Inference to the Best Explanation
Detectives, scientists, and apologists all run the same move: whatever story explains the most, wins.
Grounded in the research on inference to the best explanation
Start with the clues
You don't deduce a verdict, you weigh stories. You've got a pile of facts that need accounting for, and several hypotheses competing to explain them. This is abductive reasoning. Charles Sanders Peirce flagged it back in the 1860s as a third move alongside deduction and induction. Gilbert Harman named the modern version 'Inference to the Best Explanation' in a 1965 paper in The Philosophical Review. It's how a doctor reads symptoms and a detective reads a crime scene.
Rank the rival stories
The whole game is comparative. You don't accept a theory just because it fits the data. You accept it because it fits better than every rival on the table. Harman's point: 'best' is doing real work. A bad explanation that's the only one standing still loses to nothing. You need live competitors, then you grade them. The hypothesis that explains more, more cleanly, with fewer cheats, takes it.