24 weeks to Kyoto
Sub-5:00 needs 7:07 per kilometre. You held 7:26 for the Osaka half and 7:40 to 30 km. So this block is not only about running further — it moves your whole pace curve, then makes the new pace survive 42 kilometres.
Weeks 1–6 · 7 Sep – 18 Oct · 32–44 km
Get back to five days a week without a single crash week. No hero sessions. The only target is showing up.
Weeks 7–12 · 19 Oct – 29 Nov · 41–57 km
Marathon pace is currently faster than your easy pace. This phase moves the whole pace curve, and puts hills under your legs.
Weeks 13–19 · 30 Nov – 17 Jan · 44–67.5 km
Long runs grow teeth. Marathon-pace blocks, fuelling rehearsed every week, and the sessions that decide the race.
Weeks 20–24 · 18 Jan – 21 Feb · 21–68 km
One more peak, then take the foot off. Taper is training too. Arrive fresh, not fit-and-flat.
Bangkok evenings run 30°C and humid. Kyoto in February is 3–10°C. Train to the Bangkok pace at home; race to the Kyoto one.
Heart rate under 140. If you can’t talk, slow down.
Steady and boring. The adaptation is the duration, not the pace.
A notch quicker than the long run and a notch slower than race pace. Short sentences only.
Goal pace for sub-5:00. Practise it in heat, race it in cool.
Comfortably hard. You could hold it for an hour if you had to.
Hard but controlled. Full recovery between reps.
Half marathons, run as long runs rather than as efforts. A race is a good place to step the long run up: the course is measured, the drinks are handed to you, and the crowd carries you further than a solo Sunday would. None is a time trial, and the week around each is shaped for a race day rather than a Sunday long run.
Four timed efforts that decide which goal you actually race. Sub-5:00 is ambitious, so rather than finding out at 30 km on race day, you find out here, with time to adjust.
Sub-5:00 still live. Slower than 32:30 and we recalibrate to 5:15.
The halfway read. Over 68:00 and the A-goal moves to 5:15.
Durability check. This is the exact thing that failed at Osaka.
The decision. ≤2:15 race sub-5:00. 2:15–2:25 race 5:15. Over 2:25 race 5:30 and bank the finish.
Not the target, yet. Every personal best you hold was set inside a marathon — the 5 km and 10 km inside the one you did not finish — which makes all of them soft. A fresh 5 km would beat 35:00 comfortably. Until a real time trial says otherwise, nobody knows what you are currently worth, and that is what the gates are for.
| Gate | Sub-5:00 | 4:45 |
|---|---|---|
| 5 km, week 7 | 31:00 | 29:00 |
| 10 km, week 12 | 65:00 | 61:00 |
| 30 km at race pace, week 15 | 7:30–7:40/km | 7:10–7:20/km |
| Half marathon, week 19 | 2:15 | 2:08 |
| Week | Starts | Phase | Target | Long run |
|---|---|---|---|---|
| A | Wk 1 · 7 Sep | Rebuild | 3229.2 | 1010 |
| A+ | Wk 2 · 14 Sep | Rebuild | 3434 | 1212 |
| F | Wk 3 · 21 Sep | Rebuild | 3715.7 | 148.1 |
| A+ | Wk 4 · 28 Sep | Rebuild | 3736.5 | 1414.1 |
| A+ | Wk 5 · 5 Oct | Rebuild | 40.544.8 | 1616 |
| D | Wk 6 · 12 Oct | Rebuild | 4416.1 | 180 |
| A+★ | Wk 7 · 19 Oct | Base + Speed | 4445.8 | 1818 |
| B | Wk 8 · 26 Oct | Base + Speed↓ | 4132.7 | 2015 |
| B | Wk 9 · 2 Nov | Base + Speed | 4235.4 | 2020 |
| 10 | Wk 10 · 9 Nov | Base + Speed | 4522.5 | 220 |
| 11 | Wk 11 · 16 Nov | Base + Speed | 57— | 24— |
| 12 | Wk 12 · 23 Nov | Base + Speed↓ | 43.5— | 21.1— |
| 13★ | Wk 13 · 30 Nov | Marathon Specific | 49.5— | 14— |
| 14 | Wk 14 · 7 Dec | Marathon Specific | 59— | 27— |
| 15★ | Wk 15 · 14 Dec | Marathon Specific | 63.5— | 30— |
| 16 | Wk 16 · 21 Dec | Marathon Specific↓ | 44— | 20— |
| 17 | Wk 17 · 28 Dec | Marathon Specific | 62— | 29— |
| 18 | Wk 18 · 4 Jan | Marathon Specific | 67.5— | 32— |
| 19★ | Wk 19 · 11 Jan | Marathon Specific↓ | 44.5— | 21.1— |
| 20 | Wk 20 · 18 Jan | Peak & Taper | 64.5— | 30— |
| 21 | Wk 21 · 25 Jan | Peak & Taper | 68— | 32— |
| 22 | Wk 22 · 1 Feb | Peak & Taper↓ | 56— | 24— |
| 23 | Wk 23 · 8 Feb | Peak & Taper↓ | 47.5— | 16— |
| 24 | Wk 24 · 15 Feb | Peak & Taper↓ | 21— | 42.195— |
Stops are coloured as on the road: graded once done, blue for this week · ★ decision gate · ↓ down week · the small number under each target is what you actually ran · all distances in kilometres
Three jobs, three shoes. Every kilometre between now and Kyoto belongs to exactly one of them, so no shoe is asked to do a job it is wrong for — and none of them wears out before race day.
You will run 133.1 km in this pair.
56 km on it today · 189.1 km on it after Kyoto · 510.9 km of life still spare
You already own it, and it is a light, quick shoe — right for threshold, hills and the four time trials. Keep it to those and it lasts the whole block with life to spare. Put it under the 32 km long runs instead and it is finished by January, having cushioned none of the runs that needed cushioning.
You will run 337.5 km in this pair.
165.3 km on it today · 502.8 km on it after Kyoto · 197.2 km of life still spare
This is the hole in the rotation: you own nothing cushioned, and 78% of the block is easy and long running in Bangkok heat at 78 kg. Buy it this week, not in December — the long runs cross 20 km in week 7, and the shoe wants 150 km in it before that. It carries every easy day of the block, every long run up to the 24 km on 15 November, and the October half marathon, which you run at long-run pace. One pair covers all of it and finishes with life to spare; it hands the long runs over in week 11 rather than being replaced.
Also considered
A tougher outsole and a mesh upper that breathes better than the 1080's knit, which is worth something in Bangkok.
Heavier, and a different shape of shoe to anything you have run in — try it on rather than ordering it.
Lighter and livelier, so you could push the pace in it.
The least steady underfoot late in a long run, which is exactly when your form goes.
You will run 341.3 km in this pair.
341.3 km on it after Kyoto · 358.7 km of life still spare
Your choice, and a good one: 46 mm of stack, no plate, 240 g, and a wide base reviewers rate as stable over marathon distance — built for staying comfortable for five hours rather than going fast for three, which is your actual problem. Week 11 is the last handover that keeps the 1080 inside its life, so it is also the one that leaves this pair freshest on race day. It takes every long run from the 26 km on 22 November onward — both 32 km runs included — plus the November half marathon as a dress rehearsal. By Kyoto you will have run fifteen long sessions in it and know exactly how it treats your legs at four hours. Never race in a shoe you have not run long in.
Also considered
Nothing, for you, yet.
Aggressive racing geometry is tuned for ground-contact times quicker than 7:56/km, and five and a half hours on a racing midsole punishes the tiring legs that ended both of your marathons. Revisit the day a gate says sub-4:30.
What 700 km means. 700 km is the planning life of a cushioned trainer. Makers quote 650–800 km; at 78 kg on hard Bangkok road, the bottom of that range is the honest number. It is when the foam stops protecting you, not when the shoe looks worn out.
Kilometres are never typed into this file. Each slot says which session types it covers and over which weeks, and lib/gear.ts adds them up out of data/plan.json — so if a week's volume changes, the shoe plan changes with it.
Your history has seven crashes in it: a 25 km week followed straight away by one under 12 km. Crashes come from weeks that were too big, not from weeks that were too small.
We do not push on through a broken week. The next week runs again at the same volume instead of progressing. Losing seven days is cheap; losing a month to an injury is not.
At Osaka your heart rate averaged 144 with a ceiling of 158. The engine was fine. The legs gave out. One loaded machine session, on its own day, is the actual training here, not an optional extra.
Heart rate under 140. Most runners ruin their hard days by running the easy ones too quickly, and then wonder why the long run hurts.
A gel every 40 minutes whether you want it or not. Race day is not the time to find out what your stomach will accept.
Seven kilometres is your median run, so nothing in this block is a token effort. The consequence is that a lighter week has to mean fewer running days rather than shorter ones — weeks 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 12, 16 and 19 run four days instead of five for exactly that reason.