training
Training
A weekly running log.
The week
week ending 30 aug 2026Sydney Marathon in 2:50:50 — a large PB. Fuelling, weather, heart rate and injury are all eliminated as causes. What was left was durability at continuous marathon pace, which had never been rehearsed beyond about 5 km unbroken in the whole cycle.
- runs
- 5
- distance
- 76.8 km
- time
- 5h 44m
- elevation
- 858 m
| day | session | km | time | pace | hr | load |
|---|---|---|---|---|---|---|
| mon 24 aug | Lunch Run | 5.1 | 24:46 | 4:50 | 148 | 30 e |
First run back after four days off following the 19 Aug ankle roll — days taken by choice rather than necessity. 4:50/km, heart rate 148. | ||||||
| wed 26 aug | Afternoon Run | 11.1 | 52:52 | 4:47 | 160 | 74 e |
4x400m in 82 / 83 / 85 / 85 s (3:25 / 3:28 / 3:32 / 3:32 per km) with 400m floats, then 5.9 km of trail at HR 163. Steep, stair-heavy, uneven Kirribilli and Neutral Bay terrain. Only slight ankle soreness; completed in full. | ||||||
| fri 28 aug | Half Mona fartlek | 13.9 | 1:15:25 | 5:25 | 148 | 90 e |
10.3 km easy at HR 144, then 1x90s, 2x60s, 2x30s, 2x15s at 3:15-3:38/km with equal floats, rep HR 160-175. | ||||||
| sat 29 aug | Shakeout with strides | 4.0 | 19:44 | 4:53 | 145 | 23 e |
4x1' at 3:28-3:38/km off 1' floats, rep HR 157-169. Heavily interrupted by foot traffic; 1184 s moving inside 3694 s elapsed. | ||||||
| sun 30 aug | Sydney Marathon 2026 | 42.7 | 2:50:50 | 4:00 | 173 | 278 e |
2:50:50. Large PB. Conditions 9-11 C, 94 percent humidity. Fuelled with a 30-40 g carbohydrate gel roughly every 5 km throughout, the same intake practised on every long run. Ran with the 2:45 group to about 19 km, then let them go on recognising the pace would not hold. At 26 km the legs felt like the end of a 5 km race, heavy and full of lactate — pushed through without losing much pace. Quadriceps began to hurt entering Centennial Park around 32-33 km and worsened to the last kilometre, which is downhill and was run as hard as possible to the line. No ankle or knee pain at any point. GPS 42.68 km against official 42.195. Halves 1:21:33 / 1:29:17. GPS 5 km splits 3:52 / 3:51 / 3:50 / 3:54 / 4:01 / 4:06 / 4:10 / 4:14, closing 2.7 km at 4:05 and the last mile at 3:57. First-half HR 172.8, second-half HR 172.7. No stoppage: moving time equals elapsed time. | ||||||
good
- The aerobic engine was not the limiter and never wobbled: heart rate averaged 172.8 for the first half and 172.7 for the second, sitting in the centre of the calibrated 172-177 band for 2h51m with essentially zero drift.
- Fuelling was executed properly and can be ruled out. A 30-40 g carbohydrate gel roughly every 5 km across the whole race is on the order of 85-110 g per hour, at or above the top of what is normally recommended, and it was the same intake practised on every long run of the build rather than something new attempted on the day. No stomach trouble.
- The pacing decision at 19 km was good judgement, not a blow-up. Contact with the 2:45 group was dropped deliberately on the recognition that the pace would not hold, rather than being lost involuntarily 10 km later at far greater cost.
- Finished strongly despite sore quadriceps from 32 km: the last mile ran 3:57/km at HR 176.3, quicker than the 35-40 km split, so there was headroom at the line that had not been available at 35 km.
- The ankle rolled on 19 Aug and the knee scraped in the same fall were both completely silent across 42 km.
note
- Conditions were close to ideal and are not an excuse in either direction. 9-11 C with 94 percent humidity sounds punishing, but at that temperature the air holds very little water: the vapour-pressure gradient from skin to air stays wide and evaporative cooling is barely impaired. This was a fast morning.
- The gap between the half-marathon reference and the result is itself the diagnosis. A 1:18 half predicts about 2:42-2:44 on the standard exponent of 1.06. The actual ratio between 1:18 and 2:50:50 implies an exponent nearer 1.13, well above the 1.06-1.08 typical of a trained marathoner. That is not a speed deficit — speed was there. It is a drop-off with duration.
- No continuous run at marathon pace beyond about 5 km happened anywhere in the cycle. Every exposure was broken: 4x5k with roughly 1 km of jogging between pieces on both 2 Aug and 16 Aug. Twenty kilometres of goal pace in five-kilometre parcels is a materially easier thing than twenty kilometres of it unbroken, and the difference showed up between 20 km and 40 km.
- The tissue that failed was the quadriceps, which appears nowhere else in this log. Soreness began entering Centennial Park at 32-33 km — the flattest section of the course, a 3.4 km loop with 36 m of gain — so the damage was inflicted earlier and only expressed there. The course front-loads its work: roughly 95 m of net descent inside the first 4 km, and about 69 percent of the total climbing in the first half.
- GPS measured 42.68 km against the official 42.195 km, a 485 m over-read, so per-kilometre splits quoted from Strava run about 1.1 percent optimistic. The finish time is unaffected.
watch
- The whole result reduces to one thing worth carrying forward: continuous time at marathon pace, which this build never accumulated. Aerobic capacity, threshold stability, economy, fuelling and race-day conditions were all in good order — the drop-off was musculoskeletal and it began at 20 km.
- The quadriceps are the specific tissue to keep in view, given a descending opening was paid for 30 km later. First appearance in this log.
- Day-after status: quadriceps sore but stairs manageable and walking normal, consistent with ordinary marathon damage. The ankle is still slightly sore to walk on, which is worth tracking as the one item not yet fully closed.
5 weeks
27 jul 2026 to 30 aug 2026 · 35 days- runs
- 26
- distance
- 415.4 km
- time
- 31h 26m
- per week
- 83.1 km
| week ending | km | Δ km | runs | time |
|---|---|---|---|---|
| 02 aug | 105.5 | · | 6 | 8h 05m |
| 09 aug | 111.6 | +6.0 | 6 | 8h 35m |
| 16 aug | 90.3 | -21.3 | 6 | 6h 44m |
| 23 aug | 31.1 | -59.1 | 3 | 2h 19m |
| 30 aug | 76.8 | +45.7 | 5 | 5h 44m |
12 months
01 sep 2025 to 30 aug 2026- runs
- 280
- distance
- 4360 km
- time
- 342h 42m
- per month
- 363 km
| month | km | runs | time |
|---|---|---|---|
| sep 2025 | 359.3 | 24 | 28h 04m |
| oct 2025 | 325.9 | 23 | 25h 12m |
| nov 2025 | 262.3 | 23 | 19h 49m |
| dec 2025 | 324.4 | 18 | 26h 02m |
| jan 2026 | 300.8 | 19 | 23h 46m |
| feb 2026 | 351.5 | 22 | 28h 14m |
| mar 2026 | 473.1 | 28 | 37h 34m |
| apr 2026 | 465.0 | 25 | 37h 30m |
| may 2026 | 350.5 | 23 | 29h 01m |
| jun 2026 | 346.4 | 24 | 26h 32m |
| jul 2026 | 456.9 | 30 | 35h 13m |
| aug 2026 to 30 aug | 343.7 | 21 | 25h 45m |
- fitness
- 73
- fatigue
- 74
- form
- +29
- as of
- 30 aug 2026
- fitness, 42-day average
- fatigue, 7-day average
- the gap, fitness minus fatigue
| month end | fitness | fatigue | form |
|---|---|---|---|
| sep 2025 | 67 | 61 | +5 |
| oct 2025 | 60 | 54 | +4 |
| nov 2025 | 53 | 46 | +1 |
| dec 2025 | 61 | 83 | -21 |
| jan 2026 | 56 | 53 | -4 |
| feb 2026 | 62 | 55 | -1 |
| mar 2026 | 76 | 82 | -8 |
| apr 2026 | 83 | 87 | -7 |
| may 2026 | 79 | 78 | +12 |
| jun 2026 | 72 | 52 | +25 |
| jul 2026 | 81 | 92 | -15 |
| aug 2026 | 73 | 74 | +29 |
The curve runs on a heart-rate estimate recomputed the same way for every session, not on the stored load column. That column carries the numbers each source reported, and the two sources are on different scales, so feeding the mixture to the model would put a step at the seam between them and draw a fitness jump that never happened.