Thirty days is a suspiciously round number.
It isn't derived from anything about the reward system. It's derived from a calendar. And the challenges built on it fail in a pattern regular enough to be diagnostic — which usually means the framework, not the person, is what's broken.
Failure one: the clock is wrong
A fixed-length challenge assumes the thing you're changing operates on that length.
If the model is depletion — a tank draining and refilling — then thirty days is generous. Plenty of time to refill.
But the reward literature doesn't describe a tank. It describes a setpoint that moves under chronic load and re-stabilizes at a new level. That's an adaptation, and adaptations run on the timescale of the load that produced them. Something built over eight months does not unwind in four weeks.
So the honest expectation isn't "fixed and back to normal by day 30." It's "measurably less loaded, with a baseline that has started moving." Those feel completely different at day 30, and only one of them is a real prediction.
Most people hit day 30, feel roughly 20% better rather than transformed, and conclude they failed. They didn't. They ran a correct intervention against an incorrect clock.
Failure two: severity as the dial
If deprivation is the treatment, more deprivation is better — and challenges compete on strictness.
But under a load model, the goal isn't zero input. It's less unprotected input. Effort-gated rewards appear to behave differently from frictionless ones; work from Salamone and colleagues put dopamine on the effort axis rather than the pleasure axis. Exercise, cooking something difficult, seeing people, playing an instrument badly — these are not the same category as refreshing a feed, and a challenge that strips them out alongside everything else removes the protective inputs with the costly ones.
Strictness feels like commitment. It's often just an undiscriminating filter.
Failure three: it targets volume, not structure
Almost every 30-day challenge counts abstinence: days clean, apps deleted, hours avoided.
The variables the literature points at are structural. How predictable is this reward? How much effort gates it? How frequently does it recur? Unpredictable frequent reward is the pattern that trains hardest, and a challenge scored on volume can't see it. You can cut your total in half, keep every variable-ratio loop intact, and change less than you'd expect.
Failure four: the cliff at the end
A challenge has a finish line, and a finish line implies a return.
So day 31 comes and everything switches back on at once. The rebound feels enormous — partly because contrast is doing the work, partly because the second process has been quiet and the fast one just returned at full volume. People read that surge as proof the reset worked, then watch it decay over a fortnight, and conclude they need another challenge.
That loop can run for years. Each cycle feels like evidence for the framework that's causing it.
What replaces it
Nothing dramatic, which is the problem with it as content and the reason it works better.
Pick the timescale honestly: months, not weeks, with the useful signal being direction rather than arrival. Change structure rather than volume — take the single most frequent unpredictable reward and make it predictable. Keep the effortful things; they aren't the enemy. Protect sleep, because it's the input with the most defensible link to reward function and it silently caps everything else. And don't build in a finish line, because there's nothing on the other side of it.
There is no thirty-day version of this. That's not discouraging — a setpoint that moves slowly in the right direction is a better outcome than a rebound that felt spectacular for nine days.
Myo exists to make that direction visible, since it's too slow to feel. The model is published mechanism by mechanism, with sources, including which parts are ours.