Measure Twice
Composite scores how is this calculated?
Text: 9.00 · Audio: 8.10 · Rules: 9/10
Text axes
form_as_finite_generator
9
long_fuse_plotting
10
status_as_live_wire
10
genre_register_collision
8
tonal_duality
10
structural_reveal
9
politeness_as_masking
10
Audio axes
character_distinctness
9
delivery_match
8
pacing
7
prosody_variation
8
consistency_across_episode
8
Audio
WPM: 99 · WER: 0.00 ·
seams: 4 ·
distinctness: marginal
(Δ 12 Hz)
Script statistics
95 lines · 629 words · 6.62 avg words/line
| character | lines | words | line % | word % | avg w/l |
|---|---|---|---|---|---|
| Arthur | 49 | 348 | 51.6% | 55.3% | 7.1 |
| Thomas | 45 | 276 | 47.4% | 43.9% | 6.13 |
| Announcer | 1 | 5 | 1.1% | 0.8% | 5.0 |
| third | lines | words | wpm | beats | scenes |
|---|---|---|---|---|---|
| beginning | 40 | 308 | 112 | 4 | 1 |
| middle | 20 | 121 | 80 | 2 | 2 |
| end | 35 | 200 | 77 | 4 | 1 |
Iteration trajectory (this run, 2 iters)
Each meta iteration generates a fresh concept; iter scores aren't directly comparable as "improvement." The shaded row is the iter currently displayed.
| iter | concept | autorater | audio judge | rules | seams | applied |
|---|---|---|---|---|---|---|
| 01 | Measure Twice | 8.95 | 7.00 | 8/10 | 2 | 0/0 |
| 02 | Measure Twice | 9.00 | 8.10 | 9/10 | 4 | 0/0 |
Critique summary
Composite 9.00/10. Top issues: • [tonal_shift_present] Land at least one moment in the final third where the laughter falls quiet. Cite a specific timestamp (M:SS or t=NNN form) in `analysis.standout_moments` so the rule can detect it. The 'vinegar in the sponge cake' beat.