28

The modern rules of chess, except for the stalemating being a draw, are basically equivalent to a game in which the goal is just to capture the king, not checkmate him. It's plausible that in the oldest versions of chess, this was indeed the goal, but since a rational player would then always move their king when in check, and never leave themselves in check, and would only have their king captured when there was no escaping from it, players may have decided that responding to check may as well just be made a rule.

I haven't been able to find any reliable source either confirming or refuting this hypothesis. Is there any?

Rewan Demontay
  • 17,514
  • 4
  • 67
  • 113
Jack M
  • 641
  • 1
  • 7
  • 13
  • 4
    I think capturing the King is still the goal, except the game ends one move before the actual event. Checkmate basically means, "I'm taking it next turn and there's nothing you can do about it." – Tony Ennis Apr 03 '16 at 21:10
  • 5
    Not quite Tony. Unlike having, say, a queen captured, a player in the modern game cannot mistakenly allow his king to be captured. – Hugh Apr 03 '16 at 22:21
  • 16
    @TonyEnnis that almost works, but not quite. Stalemate is the issue. If you can't move without moving your king into check, then in chess this is a draw. However, if the goal were to capture the king then this would be a losing position, because your king would be forced to move into check and would be captured the next move. – N. Virgo Apr 03 '16 at 23:04
  • 2
    @Hugh : in many speed chess variants it can happen. I've seen official speed chess tournaments where games starting with 5 minutes for each player allowed the player to lose if he left the king in check (and the other player noticed). – vsz Apr 04 '16 at 08:11
  • 5
    A comparison to other games in the chess family could be instructive here. In "Chinese Chess" there is also no actual capture of the General - but there is no stalemate, having no legal moves means you lose. – wberry Apr 04 '16 at 15:07

2 Answers2

24

Wikipedia claims that

In early Sanskrit chess (c. 500–700) the king could be captured and this ended the game. The Persians (c. 700–800) introduced the idea of warning that the king was under attack (announcing check in modern terminology). This was done to avoid the early and accidental end of a game. Later the Persians added the additional rule that a king could not be moved into check or left in check. As a result, the king could not be captured,[13] and checkmate was the only decisive way of ending a game.[14]

References [13] and [14] are to Davidson, Henry (1949), A Short History of Chess, which can be consulted on Google Books. I couldn't tell immediately what sources Davidson used.

dfan
  • 14,671
  • 40
  • 67
  • 2
    This is interesting but I'm not sure how reliable this book is. I wish Google Books would let you look at the bibliography. – Jack M Apr 04 '16 at 10:55
  • 1
    I agree that the book seems pretty chatty and not particularly scholarly. The very dedicated chess historian Edward Winter agrees here and lists more books on the history of chess (all of which he finds some fault with). – dfan Apr 04 '16 at 17:19
6

I think you may be right, but I can't find anything to support it. However, if the goal was originally to actually capture the opponent's king, this raises another interesting question: What happens if the goal is to capture the king, moving into check is allowed, and you still have no legal move? For example:

   [fen "3k4/8/8/8/1p1p4/pPpPp3/PRP1P3/NKN5 w - - 0 1"]

If it is White to move, this position is stalemate under modern rules, but if there's no stalemate, and Black has to capture the white king to end the game, does White just skip a move, and then 1…axb2 2.Kxb2, cxb2#, or would they just look at material, position, whatever, and work out who should win, agree to a draw, or just end the game with no result and replay it or something?

hkBst
  • 1,664
  • 12
  • 29
Jivan Scarano
  • 335
  • 2
  • 6
  • Yeah, I missed that. My point still holds about the position though. – Jivan Scarano May 25 '16 at 15:50
  • 2
    It is a good point (hence my upvote and attention), but could be made better perhaps. The position also seems to suffer from being unattainable in a real game. I wonder if there are attainable positions (let alone realistic positions) that fall into this same category of no moves at all (not even once that result in check. – hkBst May 26 '16 at 06:58
  • I think that position would be attainable, although very unlikely as it is obviously completely contrived. I originally tried something similar, but using bishops to block the king from moving, but if the bishops are blocked by pawns, then they can never have moved, so how did the rook get there? Also I had the entire position moved two squares to the right, another two pawns of each colour added on the a and b files, which still works, but makes it even more contrived, unlikely, and unnecessarily complicated. – Jivan Scarano May 26 '16 at 07:27
  • 1
    yes you are right that the position is attainable. – hkBst May 26 '16 at 11:19
  • I don't see how capturing the king being the objective prevents a stalemate rule from existing. The rules could say both that whoever captures the opponent's king wins and that if a player has no legal moves he can claim a draw – David Jul 07 '20 at 06:29
  • The position is attainable. Probably not the most efficient way to go about but here: 1. Nc3 h6 2. Ne4 g6 3. Nc5 f6 4. Nb3 e6 5. Rb1 h5 6. Na1 g5 7. b3 f5 8. Ba3 e5 9. Bxf8 d6 10. Qc1 d5 11. Qa3 Rh6 12. Qe7+ Qxe7 13. Kd1 h4 14. Kc1 Rh8 15. Nf3 e4 16. Ne5 h3 17. Nd3 g4 18. Rb2 Rh6 19. Kb1 Nf6 20. Nc1 c6 21. d3 a5 22. f3 hxg2 23. Bxg2 gxf3 24. Bxf3 Rxh2 25. Rxh2 Qxf8 26. Bh5+ Nxh5 27. Rxh5 Nd7 28. Rxf5 Qf7 29. Rxf7 Kd8 30. Rxd7+ Ke8 31. Rd8+ Ke7 32. Rxc8 Kd7 33. Rxa8 a4 34. Rc8 a3 35. Rb8 b6 36. Rc8 b5 37. Rb8 b4 38. Rc8 c5 39. Rb8 c4 40. Rc8 c3 41. Rb8 d4 42. Rc8 e3 43. Rd8+ Kxd8 – Sascha Baer Mar 15 '21 at 13:04
  • Answer: This kind of analysis would have been rejected as unreasonable play by both sides and therefore the result from this position does not matter. The 16 pawn deadlock is more likely and it's poorly handled until the late twentieth century. – Joshua May 05 '22 at 19:58