Runtime

Reset week: absorbs the three sessions missed in W1 without touching the rest of the plan — same weekly load, races stay anchored. If the Jul 5 long run did happen after all, check it off in W1 and treat Wednesday as a plain easy 4 km. Strides move to Friday: short accelerations, smooth and controlled, not sprints.

6K long run — new territory beyond the 5K. Keep it slow. Walk if you need to. Lift labels corrected Aug 11 across W3–W5: the rotation is Leg → Push → Pull, which is what the Strava log has done without a break for thirteen straight sessions. These three weeks still carried the old Push/Pull/Leg order from before the Aug 1 rebuild, which put Leg on Saturday the day before the long run. W1 is left as Push/Pull/Leg on purpose — that is genuinely what was done that week, before the order changed.

Deload week. Every 3-4 weeks, pull back 20-25% to let your body adapt. Not optional.

Rewritten Jul 26. The original W5 asked for 22 km and a first 8K — off a month averaging 5.7 km/week and a 5.05 km longest run, that was a fantasy. This is three short runs and a 6 km long: still your biggest week of the block, but built from sessions you've already proven you can do. Frequency is the target here, not distance. Three runs at 8:30-9:00/km beats one hard one. If you only get two in, make them Wednesday and Sunday.

One-off shape this week only: the long run sits on Saturday at 05:45 and Sunday Aug 9 is the single empty day of the whole programme. From W8 the long run returns to Sunday where it belongs. 05:45 because sunrise is ~06:15 and by 09:00 Istanbul is 32°C. Friday is upper-body only so Saturday's legs are fresh. Outcome: 9.05 of 16 km. Mon 4.03 @ 9:02 and Wed 5.02 @ 8:57 both landed; the Saturday 7 km did not, and it is absorbed into W7 Sunday rather than added on top.

Rewritten Aug 11: Arsuz is off, the week is in Ankara — bodyweight and resistance bands, no pool. The three swims become band sessions (LEG Wed, PUSH Fri, PULL Sat); Monday's AWAY-1 already happened, so the 4 km moved to Tuesday and the 6 km to Thursday. Sunday is 7 km, not 8. That is the long run missed on Sat Aug 8 being absorbed here rather than bolted on as a second one — a 7 km and an 8 km inside five days, off a six-day layoff and a 6.02 km lifetime best, is how this block ends early. W08 onward is untouched. ⚠ No basketball this week. Nothing lateral or high-impact the day before a distance PB. The tempo is gone from the whole programme: your all-out 5K is 8:01/km and your easy pace is 8:30/km — only ~30s/km between easy and maximum, no aerobic gap for tempo to sit in. Strides are the only intensity until W12. Tuesday's run changes the read on this week. 4.03 km @ 8:22 with 65.8 m of climb — 16 m/km, the first non-flat running in the block — and it is 40s/km faster than Monday-of-W6 on ground that is far harder. But Ankara sits at ~1,030 m and the route climbed the same 33 m twice: climb 1 went at 8:50/km, climb 2 at ~9:37/km even after discounting the 20s stop — ~50s/km slower, at a higher heart rate (157 vs 153, peaking 168 vs 163). Durability runs out around 2.5–3 km. That is the honest limit right now. Thursday's 6 km and Sunday's 7 km are to be run by feel, not by pace — at this altitude on this terrain, holding 8:22 is not the goal and chasing it is how Sunday's PB attempt fails. Thursday, Aug 13 — the distance was obeyed and the pace instruction was not. 6.02 km @ 8:07/km, the same loop as Tuesday run three times instead of twice: 102 m of climb, avg HR 160, max 181, relative effort 97 against Tuesday's 51. The best 5 km inside it was 40:19 — 8:04/km, fourteen seconds off an all-out 5K, uphill, at 1,030 m, on a day marked easy. Only 26% of it sat in Z2 where an easy run belongs; 65% was Z3 and the last 4:25 was Z4. There was no Z1 at all. The finding that matters is the third lap. Climbs went 8:45, 9:32, 9:38 — Tuesday decayed 47s/km from climb 1 to climb 2 and today repeated that decay exactly, then lost only 6s/km more on the third. The wall at 2.5–3 km is real but it is a step, not a slide: past it the pace holds. That is the case for Sunday's 7 km, and it was bought with a hard session three days out rather than the easy one that was written. Friday and Saturday are now upper-body only and non-negotiable, and Sunday goes out at 9:00/km or slower. The legs have had burpee broad jumps Wednesday and a Z3 hour Thursday; the PB does not need a fourth hard day to defend it. Outcome — 17.07 of 17 km. The first 100% week of the block. Three runs written, three runs run, each to the metre: 4.03 / 6.02 / 7.02. Only Friday's AWAY-PUSH went missing, and Saturday came back as a real gym 2-PULL instead of bands. Sunday, Aug 16 — 7.02 km @ 7:43/km, avg HR 150, max 164, flat at Caddebostan. Longest run of his life and the fastest running he has done: the best 5K inside it was 38:13 = 7:39/km, two minutes under the Jun 28 baseline of 40:13 — and it was set at km 1–5 of a 7 km run, not as a time trial. Read it against Thursday with care: Ankara was 1,030 m and 17 m/km of climb, today was sea level and dead flat (3 m the whole way), so the 24s/km is terrain as much as fitness. What is not terrain is the heart rate — 150 avg here against 160 in Ankara, and relative effort 72 against 97. Faster, longer, and calmer. Zones: ~63% Z2, ~34% Z3, no Z4 at all, against Thursday's 26% Z2 / 65% Z3 with a Z4 finish. This is what the easy brief was asking for, and it is the first long run of the block that actually obeyed it. Which puts a real question on the programme: 7:43/km at the top of Z2 is faster than the 8:01/km that used to be his all-out 5K. The whole no-tempo, no-intervals ruling rests on there being only ~30s/km between easy and maximum. That gap has moved and the ruling needs re-deriving before Phase 2 hardens. The honest flaw is the shape of it. Splits ran 7:20 / 7:39 / 7:23 / 8:03 / 7:49 / 8:16 / 8:01: the first 3 km averaged 7:27, the last 4 averaged 8:02, and there are two clear stops in there (km 4 and km 5.7, where pace drops under 1.5 m/s and HR falls to 123–134). He went out at a pace he could not hold and paid for it from km 4. That is the thing to fix before Sep 27 — a race has no traffic lights and no standing rests. Run the next long one at an even 8:00/km start to finish and it will feel easier and finish faster than this did.

Deliberate step back after three build weeks — this is where the adaptation actually lands. Was 26 km with 5x400m intervals and an 11 km long run; off an 18 km week that was a 62% jump into speed work you have no base for. Amended Aug 11: leg day moved Tue → Thu, push moved Thu → Tue. The original shape put a heavy leg session ~48h after Sunday's distance PB — exactly when DOMS from a novel distance peaks — and it would have been the first barbell legs in two weeks. The swap gives the legs four days after the PB before anything heavy, and still leaves three clear days before the next long run. Monday is no longer a travel day (home from Ankara Aug 16) so the run stays, cut to 3 km as a shakeout. Re-cut again Aug 16, Aral's call: he did push on Sunday alongside the 7 km, so the lifts this week run PULL Tue → LEG Wed → PUSH Sat. Leg lands Wednesday, which is the right day — three days past the PB rather than two, and three clear days before Sunday's long run instead of two. (⚠ This sentence read “four…instead of three” until Aug 18; the counts were each one too high. Wed 19→Sun 23 leaves Thu/Fri/Sat. The ranking was still right, so the decision stands.) That pushes the 4 km off Wednesday onto Tuesday, where it stacks with pull: upper-only, so it costs the run nothing, and it puts the run on a gym day, which is where his runs actually survive. Thursday and Friday are now both rest — a real two-day break before the Sat/Sun block. ⚠ Wednesday is the fourth leg-loading day in a row (Sun 7 km → Mon 3 km → Tue 4 km → squats) and the first barbell legs in 15 days, so the squat starts at 60 kg, not 70. Volume holds at 13 km. He has just PB'd twice in a morning off his first 100% week; the down week is where that adaptation lands and it is not the place to spend it. Both midweek runs and the long run carry a pacing brief now, not a pace target — go out at 8:30 and finish faster, because Aug 16 went out at 7:20 and closed at 8:16. 🔄 Re-cut again Aug 18, off the calendar. Todoist had drifted from this page by hand: leg had been dragged Wed → Fri and the Tuesday run Tue → Thu. Friday legs was the problem — first barbell legs in 17 days, one clear day before Sunday, so peak DOMS lands on the long run. That is worse than the Thursday slot already rejected on Aug 16. Leg restored to Wednesday (Aral's call). Tuesday's 4 km never happened and is not pretended otherwise — Tuesday is now the pull-only day it actually was, and the run stays on Thursday where he had moved it, one day after squats as active recovery. Shape is now Mon 3 km ✓ · Tue PULL ✓ · Wed LEG · Thu 4 km · Fri rest · Sat PUSH · Sun 6 km long. Volume still 13 km — nothing was added, one run moved two days. ⚡ The lesson worth keeping: the plan lives in three places (this page, Todoist, Google Cal) and only Todoist can be dragged with a thumb. When they disagree, Todoist is what he actually sees — so check it, don't assume this page is what he is following. Outcome — 13.05 of 13 km. The second 100% run week, back to back with the first. 3.02 / 4.01 / 6.02, each to the metre again; the lifts went 2 of 3 (leg day missed, third week without barbell legs). Sunday, Aug 23 — 6.02 km @ 8:14/km, avg HR 149, max 173, 49:32. The running partner bailed and he ran it alone at 06:38. The pacing brief was finally obeyed: splits 8:31 / 8:43 / 8:34 / 8:05 / 7:54 / 7:35 — first 3 km at 8:36, last 3 at 7:51. Every previous long run in the block went out fast and bled; this one went out at the easy pace and closed 1 min/km faster. That is the shape a race needs, and it is the thing Aug 16 was told to fix. Two things to read honestly. First, the speed was bought with heart rate, not with ease: HR climbed from ~130 to 173, and the last 600 m sat at 171–173 — km 6 was a Z4 finish, not an easy close. A race-day negative split closes hard on purpose; an easy long run should not. Second, there are two ~40 s dips to ~132 bpm at 3.8 km and 4.7 km that read as walk breaks. The 5K inside it was 40:48 (8:10/km), slower than Aug 16's 38:13 — correct, because this was the down-week long run and not a time trial. W9 runs as written: the gate was “more than two missed runs across W6–W8” and the count is one (Sat Aug 8).

PPL-3 starts here (Aug 24 → Oct 18). PPL-2 ran 29 sessions in ten weeks and the honest read of the log is that it stopped progressing, not that it got boring: the Smith squat (50→80), incline DB (20→25) and overhead extension moved; the other fourteen movements sat at the same weight for ten weeks, and the barbell row and Arnold press ended below where they started. Two causes, neither of them motivation — no written add-weight rule, and seven-exercise sessions that left the main lift under-loaded. PPL-3 fixes both: five movements and a two-movement core finisher per day, one rep scheme per tier (main 5×5, secondary 3×8, accessory 3×12), and the rule written on every task — +2.5 kg the session all 25 reps are done. Every stalled lift is swapped for its sibling: barbell bench for incline DB, standing OHP for Arnold, pull-ups and single-arm rows for the barbell row, paused Smith squat (no free rack in the gym) for the plain one, step-ups and hip thrusts for the split squat, single-leg calves for the machine. Core leans running-specific — Pallof, carries, Copenhagen plank, dead bug. The week shape is W8's: PULL Tue, LEG Wed evening after the morning run, PUSH Sat. Wednesday legs is three days past Sunday and three clear days before the next one; Thursday becomes the rest day. Down week W11 drops the fifth set, W12 Wednesday is the last leg session before the race, race week is one 70% pull and nothing else. The 4th weekly run arrives — short, Friday. Only run this week as written if W6, W7 and W8 actually happened. If you missed more than two runs across those three weeks, repeat W8 instead and tell Vega. Falling behind is fixable; stacking volume on a broken base is not.

Sunday Sep 6 is the first 10 km of your life. Three weeks before the race, on a training Sunday, with no bib and no clock — exactly where you want to meet that distance for the first time. Walk 60s every 3 km if you need to; that is strategy, not failure. Carry water. Do not race it.

Second down week. You just ran 10 km for the first time — absorb it. Long run back to 7 km. This is the week your body banks the last three.

Biggest week of the block, and the first and only quality session before Buyukada: Wednesday's middle 3 km at goal 10K effort (~8:00/km). Not a time trial — a rehearsal so race pace is not a stranger on Sep 27. Sunday's 11 km is the longest run before the race. Tuesday is the last heavy leg session until after Buyukada.

Post-race. Monday is non-negotiable rest, Tuesday's jog will feel awful and that is normal. Long run drops to 10 km — you raced 10 km eight days earlier, this one is at 9:15/km with no clock. Do not try to prove anything this week.

Now the half-marathon build actually starts. Sunday's 13 km is new territory. This is where you start practising fuelling — something with sugar at 45 minutes, water every 3 km. Whatever you plan to use on Nov 1, use it here. Leg day goes moderate from now on.

Sunday Oct 18 is the dress rehearsal — 16 km, the longest run of the programme. Same wake time, same breakfast, same shoes, same fuelling as race day. Was 18 km; 16 is the honest ceiling off this base. That leaves a 5 km gap to the half, which race-day adrenaline and a run/walk strategy will cover — but you should know the gap is there. If you finish this, you finish Nov 1.

Volume drops a third. You will feel antsy and convinced you are losing fitness. You are not — the hay is in the barn. No heavy leg work from here to the race. The only job left is arriving on Nov 1 healthy.