The Chess Variant Pages
Custom Search

[ Help | Earliest Comments | Latest Comments ]
[ List All Subjects of Discussion | Create New Subject of Discussion ]
[ List Latest Comments Only For Pages | Games | Rated Pages | Rated Games | Subjects of Discussion ]

Comments/Ratings for a Single Item

Later Reverse Order EarlierEarliest
This item is a game information page
It belongs to categories: Orthodox chess, 
It was last modified on: 2001-12-14
 Author: Hans L. Bodlaender. Shatranj. The widely played Arabian predecessor of modern chess. (8x8, Cells: 64) (Recognized!)[All Comments] [Add Comment or Rating]
Todor Tchervenkov wrote on 2020-05-02 UTC

I knew the rules of Shatranj for a few years but had never attempted an actual game. But these days I'm again delving into historic Chess variants. My primary goal is to find a few sample games of Shatranj which would hopefully let me understand why the game was appreciated (for me it is just impossible to play it: I fill lost when I open it in Zillions, I don't know where to go, what short term goals to pursue).

While searching for sample games, I discovered the astonishing lack of historic recorded games of Shatranj. I found but two, dating back to the Xth century. It turns out that apparently Shatranj was never played from the initial setup. Players would agree on a standartized position -- which could be called an opening in modern terms -- and would use it as actual setup. I found sixteen examples of such openings but without an analysis of their strengths and weakness it is still difficult to use them. One can still admire their poetic names.

In my opinion, we see Shatranj as a poor, uninteresting game just because we don't know enough about it. It would be so nice if somebody could provide us with the analysis of As-Suli, mentioned by George Duke back in 2008 in the first comment to this page. Perhaps more knowledge of the actual way this game was played would allow us to better appreciate it, since initial setup, piece movement and winning conditions don't seem to be enough?

Kevin Pacey wrote on 2018-03-01 UTCGood ★★★★

A poorish game by modern standards, especially due to the alfil pieces, but modern chess is indebted to this historic early version of it.

Anonymous wrote on 2012-04-03 UTCPoor ★
because i didnn't play a lot.

George Duke wrote on 2010-11-01 UTC
Mohsen's ''English authoritative sources of Chess history'' are still led by 100-year-old Murray 'A History of Chess'. That is unfortunate because Murray's style is not fluid. Yet the other chess historians do not deserve mention on the same level because of far less content than Murray's. How about etymology of ''King'' through Persia or Arabia?
Also HORSE is already prevalent English name for the chess piece hippogonally jumping.  In English there ought to be non-humanistic names for all six chess pieces.  Metals could be used, or animals, or birds. Here is a chart of equivalences: Metals have Pawn-Silver, Horse-Iron, Bishop-Mercury, King-Tin, Queen-Copper, Tower-Lead.    Then player promotes his Silver to Copper, rather than Iron Horse, and starts with cornered Leads, who move orthogonally. There is correspondence to Gilmanesque organization in Silver obviously being one Shogi-style pawn-type, and like Tin King, who may be imagined tinpot dictator or 'Wizard of Oz' tinman aspiring for a heart. Just ''Tin'' impartially takes the sting out of it all. Tin check.

Charles Gilman wrote on 2010-10-31 UTC
The issue of Xiang Qi actual strengthens Mohsen's case for not calling the Ferz a 'General', as General is the name generally used for another piece in that game. The Ferz represents a kind of adviser in both games, so why call it something that doesn't mean a kind of adviser?
	The Shogi analogy doesn't work at all. The names of the King, Rook, and Bishop pieces in European languages are regularly used in the context of FIDE Chess, so it is natural for Europeans to use these names even when playing Shogi - and even devise new European names for the rest of the pieces. My own Chatelaine, Helm, Point, Primate, and Wing are the most comprehensive list of such names. Incidentally one player's King in Shogi is called 'King's general' and the other 'Jewelled general', the latter without specifying a particular kind of jewel.
	Calling the Ferz a General simply doesn't compare. The Ferz is this context not an piece exotic to players alreadsy familiar with FIDE Chess, bit the precursor of the FIDE Queen. As far as I am aware it was only ever known in Europe either as Ferz or some variant spelling, or by the local names now used by the modern Queen.
	One small point about real-life bishops: they are certainly not 'Humanistic' in the religious sense, quite the reverse!

mohsen wrote on 2010-10-29 UTC
you should consider that walker, Shooter, raunner, jester and bishop all
are humanistic personas. 

my discussion is against using 'General'. If english name for this peace
before changing piece move was general, this name is certainly correct but
if thay named it ferz or another name you should use it. I'm not able to
find any english authoritative sources of chess history.

H. G. Muller wrote on 2010-10-29 UTC
We should not confuse 'European' with 'English'. In no other language I know the Bishop is actually called after a clerical person. In German is is 'Laufer' (runner), in Dutch 'Loper' (walker) or 'Raadsheer' (adviser, counselor), in French 'Fou' (fool, jester), in Spanish 'Alfil' (so they retained the Arabic word for elephant, despite the fact that the piece moves differently).

Similarly the name of the Knight piece in many languages has nothing to do with a knight. In German it is 'Springer' (leaper), in Dutch 'Paard' (horse), in Spanish 'Caballo' (horse).

All in all, Chess pieces have pretty exceptional or weird names in English, if you compare it with the rest of the World. But I suppose any language has the right to name the pieces like they want. There is no need for litteral translation. The name Lance for this Shogi piece is a much better name than the litteral translaton of the Japanese 'Incense Chariot'. (And I don't think it would be progress to refer to a Bishop as 'Angle Mover' either, there.)

As to the Ferz, it is interesting to note that in Chinese Chess (Xiangqi) the common name for this piece is Adviser (the alternatve name, '(Palace) Guard', no doubt being inspired by the unique feature of Xiangqi of the Palace board zone to which this piece is restricted). It is sometimes also called 'Mandarin', which I guess in Chinese culture is some kind of Advisor to the emperor. The Elephants there are also often referred to as Ministers. Confusingly enough, in English, the word 'minister' can also refer to a clergiman, as well as to a statesman.

In Shogi all pieces with King-like moves are called 'generals' (includingthe King itself, which is the 'Jade General').

mohsen wrote on 2010-10-28 UTC
General is obviously a wrong translation for ferz. Original name of this
peace was 'Farzin' (that in europe spelled it 'ferz' or 'ferse') and
'Vazir' (that also spell 'Vizier'). Farzin is a persian word that
commonly precieved meaning is 'Minister' or probably 'Prime minister'
but as murray mentioned and I discussed in my comment to ferz article in
your site, this is not totally true. original meaning of farzin is wiseman.
altough there is a strong relation beetween wisemanship and role of
ministers in history. main chracteristic of persian and Islamic prime
minister was extensive Knowledge and a king must elect most wisest men for
his prime minister. this reflected in many tale, folk and storys. 'Vazir'
is an arabic word that means 'Minister' or even 'Prime minister'. there
is a rationlaity for name of 'General', in time of war 'Prime minister'
is present and act as a general but this is not his name and no one call it

there is many source that translate this peace fers or vizier. if you want
to translate it, Minister or even prime minister is true. 

another wrong name in this article is 'Knight'. name of this peace in
both persian and arabic literature was a word that means horse, as sanskrit
name of peace in chaturanga and name of peace in modern chess in most
regions in asia. in europe shatranj horse changed to chess knight. I think
this transition is for giving humanistic personality to chess peaces, this
change also happend for elefant that replace with bishop and other names.

Anonymous wrote on 2010-03-25 UTC
What is difference between Shantraj and Chaturanga?

George Duke wrote on 2009-08-22 UTC
We are the keepers with enough respect of tradition for Shatranj staying alive. Nowhere else online is it played. Nicknames Pastchess and PastPastChess, the latter for this endeared Shatranj. Single ''PastChess'' is reserved for OrthoChess64, also known as f.i.d.e. chess, aka Mad Queen. Make no mistake, our wording of ''PastChess'' is not Dead Chess, perish the thought, far from it. Make that very clear. Certain small-board titled grand-masters have called OrthoChess64 Dead. In principle, we should hear nothing of it and never let it happen, because we revere the past traditions. So long as there are stars in the night, PastChesses will always be played. It is going to be played a million times more, in honour of the past 500 years. The torch unquenched, unsheath the sword.

George Duke wrote on 2009-08-05 UTC
I think Joyce misses some irony about Shatranj Shuffle in his good points. Frankly Shatranj is not worth playing but is of utmost historical importance. The infinite sliders, Queen and Bishop, are here to stay. That's as sure as the fact that infinite slider Rook has ALWAYS been here. When they radicalized Shatranj around 1492, they didn't go far enough. The concept of complementarity from 600 to 1492 meant Rook, Knight, Alfil and Ferz going to mutually exclusive squares. Variantists tend to waste time thinking up anything, ignoring the perfection inherent in complementarity of piece-moves. It was inappropriate to have Ferz any more once adopting full-range Bishop and Queen. 'Regina rabiosa', today's F.i.d.e. chess on 64 squares, improved Shatranj by instituting Rook, Knight and Bishop, one and all mutually exclusive, and necessarily dropping Ferz and Alfil both to the dustbin of history. Further extension of complementarity has to take a large board, such as enabling Rook, Knight, Bishop and Falcon on 8x10 like Complete Permutation shows -- or some other scientifically-chosen piece-mix on 8x12. (There are other solutions too, but Joyce-involved experimental proliferation is hopeless; and more importantly, the solutions are appearing impossible on little 64 squares anymore.) This business of ''high levels'' of Chess being exhausted flaunts the reverse reality. OrthoChess64 has lost whole generations at the lower echelon, where kids don't even know the rules of RNBKQP. Why? Because to 21st-century sensibilities of the general public, there is no more intrique in OrthoChess64 than in Shatranj.

Joe Joyce wrote on 2009-08-05 UTC
In a way, shatranj is like baseball. It packs 10 minutes of action into 4 hours of playing time. :-) The only people who play it, as far as I can tell, are us, pretty much. The only place I found playable shatranj online besides here was at Zillions, and that was mostly all shatranj variants. Even the one guy who offered the original gave variants on it. [Okay, I didn't look hard, but still, looking through the first few pages of a few google searches turned up only Zillions and this site for playable shatranj. Not a big audience.]

As a game, shatranj was superceded by more modern forms, which had better play value. Chess is a better game than shatranj, just as Go is a better game, with more play value, than tic tac toe. But shatranj superceded some other, more ancient game[s], because it was better and gave more play value then. [Think about that for a minute: if the alfil and ferz were significant improvements over the pieces before them, how bad was that previous game, to modern sensibilities? ;-)]

Maybe it comes down to a case, not of exhausting the possibilities, but of exhausting the probabilities. FIDE seems to have hit that point, at high level, anyway. In some senses, the 'good moves' are already taken. That's why the shuffle. Now, note this - since chess has 2 infinite slider [singly] colorbound bishops to shatranj's pair of 8-square-only alfils, and the modern queen to shatranj's colorbound ferz, clearly, the 'good moves' in chess far outnumber those in shatranj. 

For shatranj to be worthwhile as an intellectual game of the first order again, it must be updated, not just re-arranged. To keep it shatranj, the updating must involve shatranj-style pieces, and avoid the infinite sliders that mark modern chess as clearly different. The goal is to increase the possibilities without turning shatranj into FIDE.

George Duke wrote on 2009-08-05 UTC
It tends not to be noticed that C960 castling is a 2/3 reduction, because of the other accompanying reduction for Bishop-same-colour. Instead of every King placement, C960 permits only 56 of the 168 respecting the two Rooks' positioning. (In general, on 8x8 fixed castling is better, and by 8x10 free castling is better -- but that's an educated value judgment.) Since ShatranjShuffle2880 has no castling, it has 3 times the arrays available over C960, wherewith Rooks to King placements are not mattering. Hence the exact times 3, what we were trying to understand.
(1) Incidentally with mirrors 168x2 = 336, and that number 336  keeps recurring, chiefly just because it is 8x7x6, such as at Man & Beasts 04,
(2) Technically 960/40320 = 0.0238 = 1/42, the fraction being exact. So C960 uses only 1/42 of pure (8 factorial), representing every possible permutation, and most of the 960 are still ugly as a bulldog.

George Duke wrote on 2009-08-05 UTC
These comments are appropriate for Shatranj because its inventors are dead for over 1000 years and cannot complain about our superimposition of Shatranj Shuffle2880. To diehards Shatranj Shuffle can be the game for the ages with its liberal modifications. Alexandre appears to have invented random chess in the 1820s. The 1920s saw international tournaments for basically the same idea as Free Chess of Brunner. Chessplayer Fischer forced King to castle at c1 and g1 and called random chess revived for the 1990s Chess960. It's still being played, so let's compare chess960 with old Shatranj2880 from the last two comments here. Now 2880=3x960. How so? Square-colour-requirements of Shatranj Alfil and OrthoChess Bishop are comparable. No difference there. Likewise 2 Rooks and 2 Knights are pairwise indistinguishable both cases of 960 and 2880. Chess960 requires King between two Rooks, and 2880 does not, the only remaining difference with Shatranj2880. How does that make for the factor of 3x? This chart shows Rook combinations and # allowable King placements each case in Chess 960: ab 0__bc 0__cd 0__de 0__ef 0__fg 0_ac 1__bd 1__ce 1__df 1__eg 1__fh 1__ad 2__be 2__cf 2__dg 2__eh 2__gh 0__ae 3__bf 3__cg 3__dh 3__af 4__bg 4__ch 4__ag 5__bh 5__ah 6. There are 28 combinations for Rook. If there were no exclusions invalidating King placement, there would be remaining 6 squares for the King each case of Rooks. 6x28=168. Allowance of 56 of the 168 seen in the list, 1/3 of them, is precisely the Chess960 Fischer-agonized methodology. And shows why 960 times 3 equals the 2880, where there are no castling and restriction of K in the same game (Shatranj and OrthoChess being basically the same in the sense of 6-piece and 64 squares).

George Duke wrote on 2009-08-04 UTC
Shatranj2880 preserves Alfil bindings. Shatranj Shuffle opposite sides'
8 r__n__a__k__f__a__n__r   starting arrays are perfect mirrors. That goes
7  __ __ __ __ __ __ __    for the four Alfils (elephants) as well. Each
6  __ __ __x__ __ __ __x   team has one White- one Black-cell Alfil. In
5  __ __ __ __ __ __ __    the perfection of mediaeval game, there can
4  __x__ __ __ __x__ __    NEVER be Alfil x Alfil.In two words,IM-
3  __ __ __ __ __ __ __    POSSIBLE, for their respective bindings never
2  __ __ __x__ __ __ __x   impinge. In board to left, we use the standard
1 r__n__a__k__f__a__n__r   RNAKFANR for convenience of familiarity. 
  a  b  c  d  e  f  g  h   Imagine also the mediaeval game was checkered, which was not always the case. The White's Alfils are c1 and f1, Black's c8 and f8. Now c1 and f8 are on black SQUARES. Follow f8's binding-path as f8-d6-b4-d2-f4-h6 and two what we might call ''offshoots'' h2 and b8. Each of the four Alfils reach their particular 8 squares for 32 all together, half the board. Notice f8-Alfil has route that just nips White's c1 (at d2) but he does not hit him. Both f8 and c1 stay on Black squares but never correspond by reaching exact same squares. Their bindings are all of them independent of one another. There can neither be AxA nor even Alfil defending Alfil. NOT IN ALL RECORDED HISTORY. 
Shatranj2880 adds 2879 more initial arrays with the SAME CHARACTERISTIC free-ranging Alfils -- proven valid piece interaction for a thousand years from 600 to 1500 -- and good enough for any right-thinking person for the longer-term foreseeable future.
__________________''Can we speculate that, had the later, very efficient malaria protist lived in Europe during antiquity, classical culture with its influence on modern civilization worldwide would not have developed?'' --Ricardo Guerrero microbiologist

George Duke wrote on 2009-08-03 UTC
The Shatranj Shuffle. Would our new CV have saved Shatranj from extinction? The Shatranj Shuffle is the simple technique to randomize the back-rank of the mediaeval game. Similar expedient is to be lifesaver of OrthoChess64 such as RNBQKBNR, or RKRBQNBN, or QRBNNKRB, or anything else you want up to 960 beauties. BNRKRNQB is another one still your basic Mad Queen chess64, now within chess960, or any other loose -- rather than strict -- prior-named random Orthodox rules-set. For Shatranj Shuffle, (Shatranj never having had the powerful Queen or long-range Bishops) F is Ferz the one-diagonal and A is Alfil the two-diagonal-only leaper. Those two complement in their inimitable way. There is no castling in pure Shatranj. The only subrule restriction in S.S. starting array is that the Alfils must be on opposite colours. That's it. Here are some Shatranj Shuffles, also known as S.S. Saviours: ARKFNRNA, NAARKRFN, RNAFKRNA. Ferz (queen) reaches 1/2 the squares, and that's why promotion is to another Ferz, just like in the year 1400 -- but decidedly NOT the year 1492 by which time Mad Queen(Rook + new Bishop) had taken root. Does Alfil (elephant) still reach 1/8 the squares, regardless where back-ranked initially? Of course, that's the math of it, and the Shatranj Shuffle gives many more opportunities._____________________________________________ How many? Let's see. (8 factorial) divided by 2 for the indistinguishable Rooks, divided by two again for the indistinguishable Knights. Times four-sevenths (4/7) for the Alfil same-colour restriction. That gives 5760, but we need one more division by 2 for the mirrors. 2880. 2880 starting line-ups, and 2879 of them are new. 2880 permutations, far more than a chess 96 or 100 or random chess960. If Shatranj Shuffle had been implemented in time, we would still in due course be happily Alfil-two-stepping today. [The original 2880 is confirmed correct. Shatranj Shuffle is now also known as Shatranj2880.]

George Duke wrote on 2009-08-01 UTCGood ★★★★
The linked comment is vintage Betza, usually even better expressed in his polished articles. Betza comments always as ''gnohmon'' and had this to say over seven years ago at Chaturanga in 2002.
''My average of the two skills is higher than the divine Parton or superhuman Fischer.'' Like Gilman for many years, Betza did not use formal identification, and so could not revise his words. I think there are some contradictions in terms here Betza would not fully defend.  Yet this comment shows Betza's coherent/confused mindset the year he left. By August 2003, no more all-too-profound Ralph Betza.  Also some of this particular comment by Betza would be deliberate obfuscation by him just short of sabotage -- conclusion that I can justify and cross-index another time.  A couple significant sentences, midst the true account, he knows to be untrue or does not really mean for effect, in hyperbole not for purpose of satire.

M Winther wrote on 2009-06-06 UTC
I suppose, the mentality of modernity isn't compatible with historical Shatranj. This doesn't mean that the game is totally uninteresting. We still like to watch costume films about the 18th century, but nobody would like to dress up like that anymore and dance silly minuets. We are all products of our time. Should we analyze the variants created on this site, we would get a picture of the collective psychology of modernity. In my view, many variants are somewhat overbearing and high-flying, i.e. simply over the top. But far from all, of course.

H. G. Muller wrote on 2009-06-06 UTC
Standard Shatranj is an excessively boring game. It is very slow, the Pawns and Ferzes merely crawling over the board, and the draw rate between equally skilled players is about 70%. (For Mad Queen this is about 30%, for Capablanca-type variants only 16%.) Having watched many Shatranj blitz games, I can imagine why there was such an incessent drive to modernize some of the pieces.

I don't think shuffling would solve anything. And it certainly would not have to invoke Fischer, as there is no castling, and none of the special rules invented by Fischer to preserve castling in Shuffle Chess would have to be applied. (i.e. no reason to require K between R in Shuffle Shatranj.) 

Elephants would still remain useless pieces under normal shuffle rules, which would keep color-bound pieces on different colors. It would be more interesting to allow Elephants on the same color, so they can defend each other (as in Xiangqi), then they could be used in fortress building. But the risk is that the game might become even more difficult to win then, in absence of Cannons to penetrate the fortress.

Greg Strong wrote on 2009-06-05 UTC
Joe, I think that a 'randomized' Shatranj might be interesting, especially how it greatly affects the measly 8 squares that each elephant can target. I'm not sure George agrees; his messages are so complex, I frequenetly not sure what side of an issue he's on :) I would certainly dispute the notion that Fischer wasn't innovative. In addition to Fischer Random, which is a fantastic way to 'save' chess while escaping from the fact that you need to memorize hundreds or thousands of lines to be a well-rated player these days, how about Fischer clock? It's a great idea, and is widely used for getting better results from computer matches.

Joe Joyce wrote on 2009-06-05 UTC
George, you won't get me to disagree with your first sentence. I'm always happy to see more shatranj variants played. As for the central rook, I'm currently playing a game of ShortRange Courier with its designer, David Paulowich, and it uses no queen, but rather a rook in its place in the center of the board. Aside from the rooks [3/side] there are no longrange pieces in the game.

George Duke wrote on 2009-06-05 UTC
How about Fischer Random Shatranj, and they could have skipped the whole Mad Queen phase. Or Alexandre Random Shatranj. That refers to one of the great French -- and others British -- players, Alexandre, who secretly crawled inside (fooling most but not E.A. Poe) the Turk automaton a while for Maelzel, Alexandre who started random back-ranks in 1820s. Friend of Beethoven (just remember m.m.), Maelzel had purchased the Turk through Napoleon's stepson, and Alexandre also made first Chess encyclopedia. The automaton Turk spanned 1769, invented by Wolfgang von Kempelen, to 1854, when he burned to death in the Chinese Museum at Baltimore, his last words ''Echec! Echec!,'' recorded by Poe's doctor's son standing in the burning stairwells. Over 85 years Turk played Napoleon, Ben Franklin, P.T. Barnum. Bobby Fischer had nothing new, nothing at all, all talk and no cattle, no concern for the masses of chess players wanting new challenges. Paul Morphy was more innovative for his time, connecting Europe and America. There are several random embodiments from 1820s to 1990s before Fischer's slight tweaking announced at Argentina. But nobody properly went back to Shatranj, and asked what need for so powerful Queen anyway? Otherwise, if not Random Shatranj, there are probably some good solid four to six pairs of pieces, probably Rook, Knight, Bishop, possibly Falcon, and up to two, or three, more for 13 x 8 as the next logical size. Then no stand-alone Centaur (BN), Champion (RN) or Queen-type (RB), none of them. Shatranj with occasional centred Rook randomized in starting array would be power enough; and Shatranj was good enough chess for Chaucer, who may never have formally mentioned it like Shakespeare. (When Shakespeare has Ferdinand and Miranda play Chess in 'The Tempest' in the West Indies, America, it is already Modern Chess not Shatranj, although right on the cusp.)

John Smith wrote on 2008-12-19 UTC
I want that!

Nuno Cruz wrote on 2008-12-19 UTC
Take a look at this:
Nice. And old.... (1988)


H. G. Muller wrote on 2008-10-15 UTC
Standard Staunton-style piece set for this game:

25 comments displayed

Later Reverse Order EarlierEarliest

Permalink to the exact comments currently displayed.