The Science Behind Sacrifices
Building on our Positional Sacrifices article, this piece examines the deeper question: what makes a sacrifice objectively correct? Why do some sacrifices win brilliantly while others are just blunders with aspirations? The answer involves a compensation framework that every strong player uses, whether consciously or not.
The three questions
Before sacrificing, answer these three questions:
- What do I get? (King exposure, development lead, strong outpost, passed pawn, attacking lines.)
- Can my opponent neutralise the compensation? (Can they return material, consolidate, or trade attacking pieces?)
- What happens if my attack stalls? (Am I left with a lost position, or is there a fallback plan?)
A sacrifice is objectively sound when the answers are: "significant compensation," "not easily," and "I still have play."
Types of compensation
- Immediate compensation: Checkmate or material recovery by force, this is a tactical sacrifice.
- Medium-term compensation: A lasting initiative (5–15 moves of sustained pressure).
- Long-term compensation: Structural advantages, dominant pieces, or a permanent weakness in the opponent's position. See positional sacrifices for the long-term type.
Why engines misjudge sacrifices
Chess engines evaluate positions by material count + positional factors. In the early nodes of a search, a sacrifice looks like a material loss, the engine hasn't searched deep enough to see the compensation. This is why engines sometimes take 30+ seconds to "come around" to a sacrifice that a GM plays instantly. The GM is pattern-matching to known sacrifice themes; the engine is calculating from scratch.
Practice evaluating sacrifices
Study sacrifice games from the legends archive, Tal, Kasparov, Shirov, Nezhmetdinov. For each sacrifice, answer the three questions before seeing the result. This trains the evaluation framework that separates speculative sacrifices from objectively sound ones. Solve sacrifice puzzles to practise the pattern.