Talk:Dai dai shogi

Move set confusion
A great deal of the move sets are incorrect. If the Japanese wiki is the only source it is also incorrect. Try looking at the history of chess page cited at the bottom of this wiki for correct moves per George Hodges. He got the rules DIRECTLY from authorities on the matter in Japan.

DDS is a great game... I highly recommend it!

-Shumby


 * Thanks. I did need to proof this article. The problem in that Hodges is also several steps removed from the original docs. I've asked for help on Japanese wiki and been given refs, but no direct answers with the discrepencies.


 * Japanese wiki in down right now, but here's the problems I see that I have yet to note on the page: Great elephant, great dragon, savage tiger.


 * PS. The great dragon move is too interesting to be a mere misreading. kwami 22:42, 13 April 2006 (UTC)


 * All right, the great dragon and great elephant need alt diagrams, but I don't have time for that right now. kwami 23:16, 13 April 2006 (UTC)


 * That should do it. kwami 08:04, 17 April 2006 (UTC)

Kwami

Thank you for putting in the "alt" diagrams. This certainly helps people at least make up their own minds on the game's rules. But to those reading this article who may be new to Dai Dai shogi... please consider the "English" source version, as it's been verified numerous times. (if you can find the ORIGINAL documents I'd love to see them!!!!)

To the comment on the Great Dragon's interesting move set in the Japanese wiki... I'd like to point out that "interesting" doesn't mean correct. If you'll take notice to the Japanese wiki's contradiction

http://ja.wikipedia.org/wiki/%E9%BA%92%E9%BA%9F_%28%E5%B0%86%E6%A3%8B%29#.E5.A4.A7.E5.A4.A7.E5.B0.86.E6.A3.8B.E3.83.BB.E6.91.A9.E8.A8.B6.E5.A4.A7.E5.A4.A7.E5.B0.86.E6.A3.8B.E3.83.BB.E6.B3.B0.E5.B0.86.E6.A3.8B

you'll see the promoted Kylin is also called a Great Dragon and moves as the "English" sources dictate. There cannot be two pieces on the board with the same name and identifying kanji that move differently. That makes no sense at all.

But again, I do thank you on behalf of the Western shogivar community for the update. I do hope the confusion clears in due time.

-Shumby


 * I have no idea when you posted this, but yes, that is a good argument. I'll change the article accordingly. kwami (talk) 22:56, 23 February 2009 (UTC)


 * Ah, the tai shogi description suffers from the same problem. There would seem to be a more systematic problem. Perhaps the Japanese sources are ambiguous (they are very sparsely worded), and these are two interpretations of what was meant. kwami (talk) 23:20, 23 February 2009 (UTC)

Article review
This is a great artivle! Have you considered posting it for review in the japanese wiki and here? Sj

Failed GA
I speedy failed this article for Good Article status. Based on the time of my review:, the article does not have use sources. Please read again WP:CITE and supply this article with reliable sources to support the three pillars of Wikipedia: verifiable, no original research elements and neutral point of view. The article is also very long (>100Kb) which is not appropriate per WP:MOS. The reason is that the current article looks like a manual for the game itself, rather than an encyclopaedia article. Please read again what is not a WP article. If all of these matters are resolved, then you can resubmit this article again for GA. &mdash; Indon ( reply ) &mdash; 13:30, 10 December 2006 (UTC)

Who cares about the GA. I came to this page to find precise and exhaustive information about Dai Dai Shogi. Not to admire a GA page. So, forget about prize, keep this page informative and do not hesitate to make it even longer if it is useful. Thanks Cazaux (talk) 20:05, 7 April 2009 (UTC)

Split up article.
This article needs to be split up. OneWeirdDude (talk) 21:47, 23 February 2009 (UTC)


 * Suggestions? Or just ruminating? kwami (talk) 22:02, 23 February 2009 (UTC)


 * We can do it the way the Japanese Wiki article does it: An individual article for each piece. Samboy (talk) 15:44, 12 October 2009 (UTC)


 * Would we do that for all the shogi articles, or just this one? kwami (talk) 17:01, 12 October 2009 (UTC)


 * If consensus decided it should be done, we would probably only do it for the really huge Shogi variants; anything larger than Chu Shogi. Personally, I don’t think it’s necessary or a good idea; it’s a lot more convenient to download or print all of the rules at once using the current one-variant-per-page form.  I think we’ve removed the “This article is bigger than 32k!” warning we used to have when editing articles, and WP:MOS doesn’t mention anything about maximum suggested article size, so I say keep it as-is.  Broadband is pretty universal these days. Samboy (talk) 17:55, 12 October 2009 (UTC)

some of the alternate piece moves
Japanese Wikipedia now appears to quote the original Edo-era sources. I will look through this tomorrow. Double sharp (talk) 16:02, 5 February 2016 (UTC)

All right, here we go. This task is taking longer than I had originally naïvely imagined. I don't speak the language, but I hope my knowledge of Chinese and judicious use of Google Translate will stop me from making hilarious misinterpretations:

Japanese Wikipedia gives: It seems as though the one-step move (removing its colourboundedness) is missing from the original sources in dai-dai, but is present in the Shōgi rokushu no zushiki for tai. Removing the one-step move makes this piece identical to the capricorn from maka-dai-dai and tai. This, in itself, is not a barrier, as there are similar pairs of weak step movers in those two games. But I find it difficult to believe for such powerful pieces. And I do not know of pieces that occur across these games with different names, unless you count this one. It is possible that I overlooked one in the deluge of dragons and other variously adjectived animals.
 * Long-nosed goblin
 * 『象戯図式』『諸象戯図式』『象棋六種之図式』とも、角行の動きを2度できるとなっており、縦横1マス先には進めない.
 * 『象戯図式』『諸象戯図式』には言及がなく、大大将棋と同じ動き（縦横には動けないとなっている）と考えられる. 『象棋六種之図式』では泰将棋にも記述があり、本項目の説明と同じ動きになっている.

Conclusion: Keep current move (SRZ for tai). Admittedly the argument is weak, so I'd add a statement in the table entry about the step move.

Japanese Wikipedia gives this note:
 * Old kite
 * 各古書籍では以下のような動きになっている.
 * 象戯図式 - 斜め前には何マスでも動け、縦横に2マス動ける. (Shōgi zushiki, henceforth SZ: two squares orthogonally or ranges forward diagonally. This would be fBR2. However the diagram shown by Japanese Wikipedia only gives fBrlR2, making it two squares horizontally.)
 * 諸象戯図式 - 縦横に2マス動け、斜め前に1マス動ける. (Sho shōgi zushiki, henceforth SSZ: two squares orthogonally or one square forward diagonally. This would be R2fF. This is the move chosen by the English sources, and is also the description given for the same piece in tai shogi. The two games normally have the same moves.)
 * 象棋六種之図式 - 斜め前と斜め後ろに2マス、縦横に1マス動ける. (Shōgi rokushu no zushiki, henceforth SRZ: one square orthogonally or two squares diagonally. This would be WB2.)

The weight of the evidence seems to suggest R2fF as the most plausible correct move, given its consistency with tai shogi.

A further wrinkle is given with regard to tai shogi:
 * 『象戯図式』と『諸象戯図式』には解説がなく大大将棋と同じ動きであると推定されるが、『象棋六種之図式』では斜めに2マス、横に1マス動ける（大局将棋の説明と同じ動き）となっている.

So, while the SZ and SSZ simply say the move is the same as in dai-dai, the SRZ gives the move as B2rlW, the same as taikyoku shogi. But that game usually has different moves, and this would tally with how the SRZ generally disagrees with the other two sources (which usually concur with each other) on maka-dai-dai. This suggests that the SRZ should be given lesser weight than the other two sources.

Conclusion: Follow the English sources (SSZ).

I therefore feel free to ignore the SRZ's divergent reading on the hook mover: It's not even consistent with itself, and in any case one of the possibilities listed tallies exactly with the SZ and SSZ.
 * Hook mover
 * 『象棋六種之図式』の大大将棋の説明には、釣行（鉤行）の説明が2度書かれている. ひとつは飛車の動きを2度できるものだが、もうひとつ（「已下成馬」と書かれたあとの記述）は縦と斜めに何マスでも動け、後ろと斜めには他の駒を2個まで飛び越えられるとしている.

Conclusion: Keep current move (SZ, SSZ, SRZ*).


 * Poisonous snake
 * 『象戯図式』と『諸象戯図式』では、前と斜め後ろに2マス、他の駒を越えて進め、横に1マス進めるとしている. 『象棋六種之図式』では縦横に2マス、斜めに1マス進める動きとなっている. (Referring to dai-dai. According to the SZ and SSZ, it jumps to the second square directly forward or diagonally backward, or steps one square to the side: rlWfDbA. This is the move given in English sources. But according to the SRZ, it moves two square orthogonally, and one square diagonally forward: R2fF. That is the move of the old kite in the SSZ: is there some accidental transposition at work?)
 * 『象戯図式』『諸象戯図式』では動きの解説が省略され、大大将棋と同じと推定される. 『象棋六種之図式』では、斜め前に2マス、後ろに1マス進めるとなっている. (Referring to tai. The SZ and SSZ simply refer to their description for dai-dai. The SRZ gives fB2bW. Somehow Japanese Wikipedia's diagram is an odd mix of the two SRZ moves?)

Conclusion: Follow the English sources (SZ, SSZ).

Oh God. Here is the one where the English sources seem to have fouled up, as their chosen move (Q3) makes the move of the teaching king in maka-dai-dai a ridiculous pleonasm (Q + Q3 = Q). The two games normally have the same moves. On the other hand, it does have to be said that this move makes the lion dog's promotion actually desirable, when otherwise it becomes the only demotion in dai-dai.
 * Lion dog

Although there is one interesting note: The SZ and SSZ do not give a promotion for this piece. Given that, one neat way to sidestep this problem would be to say that it does not promote at all.
 * 『象戯図式』と『諸象戯図式』には、成駒についての記述が欠落している.

Moving on, we now see how the Western understanding came about: The SZ and SSZ do indeed unanimously describe the move as Q3 in dai-dai: three steps in all directions, without any indications about the lion power. (In the SRZ, it is two steps only in the forward directions.) However, they also equally unanimously (this time joined by the SRZ) describe the move with lion power for maka-dai-dai (I infer this from the fact that Japanese Wikipedia does not see fit to provide a note in that case.) Hence I would interpret the Q3 description as an incomplete one. It is indeed accurate, as even with lion power it does move to the Q3 squares. It just can do additional things along the way.
 * 『象戯図式』『諸象戯図式』『象棋六種之図式』では、全方向に3マスまで動けるとなっており（象棋六種之図式は前方のみ2マス）、他の駒を飛び越えたり、戻ったりする動きは記述されていない.

Indeed the two games usually have the same moves. I therefore assume that the TSA looked at dai-dai first and extrapolated its rules over, but in this case, it seems that maka-dai-dai is the correctly described variant. Note that one cannot reasonably assume that the moves are different in dai-dai to satisfy its otherwise rather silly promotion, because the Q3 move is given in the SZ and SSZ, the sources that do not give it a promotion at all.

It is nevertheless true that the Edo-era sources are not cited for maka-dai-dai in Japanese Wikipedia. Of course, this could just be because all are in agreement. What is true is that the Western move must be wrong.

Conclusion: Keep current move (SZ, SSZ, SRZ for maka-dai-dai).

What of its demotion, nonexistent in the SZ and SSZ?

In the SRZ, it keeps the move of the lion dog, but appears to have gained the power to jump over all other pieces except in the forward diagonal directions.
 * Great elephant
 * 『象棋六種之図式』では、縦横と斜め後ろには3個まで駒を飛び越えて何マスでも動け、斜め前に2マス動けるとなっている.

But then where did the move currently listed come from? Is this piece really in the SZ and SSZ without a piece promoting into it? But then wouldn't the Japanese Wikipedia state that?

Conclusion: I am really uncertain. Given the general unreliability of the SRZ (at least that's the way I see it), I would imagine this piece to be a spurious inclusion. Deleting it means that all promotions become beneficial. Also, it seems to be the case that pieces with special promotions in dai-dai keep that promotion in maka-dai-dai if they exist in that game too. Except this one. Something is rotten about this piece!
 * Eric Silverman ( https://drericsilverman.com/ ) appears to have unearthed some new info on this, by directly consulting the SRZ in a Japanese library. He presents compelling evidence that what the Japanese Wikipedia indicates as unlimited ranging moves are in fact TK moves: slide + 3-step radial Lion power. Eric's inerpretation of the Japanese Wikipedia is that these slides can jump over up to 3 pieces, btw, which would be a mis-interpretation of the transverse stripes in the SRZ's movement diagram. H.G.Muller (talk) 12:19, 5 June 2021 (UTC)
 * Yes, I agree with Silverman's reading of the Japanese Wikipedia. The SRZ gives an interesting reading, but I am somewhat hesitant to adopt it because it usually disagrees with the SZ and SSZ. (And also because if the great elephant does not exist, then pieces that exist and promote in both dai-dai and maka-dai-dai always promote the same way.) But I've added a diagram and a note. Double sharp (talk) 15:16, 1 October 2021 (UTC)

Apparently the SZ has the diagram and description not matching; the description shows the solitary single step to the right, but the diagram to the left. The SRZ gives the taikyoku move for dai-dai, but in tai it has simply become a wazir(!?).
 * Eastern barbarian
 * 『象戯図式』の図では縦に2マス、左に1マス進める表記になっているが、解説は縦に2マス、右に1マスの動きである. 『象棋六種之図式』の大大将棋の解説では前後に2マス、左右および斜め前に1マス進める（大局将棋の説明と同じ動き）. 同書の泰将棋では、縦横に1マス動けるとなっている.

Conclusion: Unfortunately, the SSZ reading is not given. But the symmetry argument and the SRZ statemtent for dai-dai does indeed sway me in the direction of the English sources (SRZ taikyoku).

The Japanese Wikipedia only shows the forward diagonals, as stated. But the SRZ gives only 3 squares forward or diagonally (all?!) and 5 to the back. In tai shogi, the SZ and SSZ refer us back to their description for dai-dai, while the SRZ gives it unlimited range sideways and backward, the diagonally forward moves have become 2 squares only.
 * Free demon
 * 『象棋六種之図式』では、横と斜めにどこまでも動け、前に3マス、後ろに5マス動けるとなっている.
 * 『象戯図式』『諸象戯図式』には言及がなく、大大将棋と同じ動きであると推定される. 『象棋六種之図式』では、横と後ろに何マスでも、斜め前に2マス動けるとなっている.

In taikyoku, we finally get the move from the English sources. While taikyoku usually differs, it does not always (cases in point are some moves from the SRZ that do not follow the other games in the dai-dai/maka-dai-dai/tai triangle, but instead taikyoku; see below), so this has to be taken as a possibility.

The SRZ allows ranging on all four diagonals, but restricts the horizontal move to only 5 squares. In tai shogi, the SZ and SSZ refer us back to their description for dai-dai, while the SRZ gives the same moves as the free demon.
 * Free dream-eater
 * 『象棋六種之図式』では、斜めに何マスでも、横に5マス動けるとなっている.
 * 『象戯図式』『諸象戯図式』には言及がなく、大大将棋と同じ動きであると推定される. 『象棋六種之図式』では、横と後ろに何マスでも、斜め前に2マス動けるとなっている.

In taikyoku, we finally get the move from the English sources.

Conclusion: Nothing seems to support including the backward diagonals on these two pieces in dai-dai and tai, except that it makes these demotions instead of promotions (though far less egregious than lion dog to great elephant). Perhaps that is an argument for keeping the current status quo, with both mentioned (taikyoku).

The SRZ is the one that removes the directly-backward move in dai-dai, but it also removes the forward right diagonal in tai. It looks like the authentic move is simply that of a non-royal king, that cannot serve as a king-replacement.
 * Neighbouring king
 * 『象棋六種之図式』（「雑藝叢書」による）では、大大将棋では後ろ以外の7方向に1マス動ける（本項目の大局将棋での動きや、醉象と同じ）、泰将棋では右斜め前と後ろ以外の6方向に1マス動けるとされている.

Conclusion: Non-royal king move (English sources).

BTW, the correct kanji for this name is actually 仙 according to Japanese Wikipedia.
 * Wizard stork

We have a bibliographical nightmare here: The SZ gives BbW, the SSZ gives fQbW, and the SRZ gives fBfbR.
 * 『象戯図式』ではこの動きであるが、『諸象戯図式』では前と斜めに何マスでも動け、後ろに1マス動ける、『象棋六種之図式』での動きは前後および斜め前に何マスでも動けるとなっている.

But for maka-dai-dai: Somehow all three Edo-era sources give fBrlbR, while the SZ and SSZ then proceed to shoot themselves in the foot by saying that the Chinese cock promotes into this piece, which somehow has started moving like a free Chinese cock. One must be wrong.
 * 『象戯図式』『諸象戯図式』『象棋六種之図式』とも、斜め前、横、後ろに何マスでも動けるとなっている（象戯図式、諸象戯図式の淮鶏の動く方向に何マスでも動ける）.

Conclusion: the SZ's move of BbW is a plausible corrupt form of BfRbW, and makes for greater symmetry. Keep the status quo.

Finally! The end of the madness! The SZ and SSZ give the move from the English sources, while the SRZ gives the taikyoku move of fR.
 * Savage tiger
 * 『象戯図式』と『諸象戯図式』では、縦に2マス、斜め前に1マス動けるとなっている. 『象棋六種之図式』では、前に何マスでも動ける（香車および本項目の大局将棋の解説と同じ動き）になっている.

Conclusion: Keep move from English sources (SZ, SSZ).

Of course it is not inconceivable that the really huge games (dai-dai, maka-dai-dai, and tai) would perhaps have local rule variations. It is not as if changing the powers of such weak pieces would make much of a difference in a game with hook movers running the show. Nevertheless, perhaps this sheds some more light on which moves one should adopt if one is looking for an "urtext" rulebook, though this is perhaps an unattainable fantasy in this case.

I really doubt I will do this for tai shogi, as it is insanely big and practically not playable. The initial position is so congested, and it ends up as a potpourri of all the pieces ever invented by Buddhist monks (and just how much time were they devoting to creating huge shogi variants?) that lives up to the original etymology of that word. I wouldn't touch taikyoku with a ten-foot pole either.

The games up to maka-dai-dai deserve to be noted. I will note that maka-dai-dai is exactly where H. G. Muller stopped in his survey of the large shogi variants for The Chess Variant Pages. But dai-dai is turning into a bibliographical nightmare without resolution. Double sharp (talk) 16:14, 6 February 2016 (UTC)

TL;DR summary: That's all, I think. Double sharp (talk) 16:22, 6 February 2016 (UTC)
 * Add a note stating that we are not sure about the step move for the long-nosed goblin.
 * Banish the Japanese-Wikipedia moves for the old kite and poisonous snake as not supported by anyone.
 * We can probably get away with removing the Western move for the lion dog in both dai-dai and maka-dai-dai, except as a footnote.
 * We may even be able to kill off the great elephant, as it does not appear as a promotion in the two more reliable Edo-era sources.
 * Make the all-direction move for the neighbouring king primary, relegating the other one to a footnote.
 * Stop equivocating about the fB2 move of the savage tiger, and replace it with fF.

Lion dog
This sounds like the best we can do at the moment. The information that the Lion Dog is not said to promote in SZ and SSZ is new to me, and puts the whole issue in a drastically different light. I have always been convinced thet the Lion Dog must have had something more than just Q3. Based on it being mentioned in the TK description for a reason, as well as that the initial setups (as well as the way Deva and DS, or the Barbarians promote) obviously treat it as a counterpart to the Lion. The Japanese Wikipedia proves to be a very unreliable source, however. So drawing the conclusion that all the Edo sources agree on the Lion power of the Lion Dog in Maka Dai Dai just because the Japanese Wikipedia does not mention anything about it makes me feel very uneasy. I would really like to see the original wording in these Edo sources that describe the Lion Dog, to see what they could have meant. One thing that still worries me is that the Edo sources seem to describe Furious Fiend as Lion + Lion Dog, while the Western consensus has always been that it is Lion + Q3. And there also seems to be consensus that LD promotes to Gold in Maka Dai Dai. Which makes the LD-vs-Ln promotion a huge asymmetry in an otherwise very symmetric game. If the Lion gains the Lion-Dog move on promotion, I would have expected the Lion-Dog to gain the Lion move on promotion. But that would make them promote to the same piece, which also seems to be a no-no. If Lion would gain Q3 as a sort of 'partial' LD move, I would have expected LD to gain a two-step area move (maK). But there doesn't seem to be even a hint that this could have been the case. In the interpretation with radial Lion Power the LD is an awfully strong piece, as it can pick off pieces from a distance of two by 'quasi-igui', retreating with the 3rd step. Alternative moves that might work better from the perspective of the game could be Q3 + two-step radial Lion power (as in the Chu-Shogi HF and SE), or allowing only outward steps. H.G.Muller (talk) 09:32, 7 February 2016 (UTC)
 * The lion dog appears not to promote either in tai shogi, according to the SZ and SSZ. That is what our article says, and it would make sense given that pieces which appear in a smaller shogi variant are not explained again when they recur in a larger variant in the Edo-period sources. (So the lion dog appears in dai-dai, and thus it is not explained again in tai.) Given that the SRZ predates the SZ and SSZ sources by two centuries, perhaps this is a later change. The Western consensus of Lion + Q3 may be because there was a desire to normalise the lion dog to Q3, which is how the SZ and SSZ describe it in dai-dai. But it may not be the same in maka-dai-dai, and that is where Japanese Wikipedia aggravatingly fails to summarise the contents of those sources. (Against this would be the fact that it is then the only piece which is not the same in these two games, apart from the promoted wizard stork, which could be a simple omission of the forward rook move. The difference between Q3 and the Japanese Wikipedia lion dog is so great that it is nearly impossible to attribute it to a scribal error, though.) Double sharp (talk) 05:06, 8 February 2016 (UTC)
 * The Tai article here actually says that LD promotes to Great Elephant OR Gold, in the piece table. Tai Shogi seems to be a merger of Dai Dai and Maka Dai Dai Shogi, with hardly any new pieces. Pieces from Dai Dai that do not promote there in general will promote to Gold in Tai (if they occur in the initial setup). Pieces that occurred both in Dai Dai and Maka Dai Dai (of which LD is one) keep their promotion from the latter, however. In both cases LD would promote to Gold in Tai. Anyway, we must beware of circular reasoning, which is what we would get if we use the "this would make it the only piece that is different in ... and ..." too often.83.163.204.254 (talk) 10:06, 8 February 2016 (UTC)
 * You're right: here's the exact quote. "In Evans' software, this promotes to a great elephant. However, while Japanese Wikipedia states that this promotion occurs in other large-board shogi variants, it says that it demote to gold in tai shogi." I could believe that, as the Edo-era descriptions of tai shogi pieces that occur in smaller games tend to be "refer to [smaller game]". Your logic is impeccable. I agree that once you have multiple exceptions the argument loses all its force. If the LD was the only exception, it would however be worth noting, I think. Double sharp (talk) 12:13, 8 February 2016 (UTC)
 * I do note that we do not see area moves outside tenjiku, which seems to be a later invention, as it does not appear in the SRZ (while sho, chu, dai, dai-dai, maka-dai-dai, and tai do). So maybe this possibility would not have occurred to the designers.
 * One thing I do notice is that with maka-dai-dai, we do not have as much of a bibliographical nightmare. The SZ and SSZ almost always agree with each other on that game, with the SRZ presumably providing an earlier version. The exceptions are the deva and dark spirit (we still know nothing about the lion-dog descriptions). It seems in general as though more care was taken over the rules of maka-dai-dai than dai-dai or tai (presumably because it has more interesting rules).
 * The reason I tend to trust the Japanese-Wikipedia lion dog description is that it is such a singular move, that takes so long to describe, that I cannot see how it could have been anything other than deliberate. I can understand an accidental conflation of the moves from two sources; I can understand an omission of one direction of a boring partial-queen piece. But mangling Q3 into the lion dog move cannot possibly be accidental. If this is not the historical move, then we need to track down the original vandal and get him or her to design more chess variants. I notice Japanese Wikipedia tends not to mention sources that agree with them: usually they just mention that the SRZ says something different. When the SZ and SSZ are invoked, they're usually disagreeing with each other, or with the presented rule. (Although it seems to me that keeping the Japanese-Wikipedia lion dog, but not letting it promote as in the SZ and SSZ, would work throughout all the variants.) Double sharp (talk) 05:00, 9 February 2016 (UTC)
 * Further supporting the idea that maka-dai-dai was played appreciably more than the other two huge variants (dai-dai and tai) is the presence of a mnemonic poem for the initial setup in the SRZ, given at Maka dai dai shogi. Double sharp (talk) 05:22, 9 February 2016 (UTC)
 * What worries me is that I read somewhere that these Edo texts are basically unobtainable, because the book in which they were published is sold out and out of print. Surely this must also strongly hinder the contributors to the Japanese Wikipedia, and undermine the reliability of the latter. I would feel much more at ease if I knew that all Japanese Shogi afficionados had access to those works, and the only problem is that we cannot read them... That being said, I have never seen it contested that in Maka Dai Dai the Lion Dog would demote to Gold. The Japanese Wikipedia, and even the 'modern' rules, put no question marks to the LD's Lion power, but seem equally certain on the demotion to Gold. Of course Gold is even a demotion from Q3, but most Maka Dai Dai pieces in the Q3 class suffer such a demotion, and true promotion is mainly rserved for steppers. It is interesting to note that the TSA mentioned (as a 'speculation' fueled by the description as Q + LD) the possibility the TK had Lion power, but then rejected 3-step 'omni-directional' Lion power (on the basis it would make the TK insanely strong), and never considered purely radial Lion power. (The piece-help of Steve Evans' ShogiVar still attests to this.) Strangely enough the TSA description then continues by stating it was much more likely the TK moved the same as BS, i.e. Q+Lion, and that this would not have any repercussions for the move of the LD, on which they were adamant it did not have Lion power. (But then again, if they meant with this it did not have omni-directional Lion power, they would be right about that!) And yes, I think Maka Dai Dai as I understand the rules deserves to be played more than Dai Dai, so it probably was. It seems to contain many very interesting and unique ideas. H.G.Muller (talk) 20:56, 10 February 2016 (UTC)
 * From ja:ノート:泰将棋 (the Japanese Wikipedia talk page for tai shogi), after a question from Kwamikagami on contradicting promotions: "諸象戯図式や象棋六種之図式から正確なところを知ることができるように思われますが、両方とも江戸時代の書籍で、現在入手可能かどうか不明です. 『象棋六種之図式』を複写した『雑藝叢書』がありますが、絶版になっています（図書館では閲覧できる可能性あり）. 『ものと人間の文化史 23 将棋』(ISBN 4588202316) は現在でも入手可能なので、こちらの記述を参考にするのがよいかもしれません. " That is (I think), "I would expect that you could obtain accurate info in the SZ or SRZ, but both are Edo-era publications, and it's unclear where they could be obtained now. The SRZ was reprinted in the 雑藝叢書 'Miscellaneous Monographs' [by 国書刊行会 in 1915], but that is out of print (though it may be possible to find in a library). But ものと人間の文化史 23 将棋 'A Cultural History of People and Things, 23: Shogi' (ISBN 4588202316) is still available today, and it would be useful to consult its descriptions." The presence of an ISBN seems to improve matters, and indeed Edo-era shogi sources lists it as giving the contents of the SZ and SRZ (though not the SSZ). (Additionally, the SZ alone was republished in 1909 by 松浦大六筆写, but this is probably out of print as well.) Double sharp (talk) 14:57, 11 February 2016 (UTC)

P.S. Additionally, the lion-dog chain of reasoning, taking maka-dai-dai as the primary instead, would result in giving furious fiend = lion + lion dog (as in maka-dai-dai). Since I know of no other cases where a piece moves differently in dai-dai and maka-dai-dai (except the wizard stork, which could be explained as a simple omission), it seems that the right rule (if there can be one) must be: Double sharp (talk) 09:15, 7 February 2016 (UTC)
 * lion dog = the three-square single-direction lion mover;
 * furious fiend = lion + lion dog;
 * teaching king = queen + lion dog.

I have incorporated this stuff into the article. Double sharp (talk) 12:07, 1 January 2022 (UTC)
 * And some more footnotes, mostly about wildly divergent SRZ moves. Hopefully I have them all right now. Double sharp (talk) 12:46, 2 January 2022 (UTC)

Piece names
Incidentally, I do not see why we should have to follow the TSA translation of 奔王 as "Free King". Yes, that's what it means literally. But the piece is universally known in chess as the "Queen", which is how The Chess Variant Pages translates it. The Japanese names for the rook and bishop are not the same as the Western names in meaning. Given that CVP is behind this, it should be okay to change this. Double sharp (talk) 09:15, 7 February 2016 (UTC)
 * This is probably mostly my doing, but based on exactly the same reasoning. I general names of Chess pieces in different languages are not derived from each other by literal translation anyway. BTW, I noticed that most TSA translations suffer from "poetic exaggeration". I don't read Japanese, but by know I can recognize the first kanji of a Rook as that for 'flying' (as in Flying Chariot). And that character is quite abundant on Shogi pieces, but leads to TSA translations like "Soaring Eagle", "Swooping Owl" etc. H.G.Muller (talk) 09:41, 7 February 2016 (UTC)
 * Yes. A particularly bad example is in tenjiku, where 角鷹 kakuō becomes "Horned Falcon" but 獅鷹 shiō, shitaka becomes "Lion Hawk", totally obscuring the symmetry between this pair and 飛鷲 hijū "Soaring Eagle" and 奔鷲 honjū "Free Eagle". (Obviously the other characters come from 獅子 shishi "Lion" and 奔王 honnō "Queen" respectively.) Given that we use "Kirin" instead of "Kylin" (because just about nobody uses that borrowing into English today, and so we simply use the Japanese transliteration for a Chinese mythical creature), it should be totally fine to use "Queen" instead of "Free King". I would keep "Knight", although it doesn't move like the Western one, because fooling around with the pieces already in standard shogi is not advisable. Meanwhile, the 奔獏 honbaku "Free Dream-Eater" from dai-dai should probably become a "Free Baku", since we're not translating "Kirin" into "Unicorn". TSA, in its typical ineptness, came up with "Free Tapir" – not realising that, while that is a current meaning of the character 獏 or 貘, this was a case of borrowing the name mythical creature for a real animal, and since tapirs are unknown in Japan that cannot be what the anonymous Buddhist monks have meant. By that logic "Kirin" should have been equally ineptly translated as "Giraffe", as that is indeed a current meaning of 麒麟 (ja:キリン) in Japanese. Double sharp (talk) 14:58, 7 February 2016 (UTC)
 * For tenjiku shogi (the largest one worth the trouble IMHO), the TSA names are not that bad. This trickles down into chu shogi, which really has quite good names (except that I'd use "Queen" instead of "Free King", like you do). The 四天王 shitennō are more often translated as "Four Heavenly Kings", but I prefer "heavenly tetrarch" (but no plural; the point is that you can have up to four), as "king" has connotations in chess.
 * Here is the problem with translating names for tenjiku and chu. The names 角鷹 and 飛鷲 are terrific puns (and the TSA translations miss this, although it's not completely their fault). Ever noticed that 角 and 飛 are in the names for the bishop and the rook, and appear here too? Now look at the promotion chains upwards:
 * 猛豹 (leopard) → 角行 (bishop; angle-mover) → 龍馬 (dragon horse) → 角鷹 (angled hawk(?) – yes, that's what 鷹 means: there's no native Japanese word for falcon. The Chinese word for falcon is 隼) → 角将 (bishop general) → 副将 (vice general). 角 can mean either "angle" or "horn", so this may explain the odd biology that otherwise seems to be at work here. So maybe this is really an angled or angling hawk.
 * 歩兵 (pawn) → 金将 (gold general) → 飛車 (rook) → 龍王 (dragon king) → 飛鷲 (flying eagle) → 飛将 (rook general) → 大将 (great general).
 * And now we're stuck having to render 飛 as both "rook" and "flying", while 角 has to be both "bishop" and "angled". This sort of thing is where I would introduce translator's notes.
 * If we're being purists about not translating Japanese mythological creatures, by the way, 鳳凰 has to be "hō-ō". To those going to complain that this would be unrecognisable: don't worry, there's a Pokémon called that (250), and therefore everyone obviously knows it. (My tongue is firmly in my cheek.) Since the Chinese names are more well-known, though, perhaps we should be using "fenghuang" and "qilin". Also, the fire demon would be a fire oni.
 * That's all I have to say about chu and tenjiku names, where the TSA didn't mess up as badly as I thought they might have for a second. Double sharp (talk) 15:18, 7 February 2016 (UTC)
 * (Although dai-dai was served far worse. How did 天狗 tengu, literally "heavenly dog", become a "long-nosed goblin"?! This will take a while to go through.) Double sharp (talk) 15:20, 7 February 2016 (UTC)
 * Okay, I'm going through the large variants and replacing "free king" with "queen" and "crown prince" with simply "prince", following your CVP pages. Unfortunately, I don't think we can change the other translations unless someone like you gets behind them in your articles. Double sharp (talk) 03:58, 8 February 2016 (UTC)
 * ✅ Double sharp (talk) 05:06, 8 February 2016 (UTC)
 * We probably should not rename pieces to names other than the TSA names if the result is not already in common use (like 'Queen'). 'Baku' is hardly an improvement over 'Dream-Eater'. Creatures from Buddhist Mythology are just not very convenient for western use. It would be better to completely rename these pieces to something that is slightly mnemonic for their move. (The 'Lance' being the prime example and precedent for this. E.g. in the Chu-Shogi description on the XBoard website I use 'Narrow Queen' for 'Flying Ox' and 'Sleeping Queen' for 'Free Boar', representing them by a narrowed version of the regular Queen symbol, rotated 90 degrees in the latter case. In my shrunken version 'Macadamia Shogi' I renamed pieces Lion Dog->Wolf, Deva->Priest, Dark Spirit->Pirate, Teaching King->Saint and Buddhist Spirit->Ghost.) Simplifying by omitting a redundant adjective would be advisable, though. E.g. within the context of Chu Shogi I always refer to the pieces as 'Leopard', 'Tiger', 'Elephant', 'Falcon' or 'Eagle', (just like western players of regular Shogi usually talk about 'Horse' and 'Dragon'). Doing this in a universal way for all Shogi variants simultaneously (e.g. that it is unacceptable that a 'Tiger' moves differently in DD ('Savage Tiger') and MDD/Chu ('Blind Tiger'), although we do not seem to have such qualms w.r.t. 'Pawn' referring to a different piece in Chess and Shogi) greatly complicates this, of course. (But is sort of forced on us by Tai Shogi trying to feature all Shogi pieces ever invented in one game...)H.G.Muller (talk) 10:08, 8 February 2016 (UTC)
 * I don't know when the mnemonics should start overpowering the historical names, though. Consider 鳳凰 as a name: the closest English equivalent is "phoenix". Should we go for that, or Betza's name "waffle" for the same piece, evoking its composition as wazir + alfil? I think that in the context of describing the historical variants, as they were at their last stage of evolution (represented in the SZ and SSZ), we should strive to keep the original names as intact as we possibly can in this new linguistic and cultural context. (Of course, I have nothing against renaming the pieces at will for new variants. In fact, I might give your variants a look! I'll certainly add them to the shogi variant list on Wikipedia.) Thus "baku" would be translated as what that mythical creature does: it devours dreams and nightmares, so let it be "dream-eater" (instead of TSA's inaccurate "tapir"). But what of "kirin", which would draw a blank from most English speakers, and yet is for the piece? Translating the mythical creature to "unicorn" has been done, but is odd as (1) kirins can have two horns and (2) unicorns are called something else in Chinese. And what of "tengu", which I have the impression more people have heard of than the kirin, baku, and komainu (lion dog), and yet was translated (although given its multifaceted cultural history, it's hard to translate)? There is no consistency here that I can see, so it seems that the only safe solution is following the TSA's names, keeping the original Japanese names around for those who can read and use them. (Really, this is a large obstacle to the game gaining traction in the West. Fortunately chu and tenjiku don't suffer from it anywhere near as much as the larger ones.) Double sharp (talk) 12:13, 8 February 2016 (UTC)
 * I certainly prefer 'Phoenix' over 'Waffle', also because of historic precedence. Note that 'Vnicorno' seems to have been the name of a piece in Grande Acedrex that from the description can be identified as a Rhinoceros with certainty. So how about 'Rhino' as translation for 'Kirin'? And why not 'Viper' or 'Cobra' instead of 'Poisonous Snake'? Or 'Bull' instead of 'Violent Ox'? (It seems the second kanji simply means 'Cow', judging the reaction from a novice Japanese Chu player on 81Dojo!) Most pieces could drop part of their name without creating any ambiguity: Wizard (Stork), (Mountain) Witch, (Enchanted) Badger, (Enchanted) Fox, (Old) Rat, (Buddhist) Devil, Guardian (of the Gods). Only the abundance of Elephants and Dragons is a real problem. — Preceding unsigned comment added by H.G.Muller (talk • contribs) 14:28, 8 February 2016 (UTC)
 * I think the number of steps necessary to get from "Kirin" to "Rhino" is too high (2) for me to feel comfortable about it. At least the unicorn has a slightly more similar place in Western folklore to that of the kirin in the folklore of the Sinosphere. And we do not really know which poisonous snake was intended (perhaps it's just deliberately vague). And yes, 牛 simply means "cow", although it can be further specified to get something like "ox" (去勢牛) or "bull" (牡牛). Another case of TSA poetic exaggeration, I think. I would personally favour translations to be as literal as possible, except for pieces that have well-established standard English names. So while I might be willing to replace "angry boar" and "cat sword" with "wazir" and "ferz" respectively, I can't say that about any other shogi variant pieces except the bishop, rook, queen, and king. (In fact, AFAIK, the only fairy pieces that seem to have any single dominating name seem to be the W, F, D, and A.) Double sharp (talk) 15:41, 8 February 2016 (UTC)
 * (Can't believe I forgot the nightrider – NN – and grasshopper – gQ. There are also little-used rarities like the rose – qN – that for that reason only have one name: they are not used often because they're difficult to invent and use.) Double sharp (talk) 14:36, 22 February 2016 (UTC)

History
I am increasingly suspicious that the games larger than dai were never actually played to any significant extent, and that of the larger-than-dai games, only tenjiku deserves a revival (and it may in fact have been conceived as a replacement for dai in the sho-chu-dai triangle). Allow me to quote Jean-Louis Cazaux's website. The (Shinsen) Yūgaku ōrai from c.1350 writes "Games of Dai Shogi and Chu Shogi, kemari, wind or string instrument, renku, classic poems, composition, etc. are subject to gambling for many people." (translation by Kimura). An entry from 1424 from the diary of Hanasakai Sandaikai writes "Played shogi with Sedate Motoyuki .... I received a Free King [queen] handicap". This could either be chu or dai shogi. Cazaux also writes "Many diary references of the time seem to maintain the distinction between [dai] shogi, middle [chu] shogi and little [shogi] shogi and indeed many of them imply that little shogi was generally regarded as merely a boys' game." Also in the SRZ, though sho and chu are mentioned, knowledge of them is taken for granted as they are not describe (so that it really only describes dai, dai-dai, maka-dai-dai, and tai). Some 14th- and 15th-century diaries mention playing chu shogi often, so presumably during this time dai was dying out in favour of chu (hence why the 1443 SRZ had to explain the former but not the latter). Though it still apparently retained prestige, as Ise Sayadori's Sogo Ozoshi writes "When you face your shogi opponent open the box containing the pieces and put the pieces on the board...to finish setting them up too quickly is a breach of politeness. It is also impolite to be too slow and to make the opponent wait. Practice therefore so that you can set up the pieces in just the right time. This applies also to both middle shogi and little shogi." (so unqualified "shogi" still could have meant dai shogi at this time). The true death of dai-shogi, relegating it to a curiosity like dai-dai, maka-dai-dai, and tai apparently always were, must have come by the beginning of the 17th century, with the following statistics from Minase Kanenari, maker of shogi pieces:

Evidently the disappointing numbers for the four large variants are from a few diehard Edo-period otakus. (Although since I post incessantly about these large variants I cannot judge them at all.)
 * Maybe not even that, since Masukawa's 2004 paper says that the large sets were intended only for display, not for actual playing. (Just as well since it does not appear that there were ever standard rules for the variants larger than tenjiku.) Double sharp (talk) 14:07, 19 February 2016 (UTC)
 * In light of all this, I am tempted to agree with Kaufman's conclusion that the games larger than chu (except tenjiku) are jokes that were never meant to be seriously played. Dai is of course an exception: as the first really huge variant it makes sense that it is not that fun to play, and was later streamlined into chu. That was not a joke, certainly; but chu supersedes it as it is the final evolution of dai. Tenjiku is actually well-designed, and based on the Japanese Wikipedia giving standard piece abbreviations, quite possibly has a history that is now lost to us. (Chu shogi has history too; much of it is lost to us now.) Double sharp (talk) 15:47, 29 February 2016 (UTC)

Another thing these numbers demonstrate is that chu was also declining, like dai. While before it had apparently been more prestigious and more often played than sho (for why doesn't the Yūgaku ōrai mention sho?), it was now still popular, but declining behind the less complex sho. A true blow to the already waning popularity of chu came by 1636 in the SZ, where the 2nd Meijin Soko Ohashi presented the current rules of shogi with drops. One other thing we know from this is that tenjiku is mentioned only now, and hence could have been a recent invention in the early 17th century, when chu was remembered but not dai. Perhaps this was an attempt to recreate the three-class system of shogi variants, but having a more interesting highest level, based more on chu than on the superseded dai. Another book with the same title (or is it the same one? is this a specialist appendix, showing that this was by now for a specialist audience of Edo-period otakus only?) mentioned the large variants (sho, chu, dai, tenjiku, dai-dai, maka-dai-dai, and tai), and made reference to several others (wa, ko, Tang, taikyoku, 7-person xiangqi). The incredibly well-designed starting position of tenjiku surely implies that some playtesting must have gone on to avoid instant wins. But the curtain was evidently fast going down for the large variants. Later the 1694 SSZ covered all of these again, plus wa, and apparently attributed many of them to Buddhist monks.

In 1663 the 3rd Meijin Itō Sōkan wrote the Chu Shogi zushiki (CSZ), an exposé of chu shogi alone (evidently it wasn't dead yet)! A few years later he published a book of chu shogi directmates. The Kokon Shogi Zui from the 17th century writes once more about the large variants, and that is the last we ever hear of them, save for Oyama Yasuharu's apparently final revival efforts for chu in the 20th century. (Though it was still apparently played reasonably often in Kyoto till the mid-20th-century. I did in fact check this out completely unscientifically by asking a few elderly people in Kyoto at tourist help desks in 2013 or 2014, and yes, it is still remembered, but gone.) Double sharp (talk) 09:15, 7 February 2016 (UTC)


 * BTW, since the SRZ dates from the 15th century and the other two (SZ and SSZ) from the 17th, is it not possible that the discrepancies are because the large games (dai-dai, maka-dai-dai, and tai) evolved slightly within two centuries? This would not require many players, of course; it would just require a touching up of the rules by those Buddhist monks. But if so, then it seems like the SZ and SSZ should be taken as primary, as they are the final attestations of the rules of these games, preserved as they either died completely (dai, and chu outside Kyoto) or never lived (dai-dai, maka-dai-dai, tai, and sadly tenjiku, which really does not deserve to lie here in this pair of parentheses). Double sharp (talk) 15:41, 7 February 2016 (UTC)
 * P.S. After having tried both chu and dai on SDIN (winning a game of dai in 458 plies): even dai isn't that bad. It's possible to remember the new pieces, and the fact that they all promote to golds is a great help. Chu, dai, and tenjiku really do deserve to be played more: they are really well-designed. I have to wonder though if dai-dai, maka-dai-dai, and tai are simply jokes – although maka-dai-dai alone among them could be taken seriously. Double sharp (talk) 15:44, 13 February 2016 (UTC)
 * (Although dai is probably good only for the very bored...) Double sharp (talk) 11:13, 23 February 2016 (UTC)

Hook moves
Is this not exactly the same thing as an area move (cf. tenjiku), with the exception that it is maR for the hook mover instead of maK or maF, for instance? (Takes two moves of a rook/king/ferz/whatever piece you please, but must stop when it captures.) Double sharp (talk) 16:04, 27 February 2016 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified one external link on Dai dai shogi. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20070920062503/http://www.luckydog.pwp.blueyonder.co.uk/games/shogi/daidai.htm to http://www.luckydog.pwp.blueyonder.co.uk/games/shogi/daidai.htm

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 08:27, 3 September 2017 (UTC)

Mnemonic device piece set
Try to use the following mnemonic set for Shogi variant boards. Here's what it will look like for Dai dai shogi.

Mistakes


 * Black poisonous snake: The move shown here for the mnemonic piece used in this diagram viewed from Black's perspective is sWfAbD.

301fieriw (talk) 15:49, 15 June 2023 (UTC)