promotions, the teaching king and Buddhist spirit (since these are so powerful, they will be learnt);
 * The tile general, blind bear, and old rat (since the earth general is essentially the go-between);
 * The old monkey and Chinese cock and their promotions, the mountain witch and wizard stork;
 * The lion is known, but now it is paired with the lion dog, and their respective promotions: the furious fiend and sacred image (since these are so powerful, they will be learnt);
 * The Buddhist devil, she-devil, and donkey (new limited-range riders);
 * The left and right chariots and the side flier (new riders);
 * The capricorn and hook mover (since these are so powerful, they will be learnt);
 * The new promotion rules. Perhaps pieces that become free should be marked out in some way, as should pieces that become gold.

Totalling it up, it amounts to 9 new pieces that will be remembered because they are really, really important. Besides that, 13 new weak pieces need to be committed to memory. That makes the necessary workload that much smaller, comparable to what happens when one first learns chu shogi, or dai or tenjiku based on chu. Double sharp (talk) 06:02, 9 February 2016 (UTC)
 * I agree that Maka Dai Dai is not difficult to learn at all. (But of course I never looked at it in kanji representation. That does mean I have to recognize the name from the shape of the mnemonic piece, however, and I still would have to remember how it promotes.) There is really a lot of systematics and regularity in this game. I suspect the illogical Free Boar move is contamination from the much more popular Chu Shogi, which has a piece that happened to have the same name as the regular derivation would give for a promoted Angry Boar. With Macadamia Shogi I tried to lower the threshold for getting acquainted with Maka Dai Dai Shogi even more, by predominantly keeping the Chu-Shogi pieces. So the only new minor pieces you would have to learn there are te Flying Dragon, the Left/Right Chariot and Wrestler/Guardian of the Gods (a pseudo-symmetric pair). And the new promotions of the Kirin/Phoenix. H.G.Muller (talk) 17:49, 9 February 2016 (UTC)
 * I did a similar count for dai-dai, BTW, and the result is that you need to learn 35 new weak limited queens, even assuming you've already mastered dai and tenjiku. Not something most people would do. However, I still think maka-dai-dai has a problem in that the new tier of pieces (the hook movers) aren't strong enough to really make this go any faster than chu or tenjiku. There is however one very interesting wrinkle: "The Edo-era rules appear to say only that the capricorn and hook mover move two times as a bishop (aB) or rook (aR), implying that they may capture twice." (note 23). If these are the real moves, these suddenly become very scary pieces that dominate the whole field, as now they can kill and run away, and have huge mobility. (Betza seems to have thought that aR was worth about 30 pawns on 8×8 standard chess.) Otherwise, they're like emperors: awesome mobility, can never be trapped, but as long as everything is kept defended, they can't do anything. Double sharp (talk) 15:08, 10 February 2016 (UTC)
 * P.S. In that article, Betza suggests that aF (doublemove ferz, but not allowed to jump; so if it starts at f1 instead of White's light-squared bishop, it cannot move to d3 until the e-pawn moves) should be worth about as much as the FAD. Is this true? In fact, Betza seems to suggest that a doublemove piece should be worth about the square of the normal piece – implying that a doublemove queen should be unstoppable. How close is this to the truth? Double sharp (talk) 15:12, 10 February 2016 (UTC)
 * I am very skeptical towards the idea that "two times as a Rook" could have implied double capture or rifle capture. I think all descriptions of the Lion always explicitly mention the possibility of igui. If the hook movers could do something similar it would have been described much more elaborately than just this simple sentence. Besides, Rifle Chess seems to be an utterly boaring game: you park your Queen in the center, and then startpicking off opponent piece one by one, without him being able to do much about it. Only a Knight could chase that Queen away, but it takes too long to bring it in position. With Lion-empowered double Rook even mounting an attack on it would not stop the barrage, as it would simply move back to another square on the same file. So I don't think it would give an interesting game. My guess is that in Maka Dai Dai the hook movers are intended to make sure the Teaching King and Buddhist Spirit cannot go on plunder raid through enemy lines without restraint; you would always want to keep them under attack with a hook mover. The opponent would then have to keep them protected (to make sure he gets the TK or BS back when he recaptures), meaning they cannot go on solo action. But I certainly expect a Maka Dai Dai game to take more moves than a Chu or Tenjiku game. This is one reason why I made Macadamia Shogi. This has 48 pieces against Chu's 46 (one of the extra pieces being a Pawn), and on the average the pieces are more powerful, so its games must be shorter than Chu games. (OTOH, once a player promotes to Emperor it is difficult to image how he can lose other than through total annihilation.) As to the strength of maF. I never measured that, but I would expect it to be a bit weaker than FAD (even assuming it can do a single F step as well). The A move is lame, and the D move is 'multi-path lame'. The latter I don't expect to be too detrimental. But on 8x8 the difference between the A move being blockable or a direct leap is as much as cutting all moves of range 3 and more off the Bishop: FA is about as strong as a Bishop. It makes sense that the value of a double-mover is quadratic in the value of the underlying piece. But in case of multiplication one must worry about what the unit is, and it would be unlikely to coincide with the value of the Pawn. An FAD has 12 targets, making it Rook-class,while a Ferz is only worth 150cP. That would suggest 2 times the value in Pawns squared for the ('non-Lionic') double-mover. But, some pieces 'square' better than others: KmaK covers only 24 squares, due to large overlap of the two-step moves with both one-step moves and each other. This does make the second-ring moves 'less lame', but that is not nearly as good as spreading them out over different targets. An NmaN, however, covers 32 squares with its two-step moves, none of which overlap with its 8 one-step moves. So even if K and N are about equally valuable because they each have 8 moves, the two-step N has nearly double the moves, (40 vs 24) and must be vastly more valuable. (This 'second-order mobility' no doubt contributes significantly to the value difference between the various pieces with the same number of moves.) H.G.Muller (talk) 19:58, 10 February 2016 (UTC)
 * Truly, this is most useful information! So perhaps replacing FAD with maF in the Clobberers from CWDA may be enough to get them to FIDE strength, and create a more unique piece. (This would have to be tested, though.) [Addendum: OTOH, the fact that this piece has the ability to act as "shooting Ferz" and "doublemove Ferz", and capture two pieces in a turn – though not jump – may increase its value as well. 08:53, 11 February 2016 (UTC)]
 * True, this would be a good way to weaken the Clobberers. I would have to do a real test to see if it would not weaken it too much. However, most other armies are also stronger than FIDE, so perhaps CwDA as a whole would benefit more from a method of enhancing FIDE than from weakening the Clobberers.
 * I think the main raison d'être for CWDA is that it gives players of FIDE Chess a new experience that can be treated just as seriously. For one, the only way I've ever been able to introduce different-moving pieces to some of my chess-playing friends IRL is by showing them the Clobberers and offering to use them against the FIDE army (and their unfamiliarity with the Clobberers makes me look stronger than I really am ^_^). I tend not to show them the Nutters (because the forward-backward difference seems to short-circuit the average chessplayer's brain who isn't used to dealing with it outside pawns) or the Rookies (because they're more overpowered, while the Clobberers are at least closer). Changing the FIDEs to deal with this problem may create a more balanced game, but I think it would also mess with Betza's great intentions in creating it.
 * Regarding the Rookies, I notice in the CVP comments that Fergus Duniho misinterpreted the HFD as simply an FD (the kirin). Would that be enough to reduce it to FIDE levels? (This way they only have five pieces that can force checkmate.)
 * OTOH, one problem with reducing FAD to maF (I think aF is too strong because of the igui and double-capture powers) is that while it hobbles them in the opening, they act pretty much the same in the endgame when blocking isn't an issue, and unlike riders, their longest moves are still short enough to matter even with very few pieces left. So this would have the unfortunate problem that it would force the FIDEs to go into an all-out opening or middlegame attack, as otherwise they are likely to lose in the endgame if one or more maF is still around, as the maF seems to turn from a minor piece to a rook-class piece as the board empties. (But maybe that could be a charming feature.)
 * If the maF is still too strong, I still have an idea. Since the BD is a rook-level piece that is fast to develop, unlike the FIDE rook, which makes up for the fact that the princess is about a pawn weaker than the queen, perhaps we simply need two normal minor pieces, in which case the ubiquitous phoenix and kirin (WA and FD) recommend themselves again. (Best knight substitutes ever. I love them! The WD and FA are a little too much like the orthodox rook and bishop for my taste, although the WD's awkwardly slow speed and lower value does give it a distinct character. The WF is really slow and not that good if the endgame hasn't arrived yet – even worse than the WD – and the DA just is too limited.) Double sharp (talk) 12:30, 16 February 2016 (UTC)
 * I have started some games FIDE vs these weakened Clobberers, (with alternating color), and after 110 games they are at exactly 50%. So it seems that the maF is just what is needed. It makes sense that this piece is just a bit stronger than a normal minor; the FA is similar to the Bishop, and instead of having the jump to the A squares this one has an 'almost unbockable' move to the D squares. This should be better. If we estimate the value of sliding to an A or D square to 2/3 of the value of an unblockable jump to it, and the value of a two-path move to it somewhere in between (5/6), the maF would be F + 2/3 A + 5/6 D = 2.5 atom. That should put it slightly closer to a minor than to a Rook. The Phoenix is indeed a very nice piece. I used it in Team-Mate Chess. Its advantage compared to a Knight is that it has 'semi-mating potential': in combination with another piece it can checkmate a bare King, because it can get in 3 moves from c1 to a1. (Actually what the requirement is that an uncapture + move + capture can do this. A Kirin also meets the requirement.) So a pair of Phoenixes wins, where a pair of Knights draws. OTOH a Phoenix loses to a Rook, while a Knight would in general draw there (on 8x8). H.G.Muller (talk) 18:03, 17 February 2016 (UTC)
 * Wonderful! I can return to playing fair CWDA now! (This allows the maF to move simply as F, I assume?)
 * How accurate is my other suggestion of replacing the Rookies' HFD with a kirin (FD)? Double sharp (talk) 05:30, 18 February 2016 (UTC)
 * Indeed,it was FmaF. The final verdict after 760 games was 51.4% in favor of FIDE, but the statistical error (with only 20% draw rate) is 1.6% (which corresponds to ~10 centi-Pawn). I expect reducing HFD to FD would hurt the Rookies way to much: you reduce two Rook-class pieces to minors, which should cost 3.5 to 4 Pawns. Besides, the HFD is one of the 'iconic piece' for the Rookies, the other pieces being rather unimaginative. The Kirin has nothing Rook-like, (is even color bound), and doesn't really belong. You could of course introduce a degree of lameness into the HFD, like allowing the H move only if the D squares are empty, or perhaps even the D square plus one of the F squares. But to weaken the Rookies it would be much more natural to trim some more range of the Rook (R3 instead of R4). H.G.Muller (talk) 12:01, 18 February 2016 (UTC)
 * 51.4% is OK, I think. The White advantage in chess is more than that (about 55% IIRC). I take your point about the overcorrection: I don't know what I was thinking. Yes, I think R3 would be the closest to Betza's thinking here for a short rook move (he even tried R3 at one point before changing it), just as I think maF was an OK correction since it moves to the same squares as the FAD on an empty board. (OTOH I thought the WA fitted perfectly with the Clobberers even though it wasn't colourbound and it was only half-bishopish, but it doesn't really matter.)
 * I take your point about imaginative pieces. OTOH when I fielded my amended Clobberers to a friend he was blasé about the whole thing until I mentioned the maF, after which he got interested. So maybe what you have to do is to use a sort of piece that does not exist in orthodox chess. Chess has steppers (the king and pawn), rangers (the rook, bishop, and queen), as well as jumpers (the knight). What it does not have are ranger-jumpers like the nightrider, limited-rangers like the R4, asymmetric pieces other than the pawn (so the thought of differing forward and backward moves could be fresh), and doublemove and area-move pieces. (Iron pieces would be interesting but seem too alien to me.) I do admit to having a special place in my heart for very simple, basic, symmetrical pieces (hence the phoenix-and-kirin fangirling), but that is what one finds in FIDE already, so I completely understand that one is probably looking for something new.
 * I haven't touched the Knights very much. Maybe we need to delete the sideways king moves from the charging knight. Double sharp (talk) 13:17, 18 February 2016 (UTC)
 * I'm really impressed with Macadamia, which looks less crowded and more interesting than Cashew. I haven't looked into Nutty as much because I feel that tenjiku is still playable. I do miss the dragon horses, though. Perhaps we could put them back in by replacing the weak F2 with B and B with WB, to bring back this iconic piece. I think it's an important inclusion, as it even appears in standard shogi. (Also, the bishops would start on the same-coloured squares either way, so perhaps breaking the symmetry to fix that might be conceivable. In fact, that is one annoying problem with dai.) I'd also use your suggested promotions for Ln (Ln + Q3) and LD (LD + maK), but this would make it not entirely consistent with maka dai dai. (But then chu is not consistent with dai either, as the latter lacks the rules against trading lions.) Double sharp (talk) 07:37, 11 February 2016 (UTC)
 * To preserve the 'feel' of the larger games I tried to preserve the board filling fraction. Dai Dai is indeed much more crowded than Maka Dai Dai, with the same number of pieces on a 20% smaller board. Not having the DH was indeed one of my biggest regrets in Macadamia (and not having R or B in Cashew). But one of the design criteria was to not disturb the ratios of the number of pieces in the various 'classes'. Q+DK+DH make up the class of non-promoting pieces in Maka Dai Dai, FD+VO the class of range-2 movers that promote to Gold. Entirely eliminating the second class to fully retain the other would violate that design criterion. I could have kept one DK and one DH (and similarly one VO and one FD), but that would break the symmetry, which I also considered an essential part of the 'feel' of Maka Dai Dai (and, next to the density, one of the major differences with Dai Dai). One cannot make an omelette without breaking eggs, and it seemed more important to keep the DK as an orthogonal mover than the diagonal DH, because I already eliminated the Side Movers, and you need some way to keep the Lion out. And Bishops do start on the same color also in Maka Dai Dai. I guess the choice for what would be best here depends on whether one views Macadamia Shogi as a 'miniature' version of Maka Dai Dai or optimize it as an independent game. H.G.Muller (talk) 10:26, 16 February 2016 (UTC)
 * I think the latter may be a better idea. Who's going to play maka-dai-dai? Who has time for a whole game? (Unless it is played by mail or email, in which case you spend less time on it at one go, but it stretches out for a year or so.) Presumably people in the Muromachi period had a lot of time and hence a more laid-back way of looking at things, hence why chu could sensibly be called "middle" and things like dai could be played and tenjiku be evidently designed for playing. They would have time to play for that long and remember all the rules. (Sure, it is not that bad, but it looks bad, which is what ends up dissuading people.) But to us, chu is on the high side of the spectrum of chess variant sizes, and tenjiku only looks reasonable because the added pieces are so powerful that the game is shorter than it would look (and because much of the memorisation work has been done for you if you already play chu). So if we want to rescue the best elements of maka-dai-dai (IMHO, the hook movers, the Japanese-Wikipedia lion dog, the deva- and dark-spirit-capturing rules, and the "steppers promote to equivalent riders" mechanic), it seems that they would work better in a smaller variant that is nevertheless a different offshoot, one that may not keep 100% compatibility with the larger one. Double sharp (talk) 11:47, 16 February 2016 (UTC)
 * You might very well be right. But I can always add these alternative rules in the Notes section of the CVP description. But I won't start tweeking the rules before I have an opportunity to actually watch games being played according to these rules, i.e. until HaChu can play it. And there is another side to this: people might be more motivated to do one of the few centuries old historic games than one of the thousands of recent inventions, before they fully realize how long such a game would take and how difficult it is to learn. When Macadamia is presented as a true subset of Maka Dai Dai they would then be more inclined to touch it as a preliminary exercise. H.G.Muller (talk) 21:17, 16 February 2016 (UTC)
 * Now that you mention the inconsistencies between Chu and Dai: I have always felt that there are suspiciously few of those, and suspect there has been some 'back-contamination', where some of the changes/improvements that were made when Dai was shrunken to Chu found their way back into Dai once Chu had become the standard. E.g. w.r.t. stepper promotions. In the games predating Dai (Heian, Heian Dai and Sho) all steppers (and Lance) always promoted to Gold, and the promoted forms of C, S, FL and G in Chu seem higly arbitrary. It is strange that the promotions of S and G were changed when Dai was derived from some predecessor, but not the promotions of Stone and Iron. If one wanted Bishops and Rooks to occur as promoted versions of something else, FD and VO would be much more logical choices; +FD = G could even be a demotion, and for pieces obviously stronger than G both Sho and Heian Dai have a dedicated promoted forms to make sure they truly promote. H.G.Muller (talk) 10:30, 16 February 2016 (UTC)
 * I dunno about +FD = G being a demotion, since the G is not colourbound, and I think the slight difference in range (2 squares versus 1) does not make that much difference on a 15×15 board. Indeed, the lion and lion hawk not seem to dominate in tenjiku (16×16) even after the fire demons are gone, and this seems to be because their "death zone" is a much smaller fraction of the board. In fact, the atmosphere of a game of tenjiku shogi seems to be that the FiDs burn brightly and snuff themselves out, causing damage that has to be dealt with in the next regrouping phase, when the remaining pieces have their time at stardom – except that the 16×16 board is so big that they really can't be emperors (not the maka-dai-dai piece) of the whole board like the lion in chu shogi can be if centralised and placed many ranks up. As a result it begins to feel less like "Lion Chess" like I notice chu can be – much of the danger comes from the lion starting where it can instantly jump out, and you have to keep restraining the enemy lion without hanging your own like a fool (which I keep doing against your engine; sorry! But I am getting better). Tenjiku seems much more rider-centric in general, which is a good thing in a game this big, I think.
 * Well, I am not sure about this either. But moving at twice the speed must be a useful asset, irrespective of board size. A Flying Dragon can catch up with and then block advance of several Pawns, when these are not too far apart. A Gold cannot even stop promotion if there is a single open file between two Pawns when it is standing before them. I have never watched a Tenjiku game, but I imagine that the LH is very dangerous because its sliding move transports it quickly to any weak spot in your defense, while a Lion can be kept at bay by a pair of Side Movers. But several other pieces are capable of igui as well, and especially the Tetrarchs seem very dangerous and the Chariot Soldiers easily promotable, with 3 forward slides. You would also have to watch out for Water Buffalo's promoting to new Fire Demons. H.G.Muller (talk) 21:17, 16 February 2016 (UTC)
 * The removal of all the promotions to gold in chu shogi creates a series of promotion chains that tenjiku shogi extends (which I guess confirms my theory that it was intended as a replacement for the rather boring and dying dai). (And hence we can here talk about things from the standard game Chu trickling upward into Dai and Tenjiku.) It also means that any piece in the starting setup can be promoted to (except the weakest ones, obviously). And I think the designers of chu also realised that slow-moving steppers are really boring on such a large board, which is why most of them are gone, and the ones that are kept are the strongest of the lot. (OTOH these pieces would be interesting on 8×8.) But that removes natural possible promotions like ferz→bishop and wazir→rook, or the ones you mention, so different pieces have to be looked to to get a rook and bishop as promotion choices. At least, that's what I think happened. Then of course, these new promotions, becoming familiar, would trickle into dai as you say, and become the bedrock that tenjiku would be built on. Double sharp (talk) 11:47, 16 February 2016 (UTC)
 * P.S. I think Betza wrote something about saving Rifle Chess involving autorifle: IIRC it would mean that a white Qd4 shooting a black Bg7 would automatically be shot back in self-defence, as the Bg7 attacks d4, resulting in mutual annihilation. However, a white Qd4 could shoot a black Nf6 with impunity. Double sharp (talk) 07:41, 11 February 2016 (UTC)
 * OK, looks like I remembered correctly, though there was more to it. Per The Classified Encyclopedia of Chess Variants (2nd ed.): In Autorifle Chess (Ralph Betza after Bill Rawlings, 1977) there is no obligation to capture, but if a capture is made all possible captures must be made with that piece (Nost-algia 211). The player can decide in which direction to shoot first, but must then continue to shoot along that line as long as there are targets before changing direction of fire. Both pieces are removed in the event of a mutual attack. K, N, P have additional powers: they may continue to shoot along the same line provided there is no gap in it. In order not to give immediate advantage to White, there is an initial rule that the first capture of a game must be orthodox. Double sharp (talk) 13:41, 1 January 2022 (UTC)
 * OTOH, this would then be an argument against the free eagle's move from tenjiku of QAD[aF] in favour of Q[maF], because the Edo-era sources simply say "奔王の動きに加えて、猫刄の動き（斜め四方向に1マス動く）を2度できる", that is "to the power of a queen is added the ability to move twice like a cat-sword (ferz)", not mentioning double-capture. Unfortunately the Japanese Wikipedia does not quote the Edo-era sources exactly for the hook movers. OTOH, we also say on the tenjiku page that there are descriptions (presumably historical) of the lion having a double king move, so that a "double ferz move" in Edo-era lingo would mean [aF]AD. This leaves the jury still out there for the hook movers. Double sharp (talk) 07:35, 19 February 2016 (UTC)

蝙蝠: hempuku vs kōmori
The latter (as we have as the primary in the article) would be somewhat odd, as shogi-variant pieces tend to use the Sino-Japanese readings (at least, looking at English Wikipedia: Japanese Wikipedia has a host of footnotes on readings that I haven't looked through yet) – not surprisingly, as the Edo-era sources are themselves written in heavily Sinicized Japanese. Double sharp (talk) 15:24, 6 March 2016 (UTC)

Modernisation redux
Is it from the guys at http://makadai.seiyakai.net/? (If so they appear to have botched all the limited-range pieces into leapers.) Double sharp (talk) 19:42, 12 March 2016 (UTC)
 * Interesting website. They don't seem to claim this is Maka Dai Dai Shogi, but apparently call it Maka Dai Shogi. If my Google translation is any good, I gather that the limited range pieces ('dancing pieces', as the translation calls them) are not just leaping pieces, but locusts: their moves indicated in blue (optionally) capture what they leap over. I suppose the Lion Dog can do igui on adjacent squares, in addition to a locust jump of 2 or 3 squares. Oh, and all the promoted 'free' pieces seem to be 'ski-sliders', skipping the first two squares before starting their slide. H.G.Muller (talk) 21:35, 28 July 2017 (UTC)
 * This seems to be interesting and distinct enough that it may require a separate section giving their new piece moves at the bottom of the article! Double sharp (talk) 16:59, 18 August 2017 (UTC)

It seems that they do claim that their set of rules is derived directly from the information given in the Edo-era sources, at least from what the researchers appear to be saying in this paper detailing their intention to program an engine for their form of Maka Dai Dai. I would note that there may not be any real modernisation involved: already if you read the footnotes to this article, you can see that there was already a great deal of variation between the sources regarding the very moves of the pieces, even for such iconic ones as the Emperor. Fairbairn in his 1980 article also notes the difficulty of translating odoru "dancing" pieces, which first appear in Tenjiku going up the size ladder. Their diagrams show two or more dots going in one direction, and it is not clear if the second square is reachable even if the first one is not, for example. (I wonder how the sources describe the Violent Ox and Flying Dragon of Dai Shogi? Do they use the term odoru?) The later Wa Shogi also has such pieces; perhaps an important piece of evidence is that they are treated by contemporary players as being limited ranging pieces. So it is not surprising that the wording of the sources for Dai Dai, Maka Dai Dai, and Tai could be so sparse as to allow for multiple interpretations of the material, even if we ignore the fact that they are inconsistent with each other. This seems to suggest that the games larger than Tenjiku were never actually played at all; otherwise they would surely have gotten some consistency in their rules, and even some of their promotions. It seems plausible that, since many people in that period who played Shogi wrote their own pieces and had to explain them to borrowers, and since the only known large boards were intended for display, that these "games" were more intended for exhibition in the "initial positions", and the rules invented later, independently, and haphazardly, perhaps by the initial author of the SRZ who must have felt that a game needed to have a set of rules.

It is also interesting that the researchers note that the "Buddhist spirit does not disappear permanently". Perhaps the names of the Teaching King and Buddhist Spirit were already there, and because of the religious motivations the idea of contagion was applied when the rule sets were created later. Maybe even some of the Chu and Dai Shogi pieces began this way, as otherwise it is very difficult to see how they were invented with no precedents. The Lion in particular is an absolutely singular innovation, and it is difficult to see where its move came from. But because the Shogi piece motif of "wedge with two characters on it" allowed for the unlimited creation of new pieces, I can imagine that at some point, some people started writing down names of pieces and asked each other in fun to come up with moves for them, creating the game of Dai Shogi. Someone must, in this historical conjecture, have written down the name of the Lion "狮子", and asked for a move. Now this is pronounced shishi, homophonous with "死死" "death death", and so it is not surprising at all that the first doublemove was invented! (This homophony was already observed in a comment at CVP, which I unfortunately cannot find now that I am looking for it; but the rest of this narrative is my reconstruction from the findings described in the articles by Fairbairn and Masukawa on the site of Jean-Louis Cazaux.) Double sharp (talk) 06:09, 29 August 2017 (UTC)
 * Ah, OK; here is that comment (by John Smith). Double sharp (talk) 06:52, 29 August 2017 (UTC)

H.G.Muller's small version of maka-dai-dai ("macadamia")

 * A: Guardian [of the Gods], R3fF
 * B: Bishop, B (or in my preferred version, Dragon horse, BW)
 * C: Copper [general], fKbW
 * D: Castle [Dragon Horse], RF
 * E: [Drunk] elephant, FfrlW
 * F: [Flying] Dragon, B2 (or in my preferred version, Bishop, B)
 * G: Gold [general], WfF
 * H: Hook Mover, maR
 * I: Snake [Go-between], fbW
 * J: Capricorner [Capricorn], maB
 * K: King, K
 * L: Lance, fR
 * M: Wrestler, B3rlW
 * N: Lion, NADaK
 * O: Kirin, FD
 * P: Pawn, fW
 * Q: Queen, Q
 * R: Rook, R
 * S: Silver [general], FfW
 * T: Tiger, FrlbW
 * U: Left chariot, fRbW[fl][br]B
 * V: Right chariot, fRbW[fr][bl]B
 * W: Wolf [Lion dog], [a3vmcpK] \ O
 * X: Phoenix, WA
 * Y: Pirate [dark spirit], f[bl]FrW
 * Z: Priest [deva], f[br]FlW
 * +A: Gold
 * +B: Gold [I would prefer dragon horse; if already dragon horse, I would prefer horned falcon]
 * +C: Free copper [free = running]
 * +D: none [I would prefer soaring eagle]
 * +E: King
 * +F: Gold [if bishop, I would prefer dragon horse]
 * +G: Free gold
 * +H: Gold [to avoid drawishness, I would prefer none]
 * +I: Free goer = slithering snake
 * +J: Gold [to avoid drawishness, I would prefer none]
 * +K: Emperor (royal U)
 * +L: Gold
 * +M: Gold
 * +N: Berserker (furious fiend)
 * +O: Unicorn [great dragon] (I would prefer lion)
 * +P: Gold
 * +Q: none
 * +R: rook
 * +S: Free silver
 * +T: Free tiger
 * +U: Gold
 * +V: Gold
 * +W: Gold (I would prefer berserker)
 * +X: Golden bird (I would prefer queen)
 * +Y: Ghost [Buddhist spirit], lion + queen
 * +Z: Saint [Teaching king], lion dog + queen

A nice attempt to save an honestly impractical game. (Though I doubt anything will work for it. I still think the demotions already in maka-dai-dai make it too drawish for its own good, so that only chu and tenjiku leave you really free to menace the opponent without fear with your strong pieces.) Double sharp (talk) 09:49, 13 March 2016 (UTC)
 * I finally finished a game of Macadamia Shogi against someone who had been playing a hand full of Maka Dai Dai games (against an even weaker opponent) recently. I was surprised to see how infrequently demotion occurs, with the rule that promotion is only mandatory when capturing a promoted piece. I had expected that most tactics would end in demotion, because the side that will be losing the exchange (i.e. not capture last) would always elect to demote out of spite, as he is going to lose the piece anyway, and demoting might scare off the remaining attacker(s). But it turns out the contageous propagation chain is often broken by a non-promoting piece participating. Queen and Dragon are relatively valuable, and so would naturally be used as last defender, or, when overpowered by yet another attacker, make it impossible to force that attacker to demote. Much tactics also involves picking off unprotected low-valued pieces that had no opportunity to promote yet (tradesing the, through counterstriking elsewhere rather than recapturing). As the battle initially mainly rages between pieces that demote, both sides will be reluctant to 'promote', and the majority of pieces stay unpromoted. Pieces protected by Pawns are relatively invulnarable, as you don't want to allow the Pawns to promote by trading them, but they are also quite immobile as they block the Pawn that 'anchors' them, and can mostly be ignored. I also experienced an unforseen dilemma: when the board population thins you usually get an opportunity to use strong pieces (which demote, or are already promoted) to annihilate remaining nuclei of resistance consisting of low-valued unpromoted pieces (e.g. remnants of Pawn chains). But doing that actually destroys your own ability to promote pieces for which this would be favorable (i.e. the generals, and Deva/Dark Spirit). It is not easy at all to promote these steppers; formally the promotion zone is the entire board, but you can only promote where there are poorly protected enemy pieces, and the opponent is not going to offer them to you on a platter. So you have to cross the board to find suitable targets to promote on. All in all it was a quite interesting game, although my opponent was rather weak (but my unfamiliarity with kanji pieces caused a few blunders on my part as well, which kept the forces balanced for a long time). H.G.Muller (talk) 20:49, 8 August 2017 (UTC)

Lion Dog
I noticed that the description of the Lion Dog here is different from that of the one in the Dai Dai article. It does mention igui, but it does not explicitly mention the 2-out-1-in possibility. The Betza description implies it, but not everyone might understand Betza notation. H.G.Muller (talk) 21:02, 8 August 2017 (UTC)
 * OK, I've replaced this description with the one from the dai-dai shogi article. Double sharp (talk) 05:06, 9 August 2017 (UTC)

Piece movement table
I personally think it would make much more sense if the "promoted" forms of the pieces were shown in the right-hand column, to the right of their unpromoted counterparts (rather than above them as is shown currently). Thoughts? 2A00:23C5:D012:2200:8C59:7EF7:B480:F871 (talk) 13:20, 9 November 2020 (UTC)

The translation of 教王 ?
I think "teaching king" is a mistake.The character "教" can mean not only "teach", "education", but also "religion" or "edification". The latter is more likely the correct meaning here. SmartAlice (talk) 13:38, 25 July 2021 (UTC)

as names of many pieces come from Buddhism. SmartAlice (talk) 13:42, 25 July 2021 (UTC)
 * The translations all come from George Hodges and The Shogi Association, IIRC. He used to sell sets of many of the large variants (I think his widow still sells them). I think they might have become relatively standard in the English world for them, in which case I would rather not change them even if they are mistakes, but would rather check with on this. Double sharp (talk) 15:34, 16 September 2021 (UTC)

Free pieces
I made the free dragon and free wolf consistent with the CVP, because these are the "free pieces" where the Edo-era sources don't agree, and there's an obvious alternative: to make them truly the free versions of the unpromoted piece, as CVP does. Similarly I removed the fA jump from the free bear, for consistency with the CVP as well, and because this is apparently the modern way to play it. Maybe, however, I should provide an alternate diagram for the modernised blind bear as well. Double sharp (talk) 14:38, 1 January 2022 (UTC)
 * Nah, that revival changes a lot more stuff. Double sharp (talk) 14:56, 1 January 2022 (UTC)

Mnemonic device piece set
Try to use the following mnemonic set for Shogi variant boards.

Here's what it will look like for Maka dai dai shogi. 301fieriw (talk) 15:43, 15 June 2023 (UTC)