Satire. Every score on this site is made up. The tweets are real.Read the disclaimer →
Trust Me Bro
ModelsBenchmarksIncidentsIn the WildMethodologyChart Crime Studio
Tools · Beta

Chart Crime Studio

Turn a 2.7-point lead into a generational leap. Drag the y-axis baseline until the difference looks like a launch post. Tufte's Lie Factor is computed live, so you know exactly how much you are lying.

Figure 1

TrustMeBench Pro (Verified)

Accuracy, %, higher is better

*Our model: pass@64, majority vote, internal harness, temperature tuned per question, excluding the 31 questions it got wrong. Competitor: as reported by a user on Reddit.

Data
Y-axis starts at 0.0
HonestLaunch post
Enhancements
Post on X ↗
Actual difference
+2.7 pts
+5.4% relative
Visual difference
1.05×
our bar vs. theirs
Lie Factor (Tufte)
1.0
1.0 is honest
Verdict
Honest
Nobody will share this.
Theory

The Lie Factor

Edward Tufte defined the Lie Factor as the size of an effect shown in a graphic divided by the size of the effect in the data. A value of 1 is honest. Values above 1.05 are, in his words, substantial distortion. Values above 10 are a product launch.

Lie Factor = effect shown in graphiceffect in data

where effect = (ours − theirs) / theirs, measured on bar height for the graphic and on the score for the data.

Exhibits

Found in the wild

Real posts about real charts. The studio above was built so you can reproduce these results at home.

introducing my new benchmark, "AI models randomly sorted on an X and Y axis"
introducing my new benchmark, "AI models randomly sorted on an X and Y axis" https://t.co/bV4lho8Z4e
♥ 4.7KOpen on X ↗
what in the chart crime
what in the chart crime https://t.co/tQsjck6pNN
♥ 1.6KOpen on X ↗
another day, another chart crime committed by a frontier lab
another day, another chart crime committed by a frontier lab https://t.co/Oczhfr5qpn
♥ 445Open on X ↗