Gemini 3.6 Flash gets 0.2% of high-frequency video event counts correct, and more frames make it worse.

Trace-grounded parametric profiling (arXiv 2608.06361) built 2,190 controlled videos varying event count and frequency while holding rendering fixed, each with an executable event trace for timestamp-level scoring instead of final-answer grading. At an 80% reliability threshold, Gemini 3.6 Flash reliably counts persistent state transitions up to 12 events at 0.5 and 1.0 Hz but has no reliable positive-count region for transient blinks. In the high-count high-frequency regime, 0.2% of final counts are correct and 18.1% of true events recovered. The kicker: raising sampling rate lifted Bounce Ball accuracy from 19.6% to 29.3% while the reported sequence matched ground truth only 3.7% of the time. Extra frames buy score, not perception.