A new set every day
Which headline got more clicks?
Upworthy showed thousands of readers one of two headlines for the same story, with the same photo, and counted who clicked. Pick the winner. Mimiq picks too: it made its calls before it saw any result.
Pick the headline more readers clicked. Mimiq picks too.
Real tests from the Upworthy Research Archive, CC BY 4.0. The readers drawn as faces are simulated.
Where the rounds come from
Each round is one experiment from the Upworthy Research Archive: two headlines for the same story, shown at random to thousands of readers between 2013 and 2015, where one got clearly more clicks (p < 0.01). Only the headline differed. A new set opens every day at midnight, your time.
How the calls were made
Mimiq's method was built on other experiments, frozen and pre-registered before it saw any of these. Its forecast (Claude Sonnet 4.5) reads the two headlines in both orders and says which one more people like the readers would tap; when the order changes its answer, it makes no call. Simulated readers react to each headline on its own. The plain AI is Claude Haiku 4.5, asked once with A shown first.
The full benchmark, misses includedHow a set is scored
A point for each round where you picked the headline more readers clicked. Mimiq scores only when it picked the winner: a no call scores nothing here, while the benchmark counts it as half, like a coin. A set is a game, not a measurement; the benchmark's 1,000 tests are the measurement. Your results are kept in this browser, not in an account.
Data: the Upworthy Research Archive, J. Nathan Matias, Kevin Munger, Marianne Aubin Le Quere and Charles Ebersole, Scientific Data 2021, licensed CC BY 4.0. Readers shown here are simulated.