> cat novig-favorites.md
Heavy favorites on Novig
Hypothesis
Heavy favorites on Novig win more often than their price says. Rest as a maker and collect the difference.
Test
On the outcomes that could be scored, the win rate beat breakeven by 3.95 points, interval 0.44 to 7.01. But 43.7% of outcomes could not be scored, and not at random: the borderline cases that could be scored lost. Because of what went missing, the audit's luck correction across all 53 ideas treats this interval as void, so it does not beat luck. The tested price band is withheld.
Cause of death
Blocked. The scored part looks like an edge. The unscored part is the part that would decide it, and the pattern in what went missing points the wrong way. The audit's check of this cause: consistent with its own record.
gateNOT ESTABLISHED
causeblocked
evidenceresult table with interval
auditone of 53 ideas · plan frozen 2026-10-17 · luck-corrected
the number+3.95 points, interval 0.44 to 7.01, scored outcomes only
the audit's checker compared 270,232 values across all 53 ideas (run 3, 2026-10-25); this number is not one of them: it is the gap between two values it did compare
Lesson
Missing data that is missing for a reason can manufacture an edge.
This site is not advice and gives no signals or trade ideas.
fig. 1 Above: how far the win rate beat breakeven on the scored outcomes, with its interval, in points. Below: the share of outcomes that could not be scored. The tested price band is withheld.
| Measure | Value | Interval |
|---|---|---|
| Win rate over breakeven, scored outcomes | +3.95 points | 0.44 to 7.01 |
| Outcomes that could not be scored | 43.7% |