The auction is heating up. Opponents reach 4 hearts and you're fairly certain they're making it. Partner has shown spades. You pull out the 4 spade card, accepting two or three down, gambling that the penalty is cheaper than what they'd score.
Done right, a sacrifice gains boards and frustrates the opposition. Done wrong, it gifts them a penalty they didn't earn. The line between those outcomes is mostly math.
What a Sacrifice Actually Is
A sacrifice is when you bid to a contract you expect to fail because the penalty is smaller than the opponents' score. Paying -300 instead of conceding +620 is a genuine win in a relative-scoring game like bridge.
The tricky part is “expect to fail.” If the opponents weren't actually making their contract, you've committed a phantom sacrifice. You pay a penalty and deny yourself a plus score on defense. It's the most expensive competitive bidding error in bridge, and it happens more often than players admit.
Vulnerability Is the Whole Equation
Not-vulnerable doubled undertricks are forgiving: 100 for the first, 200 for the next ones. Vulnerable doubled, the costs jump: 200 for the first, 300 for the second and third. This difference defines whether a sacrifice is smart or wasteful.
The classic conditions for a good sacrifice: you're not vulnerable, they're vulnerable. Their making game scores 620. Two down doubled not-vul costs you only 300. You've saved 320 points. In a field comparison, that swing matters.
The worst sacrifices happen when both sides are vulnerable. Three down doubled costs 800. Their vulnerable game might only be worth 620. You've paid 180 extra points instead of saving any. Always check vulnerability before you consider sacrificing.
Matchpoints and IMPs Require Different Thinking
At matchpoints, the sacrifice decision depends entirely on what the rest of the room does. If most tables reach 4 hearts not-vulnerable and make it, any sacrifice under -420 gains boards against those tables. But if the opponents are pushing too hard and half the field never reached game, your sacrifice hurts you against every table that defended quietly and collected +100.
The field problem is real. You're not just deciding whether to sacrifice against this particular pair. You're deciding whether to sacrifice against what the whole field bid and played. That's a harder calculation, and it requires a read on whether this pair tends to overbid.
At IMPs, the calculation is cleaner. Not-vulnerable two down doubled (-300) against their vulnerable game (+620) saves 320 points: roughly 8 IMPs. But if you'd have set them and gotten +100 on defense, the phantom costs you 400 points, about 9 IMPs in the wrong direction. The arithmetic slightly favors defending when you're uncertain. For a broader look at how scoring format changes decisions, IMPs vs. Matchpoints: How Strategy Changes Everything is worth reading alongside this.
Spotting a Phantom Before It Happens
How do you know if they're making? Often you don't, precisely. But certain signals say “defend.”
If you hold a trump trick, slow down. Sacrifice math assumes they'll make. A trump trick often means they won't. Same logic if partner's bidding suggested shortness in their suit, or if your count of the hand suggests a defensive ruff or side-suit winners. A quiet pass and the right opening lead sometimes gains more than any sacrifice.
Ask yourself: “If I pass and lead, do I think we can beat this?” An honest yes is usually enough reason to sit back. Not every competitive auction needs you to take the last bid.
Checking Whether You Were Right
The frustrating thing about sacrifice decisions is the delayed feedback. You won't know whether the sacrifice was correct until results go up. Most players file it under “felt right” and move on, which means they repeat the same errors.
Comparison scoring fixes this. Bridge@Home shows what happened at other tables holding the same cards: did they sacrifice too? Did the defense collect, or was the game making everywhere? Your -300 beat 70% of the room? Right call, even if uncomfortable. Everyone else got +100 on defense? That's a phantom, and now you have specific data on what to look for next time.
This is what separates players who actually improve from those who just accumulate experience. Knowing the outcome at one table tells you almost nothing. Knowing what happened across the field tells you whether your decision was correct.
The Short Version
Sacrifice when: you're not vulnerable, they're vulnerable, their contract is near-certain, and the penalty saves meaningful points.
Don't sacrifice when the vulnerability math is thin, you hold a trump trick, partner's bidding suggests a defensive entry, or you'd be competing against a contract you'd likely set.
The most common error isn't miscounting the arithmetic. It's sacrificing against opponents who weren't going to make it. Pause before you pull out that card. Ask the honest question. If the answer is genuinely uncertain, defend and take the plus.