K+N v K+P = sufficient material?

Suppose player X has K+N, while player Y has K+P, and also suppose it's a position where X can force mate, but that fact has not yet been discovered (although it could be upon analysis).  In this position, the flag on Y's clock falls, i.e. Y runs out of time.  But Y claims a draw based on X having insufficient material for checkmate.  Although this is true in the generic sense, it's false in the position.  Does the TD have to determine whether or not X can force mate?  What's the correct ruling?

It isn't whether X can FORCE mate, it's whether there is a checkmate possible, even if one side is playing the worst moves in history. So, in this case, player X shall win. But, he would ALSO win if mate was not forceable, but Y could walk into it.

Thanks.  I like the distinction -- possibility as opposed to a forcing line.  But are you answering for chess rules generally?  I noticed you're from the UK, so I have to ask if you were referring specifically to BCF or FIDE rules.  I'm only concerned with US Chess Federation and not FIDE or BCF rules ... not that I know they're different in this case, but a USCF tournament director just told me that "mating material" has to be judged independently of what the opponent's material is, and since K+N is insufficient (by definition -- even though it's definitely sufficient for mate in this case!), player Y (K+P) has a valid draw claim.  Your answer seems more intuitively correct; it seems to have more justice in it. 

I just want to know if I can ever possibly lose on time when I have K+P v K+N.  You say I can. 

Yes, you could. This rule is true for all governing bodies in chess.

well you can't checkmate with K+N so there was no chance for x to win, although Y has a chance to win if he can somehow promote the pawn. Either way if Y loses on time than he can't say it is a draw

DaPharaoh, two things:

1) You definitely can checkmate with K+N against K+P in certain positions.  See the section on Stamma's mate under "Rare checkmate positions" here:  http://en.wikipedia.org/wiki/Checkmate

2) Even if, as you (incorrectly) say, X had no chance to win due to insufficient material, then it is a draw, not a win for X.  This is what normally happens when someone's clock expires but their opponent can't mate anyway.

I think it's a win if the pawn is a rook pawn. If it's not a rook pawn, I believe mate isn't possible, so then it is a draw.

The USCF director was wrong.

 I think so, and a different USCF TD disagreed with him.  But it sounds like you're saying that whether someone has "sufficient material" depends on what the position is, including what material the opponent has.  He (the first TD) was very clear after consulting the USCF rulebook that K+N was insufficient.  But I agree with you.  One shouldn't be allowed to claim a draw on grounds of insufficient material after letting one's clock expire in the following position:

White (X) to move has N on c4 and K on h1; Black (Y) has Bs on d7 and d8, N on e7, K on e8, and Rs on f7 and f8.

White, with the move and plenty of time, hopes to play 40 Nd6#.  It seems unjust to let Black run the clock and claim a draw based on White's having "only" K+N.

Anyway, despite what that tournament director and anyone else seems to think, what I said is correct. If there is a conceivable method of checkmate appearing on the board by the K+N, even with the other side playing worst play, then the guy with the pawn loses.

Nytik: I know this is certainly true of internet chess... but I'm under the impression that in a tournament situation, a player can stop the clock (before it runs out) and claim a draw from the arbiter should the position be obviously drawn - e.g. you have the pawn in this situation.

No, marvellosity, I am almost certain this applies to OTB chess as well... for example, in the king+rook vs. king+knight scenario, if you have the rook and are about to run out of time, you're better off saccing it!!

Nytik is absolutely correct in his assessment: if mate is possible for your opponent, then you can lose on time. Marvellosity is also correct that you may claim a draw if you can show to the arbiter that you're opponent is not making a winning attempt, but you have to do that before your time runs out (as Marvellosity points out) and you probably have to make a few moves (with the clock ticking) after fetching the arbiter to substantiate your claim.

Mate is always possible with K+N vs K+p. The pawn could promote to a rook, when the following position could arise (note that this would usually take very bad play from both sides, not just the losing side):

Good point, the non-rook-pawn can always promote and help with the mate. Didn't think of that.

A draw claim is automatically also a draw offer, so if the opponent just accepts it then it is a draw.

You can claim a draw if your time is running out and the opponent cannot win in a normal way anymore, or isn't trying to. Details differ slightly between USCF and FIDE. But that's not related to this issue (such a claim is only possible before the flag falls).

Mathijs, I'm afraid you've reversed my original example.  Not to get too technical, but when you say "mate is always possible with K+N v K+P" (which I think is false, stricly speaking), don't you really mean "mate is always possible with K+P v K+N" (which I think is true)?  The difference being that "mate ... with" means the person possessing that (first-mentioned) material mates.  I wasn't asking whether K+P could possibly win -- of course, you're right that the pawn could promote (to a Q, for that matter), but whether K+P would draw after his clock expires even with K+N to move and force mate.  I think these posts have been pretty clear on the answer to that -- i.e. no, the player with K+P loses because his draw claim is invalid.

Am I right about this?  My claim is that mate with K+N v K+P isn't alwayspossible.  That is, there are positions of that sort where the K+N can't win.      

Well, there will be a few where it isn't possible (immediate stalemates, positions where a capture is immediately forced...) but they'll be extremely rare. What sort of position are you thinking of?

You should remember that we're discussing whether mate is possible -- i.e., with the opponent helping, if necessary. This is not about forced mate being possible, but rather whether there exists some sequence of moves that ends in the first player mating.

And in that case, the answer is almost always yes, as he showed - the pawn can promote to a rook, the king can walk into the center, rook goes next to it, and then the knight can mate.

So mate is still possible, and the side with the pawn loses if he runs out of time.

No there aren't, jk: because the pawn can promote and mate is always possible as has already been described.

It seems if you're extremely low on time but in the 'strong' position (e.g. having the rook vs the knight) then you're best claiming a draw with the arbiter.

I believe on playchess.com there is even a function that if the opponent only has certain material left (e.g. a knight) that if you press the Offer Draw button, the server awards the game a draw - whereas if you run out of time, you lose.

That Playchess rule is interesting.

Under FIDE rules for Blitz, a draw claim like this wasn't possible, play would go on until a flag fell. But since July 1st, the rules have been changed, and now claims like this are possible if there is sufficient supervision (like in professional tournaments, where they can have one arbiter per board).

I guess a computer deciding these sort of things could be called supervision.

You're question was is it a draw even though the knight side has a forced mate and his opponent's flag has fallen. The answer Nytik was: "no it's not a draw, it's a win for the knight. In fact it's even a win if there is no forced mate at all, but mate (for the kight's side) is possible through some sequence of moves." Scarblac then made the interesting but erroneous point that in positions of K+N vs K+pawn not on a rook file mate (for the knight's side) is not possible through any sequence of moves. I corrected this error by showing that mate (for the knight's side) is always possible through a sequence of moves involving a minor promotion.

Edit: my post crossed some others. Interesting point about the immediate stalemates and forced captures, scarblac. I had not considered that.

So, you are incorrect, mate is always possible in K+N vs K+p (but only through very bad play on both sides, there is usually no forced mate). See my previous diagram for how it's done.

Right, I had a look at the FIDE rules to clear this one up, seems everyone is about right. You lose if you flag and your opponent *can* mate you, but you can claim a draw: (my emphasis)

Mathijs, I see your point now.  I saw your diagram earlier, but I didn't click through the moves and just thought it was an example of a back-rank mate with the rook, haha. 

Still, not every situation with K+N can possibly turn into mate against K+P.  To show this, don't we just need an example with K+N v K+P that can't be checkmate for the side with the N?

e.g. White has K on c6 and N on b1; Black has K on c8 and P on c7.  White to move touches his N.  Checkmate is impossible as stalemate is inevitable, but this is still a board position with K+N v K+P, therefore it qualifies as a counterexample.

Should we now distinguish between "position" and "situation"?