Marathon Pace KM
Menu

My next build

Rebuilding after Sydney: my plan to finish the next marathon stronger

At a glance

The next-build intentions documented on 2 October 2026: repeatable weeks, controlled rehearsals and a record of recovery. This remains a proposed approach; completed training outcomes have not been added.

At Sydney Marathon on 30 August 2026, the race started feeling difficult just before 30 km. I first walked at roughly 31–32 km. I had wanted sub-four; instead, the later kilometres became repeated running and walking. The feeling I remember most was being empty, rather than sore.

My Sydney race review covers what happened. This article is about what I want to change before the next attempt. I still care about the time, but I also want a marathon that finishes differently.

Where this plan stands: these are my intentions as of 2 October 2026, not the results of a completed rebuild. This article does not set a next-race date or a return-to-running timetable. Injury concerns need individual assessment; a calendar, calculator or watch score cannot clear me to train.

A strong finish means more than a fast last kilometre

My two Run Melbourne half marathons give me a useful contrast. In 2025 I struggled late. In 2026 I finished strongly, with the final two full recorded kilometres at 4:53 and 4:51. The full split comparison shows that difference without assuming that every variable was the same between races.

That is the feeling I want to carry into a marathon: still running with control late in the race, rather than trying to rescue the final kilometre after a long slowdown. I could accelerate near the Sydney finish too. That short burst did not mean the preceding kilometres had gone to plan.

I do not need to force a dramatic negative split. I want to choose an opening effort I can support, keep fueling manageable and respond before the plan becomes unrealistic. A faster finish would be welcome; sustaining the run is the more important change.

Build from my normal training, not my biggest week

My final 12 complete training weeks before Sydney, from 1 June to 23 August, averaged 60.7 km and 4.4 runs per week. The peak was 83.9 km. The bigger dips in my running history usually came from mild injuries, so the spread was not just a deliberate training pattern.

I want to use the whole block as the starting point for planning. An 80 km week tells me I completed that week. It does not tell me that I can repeat it comfortably alongside work or that another increase is appropriate.

The practical change is to count the interrupted weeks too. If I enter recent training into the monthly planner, I will use all four weeks rather than choosing only the strongest ones. The planner is an organising aid, not an instruction to catch up missed kilometres.

Before increasing the ambition of the next block, I want a routine I can maintain. That includes recording why a run was shortened or missed. More mileage on paper is not an improvement if it produces another interruption.

The intended week has to fit my workdays

The longer-term structure in my post-Sydney review uses Tuesday, Wednesday, Thursday, Saturday and Sunday as running days, with Monday and Friday off. Tuesday and Saturday are workdays, so Tuesday needs to stay shorter and Sunday is the natural place for the longer run. Thursday gives me more flexibility.

That is an intended structure once it is appropriate to run regularly, not my current training log or a rehabilitation schedule. I do not want to turn every available day into a demanding session simply because five days appear on the calendar.

I also want to count the whole session. Warm-up, recoveries and cool-down take time and add distance. A workout described as 6 km of faster running can occupy much more of the morning than its headline suggests. On a busy day, a shorter easy run may fit the plan better than squeezing in a session I cannot complete calmly.

In the proposed build, two demanding exposures would be a ceiling, not a weekly obligation. If the long run contains substantial marathon-effort work, it would take the place of another demanding workout rather than being added on top. Some weeks would need less.

Shorter-race fitness first, marathon specificity later

The review proposes a progression through routine, shorter-distance fitness, a half marathon and then a marathon-specific block. I like that order because it gives each stage a purpose instead of treating every month as unfinished marathon training.

On a small screen, swipe the table if needed.

My proposed sequence—not a dated training programme
StageWhat I want to work onWhat I want to review before moving on
Return and routineAn appropriate return, then easy running that fits my week.Individual advice where needed, interruptions and how I respond to the routine.
Shorter-race preparationControlled quality work and, when appropriate, a 5K or 10K benchmark.Recent fitness and whether the week remains manageable.
Half-marathon preparationLonger sustained efforts and another chance to practise race execution.The whole race and recovery, not just its finishing time.
Marathon-specific preparationLong runs, realistic marathon effort and the complete fuel-and-drink plan.Late-run control, intake, recovery and whether the target still fits.

There is no fixed number of weeks attached here. The review's provisional race windows are planning assumptions, not confirmed bookings. If a stage is interrupted, I want to reassess it rather than compress everything left into the remaining calendar.

Judge the long run by its finish and recovery

My 9 August 32 km run is a useful reminder. The early two-kilometre laps were around 5:16–5:31/km; the final two full laps eased to about 6:28 and 6:47/km. Reaching 32 km was experience, but it was not the same as executing a controlled rehearsal.

Workout labels need the same scrutiny. Before Sydney, a session labelled marathon pace included 3 km blocks at about 4:43, 4:50 and 4:57/km. Those were quicker than my recent half-marathon pace. I want future marathon-specific work to match its intended purpose rather than become another test of speed.

For a rehearsal, I want to review three things together: whether I kept the intended effort through the later part, whether I could keep consuming the planned fuel and drinks, and how I recovered afterward. I do not need every long run to finish fast. An easy long run should still be an easy long run.

I have written separately about my homemade and commercial gel experiments. The next build needs to test a usable system, not just a pace or a carbohydrate target in isolation.

Keep a record that explains the numbers

I want a short note alongside each important session, rather than relying on a pace graph afterward. This is the checklist I intend to use; it is not a record of workouts I have already completed.

  • Purpose: easy running, controlled faster work or a marathon rehearsal.
  • Whole-session load: total time and distance, including recoveries.
  • Execution: intended versus actual effort, and any stops or changes.
  • Fuel and drinks: exact products, amounts and consumption times.
  • Response: late-run feel, next-day recovery and symptoms that need attention.

Shona Halson's review of training-load monitoring describes the value of considering both the work performed and the athlete's response, while noting that no single marker is definitive. That supports keeping context beside the watch data; it does not validate a particular mileage target for me.

Sub-four remains an ambition, but I want to choose the next target from the evidence available nearer the race. A predictor can give scenarios. It cannot establish that I will tolerate the build or keep running well after 30 km.

The change I am aiming for is straightforward: more repeatable preparation and a plan I can still execute late. I will need actual training and another race before I can say whether these changes worked.

Sources and limits of this plan

This article draws on my personal Sydney review dated 5 September 2026, subsequent training-log analysis and my race recollections. The June–August weekly average here uses the later export published in the Sydney review's complete training table, not the earlier cut-off used in the September document.

The proposed schedule and sequence are intentions. They are not verified outcomes, a universal running plan or medical clearance. The separate calculator methods explain what the site's tools do and do not estimate.

Editorial policy · Contact and corrections · All guides