There has been some discussion on the ChessBase website about the unfairness of stalemate resulting in a draw. I cannot agree with that sentiment. Stalemate should remain a draw! Centuries of endgame theory would be overturned if stalemate was not a draw, as the basic position of K+P versus K with the defending King in front of the pawn would no longer be drawn.
Consider the following diagram, from round 8 of the World Youth under 14 boys, with White to move. White has been defending a Bishop ending a pawn down, and has just swapped the dark squared Bishops, because he knows the pawn ending is drawn:
This is drawn because of the variations 1.Kc4 b3 2.axb3 axb3 3.Kxb3 Ke5 4.Kc4 Kf4 5.Kd5! f5 6.Ke6 swapping all the pawns, or else 4... f5 5.Kd3 Kf4 6.Ke2 Kg3 7.f4 Kxf4 8.Kf2 is a book draw, because of stalemate at the end. Yes, I know that 7.Ke3 also works.
Stalemate usually occurs through carelessness, or through a very clever defence. I saw an example of each case in round 8. In a game on one of the lower boards they had K+R vs K. The final position was wKh1 vs bKh3, with bRg3 causing a stalemate draw. That was extremely careless.
Here is an example of a very clever defence, with Black to move:
Here Black played 1... Re8 threatening Rh8# so 2.Qc7+ is forced, met by 2... Kg6 renewing the mate threat. However, after 2... Kg6 we see that the Kh4 has no legal moves, so White can sacrifice his Queen to achieve stalemate. Unfortunately White blundered and played 3.Qh7+?? Kxh7 4.Kh5 Re1 and had to resign.
Notice that any one of 3.Qb8, 3.Qc8 or 3.Qd8 would be a very clever defence to the mate threat! For example, 3.Qc8! Kf7 (3... Rxc8 is stalemate) 4.Qb7+ Kg6 5.Qxe4+ Rxe4 results in a pretty stalemate.