Also, I've created some open invitations for this game here:
http://play.chessvariants.com/pbmlogs/index.php?age=0&stat=open
I had one Chu Shogi and even one Tenjiku Shogi game going on on chessvariants.com. But since yesterday my login attempt results into an http error 500 message. I wonder if other users get the same message while trying to login for playing games.
Yesterday an OTB tournament took place and I participated - very exciting to play 3 fatiguing OTB games on one day.
https://www.facebook.com/groups/1139225779613122/
http://www.shogi.be/tourne.php?idTournoi=196
I recently discovered Chu Shogi, an ancient (1350 ad) Japanese Chess variant:
It features a very interesting piece, the Lion, which can make two King moves in one turn (possibly capturing two pieces at once). Of course the Japanese would play it with horrible pieces, all the same color and Chinese characters written on them, but the picture above shows a culture-free, highly mnemonic board representation. (Can you spot the Rooks? They are in front of the Bishops!)
Hi
Chu Shogi is kind of Japanese Chess but on bigger Board right !?
Don't you have a better diagram to show this game and explain how we play it?
Yesterday an OTB tournament took place and I participated - very exciting to play 3 fatiguing OTB games on one day.
https://www.facebook.com/groups/1139225779613122/
http://www.shogi.be/tourne.php?idTournoi=196
Wow, a draw?! Those must be rarer than triple rainbows lol. How in the world did you manage to draw a game?
We came in repetition of moves, he had a material advantage but less time so we agreed upon draw.
I recently discovered Chu Shogi, an ancient (1350 ad) Japanese Chess variant:
It features a very interesting piece, the Lion, which can make two King moves in one turn (possibly capturing two pieces at once). Of course the Japanese would play it with horrible pieces, all the same color and Chinese characters written on them, but the picture above shows a culture-free, highly mnemonic board representation. (Can you spot the Rooks? They are in front of the Bishops!)
Hi
Chu Shogi is kind of Japanese Chess but on bigger Board right !?
Don't you have a better diagram to show this game and explain how we play it?
As Chu Shogi has 21 different piece types, 18 of which can promote to yet other piece types, a full description of it here would become very long and tedious. And there already are elaborate descriptions elsewhere.
Unlike regular Shogi, Chu Shogi does not recycle captured pieces by dropping them back onto the board. This makes it much more like Chess. I am not sure in which way you want the diagram I posted to be 'better', but this diagram already goes a long way to explaining the game: the pieces are drawn in a way that indicates how they move. Many pieces move as a King, except that they lack moves in some directions (indicated by a dent in their square shape). Other pieces can slide in some directions over an arbitrary distance, like Bishops and Rooks (indicated by a radial line). A few pieces can also jump to the second square (indicated as a disconnected patch). Only the Lion (fat square) is a complex piece (basically a King that can move twice per turn).
I finally found the text for 中将棋指南抄 (元祿16年 =1703), which in theory would tell us how chu shogi was played in its heyday. My main focus was on promotion.
General promotion rule:「成駒は敵地へ入る時成る也、す馬にて入、二の手には不成、敵の駒を取りては、内にても成る也」
Pawns:「敵地の口にて成る、す馬にて入二の目には成らず、三目入四目にては金に成也」
Go-betweens and bronzes:「敵地へ行...に成る」
All other promotable pieces:「敵地へ入...に成る」. 入 simply means 'enter'. No confusion there. Totally in line with modern understanding.
So the only difference from what is accepted today (apart from the JCSA's ridiculous allowance for 12th rank lance promotion which even according to them is based on nothing but their president's whim) is that go-betweens and bronzes could potentially have had more freedom to promote in the enemy camp, after making an additional non-promoting move within.
Suggestion for improved 1-character piece abbreviations (for purpose of notation)
For FEN-style notation, 1-character abbreviation is ideal. Use uppercase for Sente's pieces and lowercase for Gote's pieces. For the special symbols £ and $, take € and ¢ to be their lowercase counterparts, respectively. To discern between promoted and unpromoted pieces, make promoted pieces bold. It is not necessary to make bold pieces which only appear after promotion, such as whales and boars.
I learnt Chu from wikipedia. Without having any particular opinion about it, I just grew familiar with the notation introduced there. What made you come up with a new system?
For in-game moves, Japanese notation is my preferred choice. I process that faster than Western notation.
For describing an actual position, you need either a diagram or the notation. Notation can do a few things that a diagram can't. My main gripe with Western FEN-style notation for chu shogi is that you need to bracket off pieces notated by two letters, which is somewhat bothersome. There are also some other improvements to notation that could be made.
Take the final position from our last game. My proposed notation would be...
I've also been working on a comprehensive repetition rule that is practical for OTB play. Right now, it's a bit unpolished.
It's not often that I find much in the way of critique when it comes to chu shogi, but there always have been a couple things that stood out to me as suboptimal, namely the frequent lack of activity for lances and RCs, and how the slowness of step movers tends to discourage their initial development. Of course, any significant improvement in their mobility would greatly imbalance the game. So I sought a solution that would only marginally improve their mobility while greatly improving their playability. Here are my proposals:
The WinBoard GUI already uses a system of 1-letter piece IDs in position FENs or SAN move notation, which IMO is more intuitive than what you propose. For one, it uses the convention that is generally used in western Shogi notation, namely to indicate promoted pieces by a '+' prefix. One could argue that because of this it is no longer a 1-character encoding, but for the purpose of FEN this is harmless, as the '+' is always known to be the first character of a piece ID, and there doesn't need to be an upper/lower-case distinction, as the second character wil reveal the color. The advantage is that it reduces the number of required ID letters from 36 to 21, so that it can remain strictly alphabetic (and thus benefit from the natural distinction between upper and lower case, rather than requiring non-intuitive additional definitions of counterparts of special symbols). Also, using style properties such as bold or italic in an encoding function is a really bad idea; even in unicode different styles are considered the same character, and on copy-pasting one would often lose the style altogether. Use of non-ascii unicode is ill-advised anyway; in Windows they might not work at all, depending on your locale.
So WinBoard has no need to encode pieces like Falcon, Eagle or Ox, as in Chu these only occur as promoted pieces. For the basic pieces WinBoard uses
As the number of pieces is so large that almost the entire alphabet must be used, it is hard to avoid a few awkward IDs, but for most pieces the ID corresponds with the first letter of the name. (And even in the comparatively trivial case of Chess the King-Knight destinction faces a similar problem.) The I and O were chosen because they resemble the move pattern, although these letters do not occur in the piece name. (One could have used W and Y, though, if one would spell Kylin instead of Kirin, as some people do.) Even using a non-occurring letter seems not nearly as bad as using a currency symbol.
For the larger variants the Latin alphabet is insufficient (Although in Dai Shogi you could just get by if you are willing to encode the unpromotable initial pieces Queen, Lion and King as +X, +O and +E, which are their game-theoretical equivalents, but that is really ugly.) For those games WinBoard uses the system of 'dressed letters': punctuation characters behind the letter can be used to alter its meaning. E.g. L, L', and L! (and +L, +L', +L!) are all considered different pieces. The use of ' and ! alone triples the number of available IDs. Which is so generous that even in the larger variants more natural letters can be chosen. E.g. L! can be Lion, F! can be Falcon, E! can be Eagle, (in Tenjiku these are base pieces), S' can be Stone... To optimize intuitive understanding I prefer to reserve the ! suffix for really strong pieces, and single quote for weak ones. (So G' would be natural for Go Between, R' for Reverse Chariot.) If more IDs are needed, ` (back-quote) and " can also be used as suffixes.
Since I have had more time recently, I decided to take another look at the KQ v KTT endgame, with special regard to the 199-move rule. I found new resources for both sides, with better designed avoidance techniques.
The chosen start position
Moves
End position
As you can see, Sente barely promotes in time to reset the 199-move draw rule and win the game. Obviously, with human analysis this long, there are bound to be some errors. If you find any mistakes, please let me know. This just affirms that the length of the draw rule is justified, since this - one would imagine - is about the slowest possible winning promotion setup. I may look into other endgames for sake of comparison.
holy crap this is old chess
old variation
This is my first submitted attempt for a functional "common sense" version of the chu shogi repetition rule. The way that the JCSA currently writes the rule is too simplistic and far too heavily gives credit to the attacker, whereas the old rule would theoretically have given the defender right of way. To honor original intent, I wanted a rule that supported defending play, in most of its aspects, yet still punished moves that are more "attacking" than necessary.
I aimed to come up with a simplistic measure of how "attacking" any given move was, relatively easy for the players to calculate as they play. And since pieces being attacked doesn't cover quite all of attacking play, I added in a few more criteria to round off that definition. After these criteria come a list of exceptions that account for it not being that player's fault, bad moves, and incidental stuff. If you can attack subtly enough or in a complicated enough manner to bypass the attacking criteria, then you've earned it. It is more important that all the elements be easy enough to calculate than for all questionable actions to be enumerated and debated.
I aimed to have the defending criteria be lenient, but not overly so. This at first was much more tricky than defining what an attacking move was, since there are many more ways to subtly play defense than there are to attack. So I came up with a new mechanic to allow a player to defend a bit more actively without violating the repetition rule, but with a downside so that this wouldn't be too heavily relied upon.
The rule elements themselves are meant to cover as many edge cases as possible, all potential ways for the mechanics to conflict with each other. If I overlooked something, let me know.
Overall, this submission could use a touch more refinement, but I think the content is quite good. The original draft had a section on arbitration, but it didn't feel right to include that here. It pains me to use such a straightforward basis for defining attacks, but the alternative just causes more trouble for everyone when a messy situation arises. It can be a nightmare to adjudge positional play, so I left that out except where it's clear to see.
All constructive feedback is appreciated.
I find the FIDE rules on repetition the most intuitive. If any player manages to get the game into a repetition, the other player was apparently not strong enough to win, which makes it a draw. In a real war or battle, if one army has a strategy to defend the position in some repetitive way and make the war go on forever, who would blame them for that?
The way piece dynamics work in chu shogi, with so many powerful ranged pieces, a lot of harassment can be done in short order, even forcibly so. There needs to be adequate prohibition to disincentivize what would be a high percentage of draws, to no fault of the stronger player. If the lion didn't exist, this wouldn't be so necessary, but alas...
Chu shogi repetition has historically allowed a defender to defend with repetition, until the JSCA with all their nonsensical rule changes came along. My version allows a defender lots of leeway, seeking to allow every sort of rightful repetition that does not come with attacking intent. What's important is that an attacker is not allowed to repeat.
Of course, a simply stated rule is best. But this is tough. If you can come up with a conciser rule that serves all the desired purposes, then please share. At least my version of the rule is not nearly as indecipherable as the official xiangqi repetition rule. Good grief, reading that rule is awful.
There are various versions of the Xiangqi cahsing rules ('Asia' or 'China Mainland' rules), and I only know the former (and was told the latter are even more complex, as even making mate-in-1 threats is considered chasing there). But I am not sure that what you propose here is simpler than XQ Asia rules. The underlying philosophy of Asia rules is that it only takes into account whether a piece is protected, and not whether it is protected enough, or whether the recapture would be a really poor move (e.g. because it exposes you to mate-in-1, or the recapturing piece was overloaded). One can criticize this for being inaccurate (i.e. leading to a ruling that feels unsatisfactory), but not for being overly complex. Of course there are some unavoidable complications in applying the basic principle, but these are due to XQ rules in general being more complex. E.g. unlike Chu Shogi, XQ has a checking rule, so attack and recapture must be defined in terms of legal moves. And the presence of the Cannon causes there to be many more repetition patterns based on alternately blocking/deactivating and unblocking/activating a non-moving Cannon that makes the attacks.
I think it is better to avoid complex rules when you can. And I am not sure that a rule as complex as what you propose here would really be needed. But I don't like the JCSA rules either. And the problem with outright forbidding all repetitions is that it can be both the attacker and the defender that have to repeat first. E.g. if a perpetual check is possible, the checking piece usually first comes from afar, from a position not part of the checking cycle. The checker must then abort the checking, and do another move. But if the opponent is not really able to make continuation of the checking impossible, and also plays an unrelated move, the checking can resume, but this time starting from a position after the evasion that is in the new cycle. Then the defender would be the first to repeat, and if he is not allowed to, it usually means he is now checkmated, as perpetuals typically occur when you only have one evasion.
For perpetual checking this is very easy to prevent by exempting check evasions from the repetition rule. A similar solution with other threats runs into the problem that it requires a definition for what constitutes a threat. Even for 'check' this is not obvious in Chu Shogi, btw, because of the multiple (extinction) royalty. I would say that an attack on King or Crown Prince when you have both K and CP does not count as a check; your spare royal is an expendabe piece. But if chasing pieces is also forbidden this would become a moot point.
Would it really be a problem if there was always complete freedom in moving pieces that just got attacked? (I.e. also play moves that repeat positions.) Irrespective of whether they were actually protected, protected enough, what their value is compared to that of the attacker, whether capturing them was bad... And not only moving the attacked piece away, but also allow interposition. (Although it will be much rarer that those occur in a perpetual; you would have to use the same piece to alternately interpose in two different attacks by the same piece.) This is a very liberal definition of 'threat', and it is hard to imagine a player would be in trouble forcing him to repat a move when the previous move of his opponent did not attack anything at all. If you want you could add mate-in-1 threats, and allow complete freedom of move choice when the opponent can checkmate you.
This doesn't unduly penalize a defender that happens to make some unimportant attacks as unintended side effect of his defensive moves. It would only grant the chaser to freely move those counter-attacked pieces, or interpose along the attack ray. But that would likely not stay on any repeat cycle, he would abort the chasing for that. Which he would not want to do if it is a futile counter-attack, and otherwise accomplishes the goal.
The main problem I see with this very simple rule is that there could be perpetuals that mutually attack, e.g. a Rook that threatens to deliver checkmate by moving to last rank, defended by interposing a Rook on another rank protected from the side. Interposing the Rook would then counter-attack the Rook thatthreatened mate. This doesn't seem to suffer from the problem of 'revival' of the perpetual with opposite obligation to change move, as neither player could afford to stay on the repeat cycle. Perhaps this can be solved by ranking threats (e.g. check - mate threat - other attack), and requiring that after a repetition-violating move another such move can only follow if it solves a threat of higher rank.
It is also a question what type of chasing we want to outlaw. In XQ, alternately attacking several pieces is allowed. It is only a chase if you attack the same piece on every move. Not sure why they do that; probably because it isn't a problem in practice.
2nd attempt for a position where a sequence of moves will capture both Black's prince and Black's king. Position shown from white.
http://scrybqj.com/images_diversen/chu_shogi_mate_king_prince_both_01.jpg