Of course in general it's "just a sacrifice," but specificically it's a sham sacrifice (part of a winning combination leading to mate) and even more specifically a decoy and deflection sacrifice.

I suppose a more casual modern convention might be to just call it a forcing move leading to mate. However, I find it useful to identify and classify such tactics more specifically as it helps me to find them in puzzle and game situations.

White to play (20.?) "Very Easy"

By contrast, a sham sacrifice could be when you give up material in order to guarantee a forced win. Giving up material is a sacrifice, but there is no hardship involved because the win is assured.

It's a bit like the masochist who liked a cold shower every morning. So he took a warm bath instead...

<patzer2>'s use of "sham" is a foible, a trivial personal quirk that he chooses to adopt and which neither deserves to be attacked nor needs to be defended. Perhaps he employs it because he uses an engine to check and therefore always knows whether an offer of material is sound or unsound.

But there is a deeper underlying question. In a complicated position, when a GM gives up material in return for tactical advantage, does he do it because he has analyzed the ensuing situation and calculated that it wins for him in all lines? Or does he do it at least partly by "intuition," i.e., because he evaluates it to be favorable on positional grounds?

I can see the point in distinguishing between sacrifices which win by force and sacrifices where you have generally given something away.

This position, with white to play, is trivial:

click for larger view

Sure white gives up his queen, but he wins the game by force. In strict terms, this is a sacrifice. But in reality white is not losing anything, as mate follows swiftly afterwards.

By contrast, a "true" sacrifice happens when you give something up without immediate benefit. The gambit pawn in the King's gambit is a sacrifice, because you have no guarantee that you will get it back. Instead, white is accepting one weakness (the loss of a pawn) in return for other benefits (usually gain of time and the centre).

Using the same term for both kinds of sacrifice is a little unsatisfactory. It's a bit like saying that a rustbucket and a new ferrari are both cars. Strictly speaking it's true, but it doesn't do justice to the difference between them. That's why we invented words like "sports car" and "supercar".

Your trivial example of a back-row mating combination is obvious to you and me, but it might not be obvious to a beginner. Likewise, some combinations that lead to a forced mate may be obvious to Carlsen and Anand, but not to us. But -- and this is the important part -- all of these are logically equivalent. There is no difference in principle between a move that almost anyone can analyze as winning, one that only a strong player can analyze, and one that no one (yet) can analyze to a conclusive end.

A chess logic pedant would argue that any position is a forced win, forced loss or dead draw with best play. If two equally omniscient beings sat down to a game, there is only one possible outcome from any given position. Then chess stops being a game and becomes nothing more than a logic puzzle to be "solved".

Under that scenario, many chess terms become redundant. We would stop talking about "advantage" to one side or another, because any intermediate position would simply be an inevitable step towards a pre-determined outcome. There would be no element of risk, of positional strengths or weaknesses, no sense of attack and defence. And, yes, all sacrifices become equally trivial under this way of thinking.

On the other hand, a linguistic pedant might argue that sacrifice has to involve giving something up in return for a different kind of reward. In other words, there needs to be an element of hardship or loss willingly endured. An early civilization might sacrifice a goat in the hopes of appeasing the gods and getting a good harvest. A good harvest is worth more than one goat, but you do have to say bye-bye to Billy. And there are no money-back guarantees with this sort of transaction. You just have to hope that the gods (a) exist (b) are listening and (c) are not annoyed with you for some other transgression.

By contrast, swapping Billy boy for a bag of fertiliser isn't a sacrifice, because you have a more certain chance of growing beans next spring.

Seeing as we are not omniscient, especially when sitting OTB in some draughty church hall with the clock ticking, this suggests that there are different degrees of sacrifice. At the most trivial, there are sacrifices where a mate is guaranteed shortly afterwards. Then we graduate up by degrees through sacked pawns in the king's gambit all the way through to that queen sac in Fischer's game of the century against Byrne.

I have never seen anyone other than <patzer2> use the term sham sacrifice. But I have seen chess authors talk about genuine sacrifices and true gambits - for example, the QGD is sometimes described as not a true gambit because white can always recover his c4 pawn. And if chess authors talk about true sacrifices or genuine sacrifices (bye bye Billy), surely we need a term for the opposite - "not a true sacrifice".

I think so. If you knew that a sacrifice was certainly going to be repaid, where's the hardship? For me, a sac has to hurt the person making the sacrifice in some way. The risk of course is that we splash billy's innards across the stones of our pagan altar, yet still we get a rubbish crop next spring. Not good.

<When I make a sacrifice in a chess game, I am currying favor with Caissa. Does this mean that if I play the Danish gambit I will always win?>

<once> <I have never seen anyone other than <patzer2> use the term sham sacrifice. But I have seen chess authors talk about genuine sacrifices and true gambits - for example, the QGD is sometimes described as not a true gambit because white can always recover his c4 pawn. And if chess authors talk about true sacrifices or genuine sacrifices (bye bye Billy), surely we need a term for the opposite - "not a true sacrifice".

My own definition is this - a true sacrifice is when the person making the move cannot be sure that it wins by force or that the material can be recovered. If it's a case of "bye bye Billy", it's a sac. If it's a bag of fertiliser it's just chess.>

One thing that might help answer this is to ask what value the King has. Many books ignore this question, or say that the King has no value, but others say that the King has a value of infinity.

That might help us define what a sacrifice is. If I capture a Rook, but lose a Bishop to do it, nobody calls that a Bishop sacrifice, because I gave up 3 points of material to gain 5. Net +2.

In once's example, where he begins a mate in 2 by swapping a Queen for a Rook, it's also not a sacrifice, only an exchange where he comes out far ahead. In fact, that example couldn't be considered a sacrifice in any sense, because even if Black's g7 pawn was on g6 (so no checkmate), White gives up 9 points of material (the Queen) to gain 10 (the two Rooks).

But when the Queen is involved, terminology is sometimes fuzzy. Some people call it a "Queen Sacrifice" any time you give up your Queen and don't get the other guy's Queen. If you give up your Queen, but get a Rook and a Knight, that is a sacrifice, technically (give up 9 points to get 8), but "Queen sacrifice" makes it sound like a much bigger sacrifice than it really is. Under the standard conversion rate, it's a sacrifice of 1 pawn, which doesn't sound nearly as good.

Now, consider this game that I played yesterday.

1. e4 Nf6 2. e5 Nd5 3. c4 Nb6 4. d4 d5 5. c5 N6d7 7. Bd3 Nc6 8. e6 fxe6 9. Qh5+ g6 10. Qxg6+ hxg6 11. Bxg6++

Is that a Queen sacrifice? If a King is worth infinite, then no. I gave up 9 points of material to win infinite points of material. Net Gain = infinity - 9. A sacrifice is when you have a Net Loss in the foreseeable future.

Now, could the case be made that 10. Qxg6+ was a less-than-best move? I gave up the Queen when I didn't have to. 10. Bxg6+ hxg6 11. Qxg6++ would have won without giving up the Queen. Does that matter? Is one move better than the other? That depends on whether you think infinity - 9 is a lower number than infinity - 3. I think (but I'm not positive) that a mathematician would say that both are the same number. So, if a King is worth infinite points, then the Bishop "sac" and the Queen "sac" are both equal even though they don't appear to be.

Thinking of the King as worth infinite points helps, because when you make a sacrifice, you hope to eventually win the game with it. But you don't necessarily expect to win back the material you sacrificed. Take the so-calld Game of the Century. Fischer gave up his Queen, and eventually won the game, but he never did get that Queen back. Byrne still has his in the final position.

Was Fischer's sac a Queen sac? You could make the case no, because he could have forced a draw by repetition at the time he made it. So, was it a sacrifice when he gave up the Queen? Or did it become a sacrifice when he abandoned the repitition and tried for more?

Consider, that if someone gives up a Knight for no material at all, people call that a piece sacrifice. If they give up a Knight for one pawn, people will probably STILL call it a piece sacrifice, even though you've sacrificed less than that. If you get two pawns, they may still call it a piece sacrifice.