We publish the probability before the outcome, attach a falsifier and a resolve-by date, then Brier-score it in the open. Hits and misses both. No quiet deletes.
Missed. The strait never fully closed, so a literal reading could call this a hit. We don't. The US resumed its blockade on July 14, and Lloyd's List counted 78 to 95 transits a week in early August, a small fraction of the roughly 125 a day before the war. (Corrected Oct 24, 2026: we first wrote about 100 a day, a figure we could not source. Reuters reported about 125 a day before the war.) We also never published written criteria for this call. We should have.
Missed. Iran struck three ships on July 7, President Trump called the deal over, and by August 16 Iran said it had suspended all its commitments. This call also went out without written criteria. It misses under any reading.
Missed. The energy fund XLE closed at $59.20 on July 22, 10 percent above its June 18 close of $53.77, when we made this call, and stayed above that line on July 23, 24 and 31. We never printed that baseline in the story. Measured from our July 7 publication date instead, it would score as a hit. We score from when the call was made.
Hit, narrowly. From mid-July to late August, broker quotes for a Hormuz transit ran 3 to 10 percent of a ship's value, far above our 2.5 percent line. One July 10 Marsh quote put rates at "anywhere between 2% and 6%" without saying for which ships. We scored the large-tanker quotes. A stricter reading of our own falsifier would score this a miss.