Gap to the claim

Check how far the inference has to jump from what was observed to what is being claimed.

A fund manager's pitch deck shows he beat the market five years running, and concludes: skill. The observation is real. The question is the length of the leap from what was observed to what is being claimed, and this leap is longer than it looks.

The move is to measure the inferential distance between the data and the conclusion. The cue is a modest observation carrying an ambitious claim on its back. Steps: state the observation plainly (five winning years), state the claim (this man has skill, so give him your money), list the rival explanations that fit the same observation (among thousands of managers, luck alone produces some five-year streaks; a strategy may have ridden one regime that just ended), then ask whether the evidence discriminates between the rivals at all. A separate jump hides at the end: even if he was skilled, does past return support the intervention of investing now? When it works, "impressive" gets relabeled "consistent with luck or skill," which changes what you do with your savings. Skip the full audit for small reversible stakes, where a wide gap between data and claim simply doesn't cost enough to matter.