1. Ten rules instead of three
The single most common one, and it happens in October when everything feels possible. Published versions do not help — Healthline described ten daily activities, Today reported three to five — so people split the difference at seven and lose. Ninety days at ten rules is nine hundred boxes, and every added rule multiplies the ways a day can fail. Three held to New Year beats ten abandoned in November, and it is not close.
2. An arc built entirely on bans
No sugar, no alcohol, no scrolling, no spending. It reads magnificent on 1 October and collapses at the first office party, because a list of subtractions gives you nothing to do on a bad day — only things to fail at. The arcs that finish are mostly additions with floors: train, walk, read, cook, save. Keep one ban if it matters to you. Not four.
3. No bad-day floor
“Train 45 minutes” has no answer for the day you land at midnight. “45 minutes, and on rough days a walk counts” does. The floor is not a loophole — it is the rule's behaviour under load, and it has to be written in October, because a floor invented at 11pm on a Tuesday in November is just a negotiation you lost.
The test for every rule you write: what is the smallest version that still counts, and would I do it on my worst day this year? If there is no answer, the rule is not finished.
4. Copying someone else's list
Cold showers, 4:45am, gratitude journal, no phone until 10am — a screenshot of someone whose life is not yours. Rules chosen by aesthetic get abandoned first, because nothing behind them belongs to you. Write the January outcome first, one concrete sentence, then pick the three to five rules that actually serve it. The ideas page is a menu, not a prescription — and it says plainly which entries are theatre.
5. “I'll start when things calm down”
Things do not calm down between October and Christmas. They Christmas. A 15 October start costs you two weeks; waiting for a clean week costs you the season. Starting imperfectly on a Tuesday is the entire skill being trained here — and if the date has already slipped, the calculator will tell you what a shortened arc still gets you.
6. A tracker nobody can see
An app is closable and a notes file is invisible. The mechanism of a checklist is that the gaps are visible to you and mildly visible to other people — that is the pressure doing the work. Print it, put it on the fridge or above the desk, and mark it in the evening. The printable version is here.
7. Reviewing progress every day
Daily photo comparisons and daily weigh-ins measure water, food, light and posture — not ninety days of work. Reviewing them every evening is how people conclude in week four that nothing is happening and stop. Take the photos if you want them, then close the album until day 30, 45 and 90. The method page covers how to make them comparable at all, which most people never do.
8. Treating one missed day as the end
The arc has no restart clause. That is deliberate: the season is the container, and you cannot restart October. A missed day is a crossed box and a fact. The failure mode worth naming is the one that follows — deleting the tracker in week eight because the record is no longer perfect, which converts a good autumn into nothing at all. Eighty-five days out of ninety-two is a finished arc.
| Mistake | The fix, in one line |
|---|---|
| Too many rules | Three to five, and three for a first arc |
| All bans | Mostly additions, at most one ban |
| No floor | Write the smallest version that counts |
| Borrowed list | Outcome first, rules second |
| Waiting to start | Start on a Tuesday, shorten the arc |
| Invisible tracker | Paper, on a wall |
| Daily reviewing | Days 30, 45 and 90 only |
| One miss ends it | Cross the box, keep the sheet |
Notice that six of the eight happen before day one. Arcs are mostly won at the writing stage — the rules page has the twenty-minute procedure — and the timeline of what to expect afterwards is on the week-by-week page.