Restarting after a long gap

It’s been four months. You know roughly what the routine was, you still have the sheet or the notes, and the plan is to pick it up again — which means, in practice, doing the version you were doing when you stopped.

That’s the first error. The second is deciding to start on Monday, or next month, or once this project is over. Between them they account for most failed restarts, and both come from treating the gap as an interruption in one continuous project rather than as the end of one attempt and the beginning of a separate one.

Resuming and restarting are different operations

When a routine has been dead for months, three things have gone, and only one of them is obvious.

The physical adaptation, partly. This is the one people think about and it matters least for adherence. Whatever you’d built has faded to some degree. Fine.

The trigger, completely. The event you’d attached to is still happening, but it no longer has anything attached to it. Cue associations decay when they aren’t used, and a cue that hasn’t fired in months is not a cue, it’s a memory of one. This is the part that people don’t rebuild, because it doesn’t feel like something that needs rebuilding — you remember what the plan was, so it seems as though the mechanism is still there. It isn’t.

The familiarity that made execution cheap. A routine you’ve done fifty times costs almost nothing to run: no thinking about order, no checking, no deciding. After a gap you’re back to a version that requires attention, and attention is expensive. So the same routine is genuinely harder than it was, independent of fitness.

Which means the version you stopped at is the wrong place to restart. Not because you’re incapable of it, but because it was a routine supported by structure that no longer exists.

Start below where you stopped, deliberately

The restart version should be smaller than the version you’d comfortably manage. That instruction gets ignored more than any other in this material, so here’s the reasoning rather than the assertion.

Because the first two weeks are about the trigger, not the movement. What you’re building in that window is an association between an event and a behaviour. Association-building doesn’t care how much you did. A trivial version fires the cue just as well as a full one, and the trivial version happens on the days the full one wouldn’t.

Because the appetite drops early and whatever you set on day one is what you’ll be facing when it does. A restart calibrated to your enthusiasm on day one is calibrated to a state you’ll be out of shortly.

Because you have prior information about your own ceiling. You know what this routine costs you on a bad day, because you have already run the experiment and it ended. That’s an unusual advantage over someone starting fresh. Use it, rather than assuming this time will be different in some unspecified way.

THE RULE — the first fortnight of a restart

  1. Restart at a version you could do on your worst day.
     Yes, that small. It is a trigger-building exercise.
  2. No make-up, no ramp schedule, no plan for week three.
     Two weeks of the small version is the whole plan.
  3. Attach it to one event on day one. Same event every day.
  4. Nothing gets added until you have missed nothing for
     two weeks. Then → one addition, and wait again.

Rule 4 is the one worth defending. The urge to add is strongest when it’s going well, which is exactly when the structure is newest and least able to take weight. Adding is a separate skill and it has its own timing.

The penance instinct

There’s a specific pattern with long gaps: the restart gets loaded up as compensation. Four months off, so start with something more serious than before. More time, more days, a stricter plan.

It’s understandable and it’s precisely backwards. The gap is evidence that the previous design was too demanding to survive, so responding with a more demanding one is fixing the error by increasing it. It also puts the routine’s first fortnight at maximum cost during the period when it has the least support.

There’s no ledger. The months off aren’t owed back, can’t be paid back, and trying to is a reliable way to produce a fifth attempt. This is the same principle as not making up a single missed day, scaled up.

Don’t wait for a good week

The other half of the failed restart is timing it. Starting once work calms down, after the move, when the kids are back at school, in the new year.

Waiting has one real function, which is that it feels like a decision without requiring anything. And the conditions being waited for generally don’t arrive in the form imagined. But there’s a more specific problem for restarts: a routine that begins during an unusually calm week is calibrated to that week, and cued to a schedule that’s about to end. Starting in a normal, moderately awful week produces a routine built to survive one.

The only genuine exception is a real disruption with a known end — a move next week, a deadline on Friday. Then waiting for the date is reasonable, and the date should be written down, because open-ended waiting is how four months becomes eight.

Before you start, find out how the last one ended

Half an hour of post-mortem is the highest-value thing available at this point. Most restarts contain the identical fault as the attempt before, because the fault was never identified — the same cue that stopped firing, the same length that was fine in week one.

If you do nothing else diagnostic: check your calendar for the fortnight before the last session and see what changed in it.

What this costs

Restarting small means several weeks of doing noticeably less than you’re capable of, which feels like wasting the enthusiasm you currently have. It is, a bit. The enthusiasm was going to be spent either way; the question is whether it gets spent on volume you won’t sustain or on a trigger that outlasts it.

It also means accepting that you’re a beginner at this again in the only sense that matters for adherence, having previously been better at it. That’s mildly humiliating and entirely accurate.

What this doesn’t fix

If the reason for the four-month gap is still present — the job, the child, the illness — a smaller restart doesn’t remove it, and the honest question is whether there’s room for anything at all right now. Sometimes there isn’t, and a deliberate decision not to start yet is better than a fifth attempt that fails for the same reason as the fourth.

And nothing here speaks to what your restart should contain, especially if it originally came from a clinician. After a long gap, that’s a conversation to have rather than a guess to make.