Analysis of the oldest known shatranj game

Sort:
chrisbao

I've vertically mirrored the game so that the ferz starts on e1, but this does not functionally change anything.
White is As-Suli, an ancient master, said to have been one of the greatest of all time at the game. Black is Al-Lajlaj, another great master of his time.

1. c3 c6 

2. c4 c5 

3. Nc3 Nc6 

4. b3 b6 

5. Rb1 Rb8 

6. a3 a6

7. d3 d6

8. b4 cxb4

9. axb4 b5

10. cxb5 axb5 There is not much to say about these first moves, all three superhuman engines (Stormphranj, Fairy-Stockfish, my engine Prolix) think the position is completely equal here, even if they would have chosen different tabiya.

11. e3 e6

12. d4 d5

13. Bd3 Bd6

14.Nxb5 Kd7?! A slight mistake, all three engines recommend Bxb4 instead.
On white's next move, the three engines all recommend Qd2 (Nxb4 Nf3 f6 Ne1) instead of f3, with a slight advantage for white.

15. f3?! Nxb4 The position is back to being dead equal.

16. Kd2 f6

17. e4 e5

18. g3 g6

19. Ne2 Ne7

20. Qf2 Qf7

21. Qe3 Qe6

22. Nec3 Nec6

23. Ba3 Ba6

24. Bc5 Bc4 Black can play f5 here first, a unanimous suggestion.

25. Rhf1 h6?! f4 for White immediately, was also an option. For reasons unknown to me, h6 is an inaccuracy, but Ra8 (which also appears to ignore f4) maintains equality.

26. f4 Rhf8? Black rather should have opened the position with exd4 and dxe4. The played move completely ignores White's next move.

27. f5! gxf5 dxe4 (majority, but not unanimous, suggestion) was still possible here, though the played move also should be fine.

28. Bxf5+ Kd8 Black has a tough defensive task ahead, but the position should still be holdable.

29. exd5 Nxd5? After Qxd5, black should have at least some compensation for the lost alfil.

30. Nxd6 Rxb1? Amusingly enough, Nxc3 without Rxb1 is recommended by the engines.

31. Rxb1 Nxc3? Ndb4 or exd4 were alternatives.

32. Kxc3 But now Black is simply down an alfil, and will lose more material in a few moves.

Pokshtya

Monument to al-Suli in Ashgabat, Turkmenistan.

samuelebeckis

As-Suli was from Iran or Turkmenistan?

It is not clear to me! sad

coolthing

Nice analysis!

25. ... Ra8 may be better because 26.f4 allows 26. ... fxe4, where Black may soon infiltrate with the rook. For instance, 26.f4 fxe4 27.Nxe4 Ra2+ will win the d3 alfil or worse: 28.Kc3 Nd5+! 29.Kxc4 Na5#.

Pokshtya
samuelebeckis wrote:

As-Suli was from Iran or Turkmenistan?

It is not clear to me!

Abu Bakr bin Yahya al-Suli

https://en.wikipedia.org/wiki/Abu_Bakr_bin_Yahya_al-Suli

samuelebeckis

Thanks my friend for the link. I did read the article on Wikipedia but tbh I think there is a mistake, because Gorgan currently may refer to the city, or the district, or the river of the same name (and its drainage basin). Well in any case they all are located in Iran, not in Turkmenistan, however the article may also probably refer to the historical region of Gorgan (Middle Persian: Gurgan, Old Persian: Varkana, Greek: Hyrkania). This region of ancient Persia also included the steppes in southern Turkmenistan, therefore as-Suli could have been born in ancient Gorgan, yes (i.e. Hyrkania), but still falling in present-day Turkmenistan. In other words I'm not sure Wikipedia redirects to the right meaning of Gorgan, and that statue in Ashgabat, capital of Turkmenistan, is probably a clue of that.