3 tasks. May 21, 2026 closed at 36.3x weighted leverage across 69.0 human-equivalent hours in 114 minutes of wall-clock time. Supervisory leverage came in at 318.5x.
That is 1.7 weeks of human-equivalent throughput in 1.9 hours. The ceiling was 55.4x; the floor was 8.6x. 3 of the 3 entries came from a single project.
Task Log
| # | Task | Human Est. | Claude | Sup. | Factor |
|---|---|---|---|---|---|
| 1 | Deployment | 60.0h | 65m | 6m | 55.4x |
| 2 | AP Precalc spec audit + math content rollout plan (spec issue identification, activity catalog inventory, 8-phase plan covering spec fixes, math-specific content shapes, 5 new Tier A activities, v2 atom… | 4.0h | 14m | 3m | 17.1x |
| 3 | avian-api native-mode wiring (bcrypt pin, settings hardening, certs->certifications fix, native entitlement path, commit-on-exit deps, event_type kwarg drift, secure-cookie toggle) + seed_test_user.py +… | 5.0h | 35m | 4m | 8.6x |
Aggregate Statistics
| Metric | Value |
|---|---|
| Total tasks | 3 |
| Total human-equivalent hours | 69.0 |
| Total Claude minutes | 114 |
| Total supervisory minutes | 13 |
| Total tokens | 360,500 |
| Weighted average leverage factor | 36.3x |
| Weighted average supervisory leverage factor | 318.5x |
| Human-equivalent weeks | 1.7 |
Analysis
The highest factor of the day came in at 55.4x and the lowest at 8.6x, a spread of 6.5 times between the two. That is a wide range for a single day, and it usually means the day mixed mechanical work with work that needed real judgement.
The largest single entry accounted for 60.0 of the 69.0 human-equivalent hours, or 87 percent of the day. One task carrying that much of the total is worth noting; the average is doing less work than it appears to.
Supervisory time was 13 minutes against 114 minutes of execution, a ratio of about 1 to 9. Supervisory leverage of 318.5x is the figure I find most honest, because it measures the hours I actually spent rather than the hours a machine spent on my behalf.
Every figure here is recorded at the time the work is done rather than reconstructed afterwards. The human estimate is my own judgement and carries the uncertainty that implies; the minutes and tokens are measured. The full dataset, including this day, is available for download.