Values transcribed from the essay text in /tmp/p3-snap.html (Essay 09 build,
October 2026). Nothing was recomputed and no data points were invented.
Where the essay shows a derivation (e.g. 5 mmol/kg/km × 42.2 km ≈ 210 mmol/kg),
the CSV records the same arithmetic the essay shows.
clock_stops.csv — the 7 race-clock stops plus the +24 h coda: clock time,
distance, beat title, which runner(s), and the key claim at each stop.fueling_bands.csv — the carbohydrate-intake guideline bands behind the
“What if you fueled differently?” slider (30 / 60 / 90 g/h).fueling_mechanism.csv — the Tsintzas et al. 1996 mechanism shown beside the
slider: 25% type-I fiber glycogen sparing; exhaustion 104 → 132 min.glycogen_arithmetic.csv — the Sherman glycogen math behind the wall claim.sweat_drift.csv — sweat rates, body-mass loss, the 42.5% sweat-loss
underestimate, and cardiovascular-drift figures.wall_rates.csv — self-reported vs pacing-defined wall rates (Buman 2008,
Smyth 2021, Berlin 1999–2025, Boston 2015–17).hyponatremia.csv — Almond et al. NEJM 2005 Boston figures.ck_enzymes.csv — post-marathon creatine kinase figures (JAMA 1979/1981).core_temperature.csv — core-temperature plateau and the hot half-marathon
outlier group.Every value traces to a source named in the essay’s numbered footnotes (notes 1–19) and the full source list in the essay’s “How this was built” section. Key primary sources: Sherman et al. (Int J Sports Med, glycogen utilization); Bergström et al. 1967 (glycogen stores); Wingo, Ganio & Cureton 2012 (cardiovascular drift); Stone et al. 2025 (drift magnitude); Lara et al. (sweat rate); Almond et al. NEJM 2005 (hyponatremia); Tsintzas et al. 1996 (carbohydrate mechanism); Jeukendrup 2012 / ACSM (fueling guidelines); Smyth 2021 and Buman et al. 2008 (wall rates); PLoS One 2025 (Boston pacing).