Measured like research, written for people who follow a field.
Questions come from practice. Methods come from research. Results go back to practice, with the misses left in.
01
People who follow a field for a living
Journalists, analysts, policy and communications teams. A weekly note says what gained share, what is only repeated, what attention did, and what no second series confirms. It does not say what the public thinks.
Fixed corpora, frozen weeks, rules filed before the results, and every miss kept. The method page lists the measures, the data dictionary and the validation tests, so a reading can be checked against its sources.
Every claim has a fixed sentence, a named measure, a deadline and a simple baseline. A script resolves it against a public source. The log shows what held and what missed.
A desk that does not own a corpus does not file. Elections is the first field and the house method. Hindsight and Murphy have fixed their sources and collect every day; each publishes weekly notes once a full week of live collection is frozen, and files claims only after its readiness check passes. Buzz opens last.
Owns: Clinical and hospital language, quantified versus adjectival claims.
Status Measuring Notes Weekly, published on Tuesdays
The claim log
Claims are filed, then resolved or missed.
A claim is one sentence about a field, with a named measure, a deadline, a kill condition and a simple baseline. Past the deadline is a miss. Misses stay up. No claims have been filed yet. An empty log is the honest starting point.
Claims by desk · checked Oct 24, 2026 at 01:58 UTC
Desk
Open
Resolved
Missed
Void
Elections
0
0
0
0
Hindsight
0
0
0
0
Buzz
0
0
0
0
Murphy
0
0
0
0
Election forecasts published before the claim log opened stay on the election pages with their own error tables. Open the claim log →
Closed experiment
The practice-trade test is closed.
This site used to run a practice-trade experiment. It has ended. Items still waiting at closure were not scored and are not part of the claim log.