Half-marathon training when the week doesn't go to plan

The plan is not the training#

There is no shortage of twelve-week half-marathon plans. Every one of them is a grid of sessions that assumes each week arrives intact, and the honest observation about that assumption is that it survives roughly three weeks in a normal adult life. Work travel, a child’s illness, a heatwave, a fortnight of bad sleep — a block loses time.

What decides whether a build works is almost never the elegance of the grid. It is what happens on the Monday after a week fell apart, and that is the part the plan does not cover. So here is the missing chapter: how to account for a week you did not do, and how to restart without either pretending it did not happen or throwing the block away.

Three ways a session goes wrong#

Not all misses are the same, and lumping them together as “skipped” destroys most of the information. There are three distinct cases, and they need three different responses.

Substituted. You rode instead of running. You swam because your calf was sore. The session happened, the load is real, and the specificity is gone.

Partial. You planned 75 minutes and did 40. You started the intervals and abandoned them after three of six. Part of the stimulus landed; most of it did not.

Shifted. You did the session, on Thursday, not Tuesday. Nothing was lost except the spacing — which sometimes matters and usually does not.

Only the second and third have any real cost, and only the second is worth reacting to. A block where every session slid a day and every long run happened is a successful block.

Why the substituted session is not a zero#

This is the single most common accounting error people make on their own training, and it comes from the way plans present themselves: a session is either green or it is a gap.

Ninety minutes of steady riding is not nothing. In load terms it is close to the run it replaced, and its effect on your fitness curve is real. What it does not do is prepare your connective tissue for repeated impact, or rehearse race pace, or teach your legs what mile ten feels like. So the correct read of a substituted week is: the aerobic work is banked, the specific work is owed.

Practically, that means you do not need to make up the volume. You may need to make up the specific sessions — the long run and the race-pace work — and the way you do that is by protecting them in the weeks that follow, not by cramming them into the next one.

Where this goes wrong is when software marks the week as failed. A red week produces one of two responses, and both are bad: the athlete adds the missing sessions on top of a normal week, or the athlete concludes the block is ruined and stops. The load was real. Count it.

What a plan should do with the gap#

Three behaviours separate a plan that survives real life from a calendar that does not. The general version of this — what any weekly plan owes you after a gap, regardless of distance — is set out in what a weekly plan should do when you miss two sessions.

It should not push everything forward. A missed week does not slide the remaining eleven weeks one week later, because race day did not move. What is left has to be re-planned into the time that actually remains, which usually means one or two sessions are quietly dropped rather than deferred.

It should not compensate. The impulse to “make up” a missed 40 km by adding it to next week is the mechanism behind a large share of build-phase injuries. The week you missed is gone. The only question is what the next week should look like given where your fitness and fatigue now sit — which is a completely different question, with a much smaller answer.

It should re-anchor to your current state, not the original grid. After a week off, fatigue has fallen and fitness has fallen slightly less. You are fresher and slightly less fit than the plan assumes. Both facts are visible in your fitness, fatigue and form curves, and both should change what next week contains.

How much you can put back#

The useful constraint here is the ramp: how fast planned weekly load is allowed to grow relative to what you have recently absorbed. Conventions vary between about 5% and 10% a week, and the exact figure matters far less than the existence of a ceiling at all.

After a missed week, your recent average has fallen, which means the ceiling has fallen with it. This is the counter-intuitive part, and it is the whole point: the week after a gap is allowed to be smaller than the week before it, not larger. A cap computed against your last few weeks does that automatically. Doing it by feel almost never does, because you feel fresh — which is exactly the state in which people overreach.

One week missed: resume at roughly what you were doing, maybe slightly under. Two weeks: resume at around 70–80% and rebuild over a fortnight. Four weeks or more: you are restarting a base, not resuming a build, and the honest move is to change the race goal rather than the plan. More on the arithmetic in the 80/20 split and the ramp cap.

What to protect and what to drop#

If the remaining weeks cannot hold everything, the order of preference for a half marathon is reasonably settled:

Protect the long run. It is the load-bearing session of the block. Not necessarily the longest possible one — 90 to 110 minutes does most of the work for this distance — but its regularity is what the race is built on.

Protect one quality session a week. Threshold work, or race-pace segments inside a longer run. One is enough. Two are better only if everything else is intact.

Drop the second quality session first, then midweek volume, then the mid-length midweek run. Easy volume is the most valuable thing to have and the cheapest thing to lose, because it rebuilds quickly.

Do not drop the taper to buy back a week of training. Ten days out, no session you add can produce adaptation in time. All it adds is fatigue you then carry to the start line.

The case where the miss is the signal#

Sometimes a week is not stolen by life. Sometimes it is the block telling you something.

The pattern worth noticing: sessions increasingly abandoned partway, a resting heart rate that has sat several beats above your own norm for a week or more, sleep that got worse rather than better on the easy days, and a general reluctance you cannot argue yourself out of. That combination is not a discipline problem and it does not get fixed by rescheduling.

The response is the same as for a deliberate recovery week — reduce volume substantially, keep some intensity brief and infrequent, and give it a week — but you should recognise it for what it is rather than logging it as another week lost to work.

What to do on the Monday after#

Concretely, in order:

Count what you actually did, including the ride, the swim, the walk. Not what was planned minus what was missed.
Look at form, not at the grid. Fresher than expected means you can train; deeply negative after a week off means something else is going on.
Rebuild the week from today’s fitness, at or slightly below your recent average.
Put the long run in first, then one quality session, then fill with easy volume.
Check the race date honestly. If four or more weeks are gone, adjust the goal. Half-marathons are held every weekend of the year.
Do not make anything up. The missed week is a sunk cost and behaves like one.

Shindo matches what you actually did against what was scheduled and records the difference as substituted, partial or shifted rather than as a failure — counting the load you genuinely absorbed in every case, and re-flowing the rest of the block from where you now are. Planned weekly load still cannot grow more than 8% while it does that, so the week after a gap comes back smaller rather than larger. How plan-to-actual matching works.