An AIC (*Alternating Inference Chain*) is a flexible chain of strong and weak inferences. It can hop across digits and pattern types — X-Chain and XY-Chain are special AICs, and many named patterns are “just short AICs.”
Before this: X-Chain, XY-Chain, and many wings/fish endings are special AICs.
Why it works
Strong inference: if A is false, B is true. Weak inference: if A is true, B is false.
Alternating them from end to end means the two strong endpoints cannot both be false. Anything that sees both endpoints is eliminated — same “stung either way” idea, with more kinds of links allowed.
How to spot it
Prefer chains that are strong at both ends. Think: if this end is false → the other end is true.
Demonstration
Follow alternating strong and weak inferences across the grid. Prefer strong at both ends.
The green cells outline an AIC involving 2.
The chain proves the two ends cannot both be false — if one end is false, the other must be true.
Eliminate candidates that see both ends (red) (here 2).
Example 1
Trace an alternating inference chain.
This AIC yields eliminations on 2.
Remove the marked candidates (red) (here 2). Both ends cannot be false, so shared neighbors are stung.
Example 2
Trace an alternating inference chain.
This AIC yields eliminations on 7.
Remove the marked candidates (red) (here 7). Both ends cannot be false, so shared neighbors are stung.