Arabian mate not in a corner.

[Ajedrecista]

Hello:

I played other interesting-to-me 5+0 blitz game today where I was able to give an Arabian mate not in a corner! Please do not expect top level:

[pgn][Event ""]
[Site ""]
[Round ""]
[Date "2026.08.12"]
[TimeControl "300+0"]
[Result "0-1"]
[PlyCount "50"]

1. d4 d5 2. Nf3 Nf6 3. Bf4 e6 4. Nc3 Bd6 5. Bxd6 Qxd6 6. e3 O-O 7. Bd3 c5 8. dxc5 Qxc5 9. Ng5 Nc6 10. h4 a6 11. Qe2 Bd7 12. O-O-O Rac8 13. g4 b5 14. f3 Na5 15. Nh3 Nc4 16. Rde1 b4 17. Nb1 a5 18. g5 Ne8 19. h5 a4 20. g6 h6 21. gxf7+ Rxf7 22. Bg6 Rf6 23. Nf4 Nxe3 24. Nd3 Qxc2+ 25. Qxc2 Rxc2# 0-1[/pgn]

I missed the simple, much better move 16.- ..., Nxb2 after 16.- Rde1?? I played 23.- ..., Nxe3 with two lines in head: the correct 24.- Qxe3, Qxe3+; 25.- Rxe3, Rxf4 (winning a pawn); or the disgusting 24.- Qxe3, Qxc2+??, expecting a checkmate, but the white bishop on g6 defends c2. Luckily, my opponent played 24.- Nd3?? and I spent few seconds to recognise the pattern and play it.

Later, I googled 'Arabian mate in the middle of the board' and I found some examples of games, for example this one of 1947 (via a Chessgames compilation of Arabian mates) and this recent casual game.

The Wikipedia article says '[...] the Arabian mate is also an important topic in the context of history of chess for being mentioned in an ancient Arabic manuscript dating from the 8th century CE. [...]'. Google IA pointed me to a book of 1257 (link) and a copy of facsimiles of a lost book around the 9th century with modern diagrammes (link). I tried to find early documented occurences of the Arabian mates in those books, but neither the Google IA nor me were able. I found that Yet Another Chess Problem Database (YACPDB) has few problems (not Arabian mates) from 'lost manuscripts' (Al-Adli, Al-Khhadim and As-Suli), just typing 'Lost A' at 'Source (issue/id)' search box. I think that the copy of facsimiles is from Al-Adli according to Google IA. There are problems of other manuscripts at YACPDB, just typing 'manusc' and 'manusk'.

------------

To make this thread more computer chess related, I wonder if chess engines give this checkmate pattern not in a corner. I guess that some info could be extracted with PGN-extract tool. I would like to see some examples, expecting more if games are played up to checkmate, when engines are not paired Elo-wise and forget about top vs top engines.

Regards from Spain.

Ajedrecista.

[Ajedrecista]

Hello:

I wondered yesterday how soon an Arabian mate can be reached from the starting position, regardless of the mated king being in a corner (as usual) or not. I googled this and did not found anything, surprisingly. Then, I asked the IA Mode of Google and you cannot imagine the bunch of IA hallucinations and illegal moves in its tries to compose a shortest proof game (SPG).

So I had to think on my own, as a chess riddle. I got a SPG of 6.0 moves (12 plies) in few minutes, with the following idea (I share the version that is first in alphabetical order in English algebraic notation):

[pgn][Event "SPG of Arabian mate"]
[Site ""]
[Round ""]
[Date "2026.08.14"]
[Result "0-1"]
[PlyCount "12"]

1.e3 h5 2.Ba6 Nc6 3.Qxh5 Nd4 4.Ne2 Rxh5 5.O-O Nf3+ 6.Kh1 Rxh2# 0-1[/pgn]

I think that the fastest way is that the mated king must end in a corner, to minimize rook moves from the mating side; and that my SPG is optimised in this sense.

Any improvements of my SPG (5.5 moves or less) are welcome!

Regards from Spain.

Ajedrecista.

[peter]

Any improvements of my SPG (5.5 moves or less) are welcome!

Hi!
I don't think, there'd be such, at least Poypeye didn't find one, with 6 moves only transpositions to your line as well.
Yet, if you'd rather want another one, not in a corner, here's a 6mover to such, again Popeye agrees with SPG here (no solution with 5+1/2 nor with 5)

[pgn][Event "Talkchess"]
[Site "?"]
[Date "2026.08.07"]
[Round "?"]
[White "Ajedrecista, J.."]
[Black "SPG, Arabian"]
[Result "*"]
[ECO "A00"]
[WhiteFideId "-1"]
[BlackFideId "-1"]
[PlyCount "12"]
[GameId "2343514568663077"]

1. e3 h5 2. Ba6 Nc6 3. Qxh5 Nd4 4. Ne2 Rh6 5. Qf3 Rxh2 6. O-O Nxf3+ *[/pgn]

[peter]

Sorry for that nonsense, can't delete it anymore. Simply changed the final position without checking properly, if it was mate at all, which of course it wasn't.

Yet the SPG is still a valid one

[peter]

And in 7moves it yet would work fitting to the title of the thread too:
[pgn][Event "Talkchess"]
[Site "?"]
[Date "2026.08.07"]
[Round "?"]
[White "Ajedrecista, J."]
[Black "SPG, Arabian"]
[Result "0-1"]
[ECO "A00"]
[WhiteFideId "-1"]
[BlackFideId "-1"]
[PlyCount "14"]
[GameId "2343514568663077"]
1. e3 h5 2. Ba6 Nc6 3. Qxh5 Nd4 4. Qd1 Rxh2 5. Qf3 Rxg2 6. Ne2 Rh2 7. O-O Nxf3# 0-1[/pgn]

[Ajedrecista]

Hello Peter:

And in 7moves it yet would work fitting to the title of the thread too:
[pgn][Event "Talkchess"]
[Site "?"]
[Date "2026.08.07"]
[Round "?"]
[White "Ajedrecista, J."]
[Black "SPG, Arabian"]
[Result "0-1"]
[ECO "A00"]
[WhiteFideId "-1"]
[BlackFideId "-1"]
[PlyCount "14"]
[GameId "2343514568663077"]
1. e3 h5 2. Ba6 Nc6 3. Qxh5 Nd4 4. Qd1 Rxh2 5. Qf3 Rxg2 6. Ne2 Rh2 7. O-O Nxf3# 0-1[/pgn]

In 7.0 moves? I raise my bet and win in 6.5 moves!

[pgn][Event ""]
[Site ""]
[Round ""]
[Date "2026.08.16"]
[Result "1-0"]
[PlyCount "13"]

1.h4 e5 2.Nc3 Qxh4 3.Nd5 Qf6 4.Rxh7 Ba3 5.Rxg7 Ne7 6.Rh7 O-O Nxf6# 0-1[/pgn]

That Qh5-Qd1-Qf3 manoeuvre was telltale that could be shortened changing the side (just Qh4-Qf6). To be fair, full credit for you.

Regards from Spain.

Ajedrecista.

[peter]

Hi Jesus!

To be fair, full credit for you.

Not really, considering the thematic try was yours from the very start anyhow and I could have had the idea of changing sides to spare a ply by myself finally yet still