The Lewsearch Report · Issue 04
Issue 04Pre-registered · resolves Jan 8, 2027

225 Pre-Registered Predictions for 2026–27

225 questions frozen with sources before fielding; hashes c73a440e and 801ed3f3.

September 1, 2026 · 225 questions · n=1,000

Methods

What. 252 questions were drafted. 27 forecast-framed items (holiday Over/Under wording) were withheld before publication and rewritten as respondent questions for Issue 05. The published book is 225. The withhold came after seeing predictions and before any outcome. Editorial note, September 9, 2026: Issue numbering changed after publication. These 27 questions remain withheld for a future Issue and are not part of Issue 05.

Panel. n=1,000 synthetic respondents per question. National items use a U.S. adult draw. State items use that state's census panel. The District of Columbia is excluded from the state map, as pre-registered.

Aggregation. Each simulated person states chances on the options; a small floor keeps no option at exact zero. No further calibration.

Scoring. Every one of the 225 items is scored. None are dropped after the fact. We publish how many points we miss the poll's shares, and whether we pick the same winner as a dummy that always picks the first listed option. Each item is graded as its source publishes, through January 8, 2027.

Comparison arm. A comparison run that used letter answers is archived and will be scored at resolution. It is not the headline. The published book is the version where each person says how likely each answer is, plus a small safety rule so no option sits at zero.

Hashes. c73a440e commits the questions, options, resolve dates, and sources, and excludes distributions. 801ed3f3 commits the same book after Lewis option shares were filled.

Showing 225 of 225

Data

Aggregates JSON (public)