The Chess Variant Pages
Custom Search

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

Comments/Ratings for a Single Item

Earlier Reverse Order LaterLatest
This item is a game information page
It belongs to categories: Orthodox chess, 
It was last modified on: 2001-09-29
 Author: Hans L. Bodlaender. Inventor: Bruce  Zimov. Knightmate. Win by mating the knight. (8x8, Cells: 64) [All Comments] [Add Comment or Rating]
Robert wrote on 2002-08-13 UTC
I'm assuming that all the rules regarding stalemate, en passant, etc still apply, and that pawns can be promoted to any piece except a Knight. In other words, they can be promoted to a King if you so desire. I have a challenge. Can anyone contruct a chess problem where one side promotes a pawn, such that, promotion to a King wins the game, but promotion to any other piece, (Queen, Bishop or Rook) either loses or draws. It's relatively easy to imagine a conventional game of chess, where promting to a Knight is desirable, because of the Knight's unique movements. But it's not so easy to envision a scenario where you would want to promote to a King in a Knightmate game.

Peter Aronson wrote on 2002-08-13 UTC
What about this situation (white to move): <b><pre> +---+---+---+---+---+---+---+---+ 8 | |:::| |:::| |:::| |:::| +---+---+---+---+---+---+---+---+ 7 |:::| P |:::| |:::| |:::| | +---+---+---+---+---+---+---+---+ 6 | |:::| |:::| |:::| |:::| +---+---+---+---+---+---+---+---+ 5 |:::| |:::| |:::| |:::| | +---+---+---+---+---+---+---+---+ 4 | |:::| |:::| |:::| |:::| +---+---+---+---+---+---+---+---+ 3 |:::| |:::| |:::| |:::| | +---+---+---+---+---+---+---+---+ 2 | |:::| |:::| |:::| |:::| +---+---+---+---+---+---+---+---+ 1 |:n:| |:::| |:N:| |:::| | +---+---+---+---+---+---+---+---+ a b c d e f g h</pre></b> If white b8=Q or b8=R, then stalemate. If b8=B, then I believe there is insufficient material for mate -- thus b8=K. I'm fairly sure that K+N can mate bare N. If not, add a B at g1. Two B on the same color + N vs N is still not promising, but K+B+N vs N is surely enough!

Peter Aronson wrote on 2002-08-13 UTC
Hmm. A Problem with my previous post -- White could move their Knight, then on the following move promote the Pawn to Queen. <p> However, this situation ought to do it. <b><pre> +---+---+---+---+---+---+---+---+ 8 | |:::| |:::| |:::| |:::| +---+---+---+---+---+---+---+---+ 7 |:::| P |:::| |:::| |:::| | +---+---+---+---+---+---+---+---+ 6 | |:::| |:::| |:::| |:::| +---+---+---+---+---+---+---+---+ 5 |:n:| |:::| |:N:| |:::| | +---+---+---+---+---+---+---+---+ 4 | |:::| |:::| |:::| |:::| +---+---+---+---+---+---+---+---+ 3 |:::| |:::| |:::| |:::| | +---+---+---+---+---+---+---+---+ 2 | |:::| |:::| |:::| |:::| +---+---+---+---+---+---+---+---+ 1 |:::| |:::| |:::| |:::| | +---+---+---+---+---+---+---+---+ a b c d e f g h</pre></b> Since white better promote the Pawn or black will take it.

Robert wrote on 2002-08-13 UTC
I guess that raises a follow-up question. What material is sufficient to
force checkmate. 
Q + N vs N ?
K + N vs N ?
R + N vs N ?
two Rooks + N vs N
two Bishops (on opposite colours) + N vs N

The Queen mate is easy and in fact the Queen could force mate on it's own
without help from it's own Knight. I'm pretty sure the two rooks can do
it. I'm not sure about the one Rook though. The King may be able to but
I'd have to work it out. The two Bishops, probably, but again I'd have to
work it out. Had anyone actually played this game. Are strategies similar
to the conventional game, ie, control the centre, develop your pieces etc.

Peter Aronson wrote on 2002-08-14 UTC
According to the Encyclopedia of Chess Variants, in Knightmate a K can
mate a N unaided, so my solution to your challenge stands.

Robert wrote on 2002-08-14 UTC
While it is possible to get a configuration whereby a King alone can mate a Knight without help (the Knight would have to be on a corner square) I'm not sure it's possble for a King to force the Knight into that position. I'll have to try it one day and see if it works.

Peter Aronson wrote on 2002-08-14 UTC
Valid point. However, this version of the problem, with white to move: <b><pre> +---+---+---+---+---+---+---+---+ 8 | |:::| |:::| |:::| |:::| +---+---+---+---+---+---+---+---+ 7 |:::| P |:::| |:::| |:::| | +---+---+---+---+---+---+---+---+ 6 | |:::| |:::| |:::| |:::| +---+---+---+---+---+---+---+---+ 5 |:n:| |:::| |:N:| |:::| | +---+---+---+---+---+---+---+---+ 4 | |:::| |:::| |:::| |:::| +---+---+---+---+---+---+---+---+ 3 |:B:| |:::| |:::| |:::| | +---+---+---+---+---+---+---+---+ 2 | |:::| |:::| |:::| |:::| +---+---+---+---+---+---+---+---+ 1 |:::| |:::| |:::| |:::| | +---+---+---+---+---+---+---+---+ a b c d e f g h</pre></b> Should do it. After b8=K, black Nb3 is forced. After white Kb7, the black N is pretty well trapped.

H.G.Muller wrote on 2006-01-09 UTCGood ★★★★
I ran end-game data-bases for this game, and it turns out that for the
single-piece end-games only N+Q can force the royal Knight into
Not even a Rook suffices, and neither does a (Commoner) King.

Any two-piece advantage can enforce checkmate, though, except N+N+N, but
in Knightmate that is even outlandish as K+K+K in FIDE chess (but, by the
way, easily beats K+R there!). In particular N+K+B is an easy win over a
bare Knight, the position after b8(K) in the proposed problem is a mate

1. Kc7 Nb3
2. Nf3 Na5
3. Kb6+ Nc4
4. Bc5 Nb2
5. Kb5 Nd3
6. Be3 Nb2
7. Kb4 Nd1
8. Ba7 Nb2
9. Kc3+ Nd1
10. Kc2++

Peter Aronson wrote on 2006-01-09 UTC
Any two-piece advantage can enforce checkmate, though, except N+N+N
Does that include N+B+B vs N when both Bishops are on the same color? If that's the case, then I still haven't come up with a case where promotion to Commoner is required.

David Paulowich wrote on 2006-01-10 UTCExcellent ★★★★★
White: Knight (d3) and Pawn (e7), 
Black: Knight (a6) and Rook (d8) 

8  |   |:::|   |:r:|   |:::|   |:::|
7  |:::|   |:::|   |:P:|   |:::|   |
6  | n |:::|   |:::|   |:::|   |:::|
5  |:::|   |:::|   |:::|   |:::|   |
4  |   |:::|   |:::|   |:::|   |:::|
3  |:::|   |:::| N |:::|   |:::|   |
2  |   |:::|   |:::|   |:::|   |:::|
1  |:::|   |:::|   |:::|   |:::|   |
     a   b   c   d   e   f   g   h

leads to 1.exd8(K) Nb8 2.Kc7 Na6 3.Kb7 mate! Promotion to a Queen is stalemate. To meet Robert's challenge in the first Comment, we need to verify that promotion to either Bishop or Rook in this position will also lead to a draw. This 'capture promotion' ending could reasonably happen in a real game after the Black Rook had captured on d8. But it would be extremely rare.

H.G.Muller wrote on 2006-01-10 UTC
No, Bishops on the same color cannot force a checkmate. I failed to mention this, but it is almost universal that piling up power on one color doesn't buy you anything as long as you don't have any influence on the other color. (Extreme case: King + 32 Bishops that loses to K+B.) So the original problem is validated.

Peter Aronson wrote on 2006-10-20 UTC
This game is now available (for members, anyway) on Unfortunately, they did not give credit to Bruce Zimov for the invention.

Gary Gifford wrote on 2006-10-21 UTC
Bruce Zimov created this interesting game back in 1972.  I only became
aware of it today through Peter's comment.  I believe Mr. Zimov's game
deserves a pre-set, so I just made one.  It is located at this link.  The
pre-set has a link to the rules.


H. G. Muller wrote on 2008-05-12 UTC
Note that there has just been released a WinBoard compatible version of the variant-capable engine Dabbaba of Jens Baek Nielsen. One of the games it knows is Knightmate. You can currently watch it play a Knightmate match live against my own engine Fairy-Max, on my Chess-Live! webserver

for the next one or two days.

If anyone knows any other WinBoard engines that can play Knightmate, let me know; then I could hold a tournament.

H. G. Muller wrote on 2008-05-14 UTC
My small live tourney has led to a proliferation of WinBoard-compatible Knightmate engines. We now have:

JokerKM (
CCCP-Knightmate (
Fairy-Max (, do not forget to download the accompanying fmax.ini with game definitions!)
Dabbaba (

JokerKM is the strongest, CCCP and Fairy-Max are both about 400 Elo points weaker. Dabbaba is a rebuild of an old DOS engine from the 90s, and is some 300 Elo points behind that.

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

George Svokos wrote on 2009-10-31 UTCExcellent ★★★★★
This is a fun and challenging chess variant. It would be nice if someone produced a 'KnightMate' specific set like the one shown in the below comments. I like the Mate the Knight concept and it might be interesting if there were a smaller variant played on a 6x5 board. The royal knights would be placed on c1 and c6, cardinals/archbishops on a1 and a6, and marshals on e1 and e6. The second row would consist of pawns that could move one square without en passant capturing. Pawn promotion would be for lost pieces, or, if no piece has been lost upon promotion, it becomes a non-royal knight.

George Svokos wrote on 2009-11-06 UTCExcellent ★★★★★
If the royal night is too easy to checkmate in my game design in the below comments, then it can be changed to an equus rex/crowned knight to give it the king moves in addition to the knight moves.

H. G. Muller wrote on 2010-01-04 UTC
The number of WinBoard-compatible engines that can pay Knightmate now has risen to 10:

JokerKM 1.1.14
Lime KM 62
Faile KM 1.4.4
CCCP Knightmate
Fairy-Max 4.8v
Dababba 2.62g
Gerbil-KM 2.0
VanillaChess KM 1.4.4
Fimbulwinter KM 5.00

High time to hold another computer tournament. I therefore started 'The Knightmate Challenge 2010'. The games can be watched live on the web page:

Anonymous wrote on 2010-07-09 UTC
What if knightmate army is played against FIDE army? Who have more chances?

H. G. Muller wrote on 2010-07-10 UTC
The FIDE army is stronger. I know for sure that Commoners are weaker than Knights in a FIDE context. The Royal Knight also seems weaker than a normal King: it is very vulnerable to being chased over the board by a Queen, and subsequently mated, and it is not able to defend its own Pawn shield like a King does.

It is difficult to quantify this, however. I tried to play the two armies against each other with Fairy-Max, but it crashes very often because of mutual perpetual check. (This is something Fairy-Max cannot handle, as it extends checks always by 1 ply, so mutual perpetuals leed to infinitely deep search. This is already a problem in games like Xiangqi, where the Cannons allow you to set up mutual perpetuals. But when the two Kings move differently, this becomes really frequent. The Kings just have to approach each other, and bang...! They start to attack each other directly.)

Daniil Frolov wrote on 2010-08-07 UTC
I'm not sure that commoners (non-royal kings) are weaker than knights: they are neither colorswitching, nor colorbound and can checkmate lone FIDE king (not knightmate king!) with help of any other piece.

David Paulowich wrote on 2010-08-09 UTC

This table of piece values states that the endgame value of a COMMONER is halfway between a Knight and a Rook in FIDE Chess. Betza once pointed out that replacing the White Knights with Commoners in the initial setup can lead to a difficult game, while replacing the White Bishops with Commoners allows the White pieces to develop more quickly.

Ralph Betza also says on this web page: 'Surprisingly enough, a Commoner (a piece that moves like a King but doesn't have to worry about check) is very weak in the opening, reasonably good in the middlegame, and wins outright against a Knight or Bishop in the endgame. (There are no Commoners in FIDE chess, but the value of the Commoner is some guide to the value of the King).'

My [2006-01-10] comment here shows that Royal Knight and Commoner can sometimes checkmate a Royal Knight. This example also works with FIDE Knight and King against a Royal Knight.

H. G. Muller wrote on 2010-08-13 UTC
Indeed. So those dreaming up the table, as well as Betza, were wrong...

Set up any symmetric position with Kings + 3-6 Pawns on either side, and add one or two Commoners for one side, and one or to Knights for the other, in mirror-imaged positions, and then take any strong Knightmate engine to play out the game a few hundred times. If you can find any position where the Commoners would have the upper hand, I would be interested to see it.

Give the other side two Bishops in stead of two Knights, and the Commoners are completely crushed...

Charles Gilman wrote on 2010-08-14 UTC
If this piece is stronger (on 8x8) than the Knight or the Bishop that's perhaps another reason for calling it a Prince rather than a Commoner, but I digress. Certainly it has the 'major piece' quality of being able to Checkmate with the assistance of the same player's King. The implication of this interesting in that any piece of whose move the Prince's is a subset is also a major piece. Thus promoting a Bishop in Shogi is a big deal indeed. Note that my illustration of how the Harlequin in Commedia dell'Arte Chess is a major piece uses its one-step moves rather than any of its longer-range odd-step moves.

25 comments displayed

Earlier Reverse Order LaterLatest

Permalink to the exact comments currently displayed.