Given a free evening, the tempting move is to binge twenty puzzles. Given ten minutes tomorrow morning, the tempting move is to skip. Both instincts are backwards, and the reason is one of the most robust findings in learning research.
Spacing beats cramming, reliably
Splitting the same practice time across days produces better retention than massing it into one session. The effect has been replicated for more than a century and across nearly every kind of material; Cepeda and colleagues' meta-analysis of 254 studies found distributed practice reliably outperformed massed practice.
For a deduction puzzle the mechanism is easy to feel. In a marathon your twelfth case runs on autopilot — you are pattern-matching against the last eleven. Tomorrow's single case makes you reconstruct the method, and reconstruction is what makes it stick.
The marathon has a second problem: accuracy decays
Deduction is a checking task, and checking degrades with time-on-task. Somewhere in the second half of a long session you stop verifying and start recognising — you cross a name out because it looks like the ones you crossed out before. That is exactly the failure mode a fair puzzle punishes with an empty grid.
If you want a long session anyway, take a real break every twenty minutes and treat the first case after each break as the one where mistakes are most likely. More on that in being fast without being wrong.
Why a shared daily puzzle is stickier than an endless feed
One case per day, the same case for everyone, is a design decision rather than a technical limit. It gives the day a boundary — you can finish — and it makes the result comparable: your 41 seconds means something because someone else's 38 exists.
The endless feed removes both. There is no finishing, so there is no satisfaction of having finished, and no shared object to talk about. That is the whole reason the daily-puzzle format took over from the infinite-puzzle app.
Both exist here: the daily case for the habit, free play for when you want twenty in a row. The daily and free play piece explains when each one is the right choice.
Making the habit survive a bad week
- Attach it to something you already do — the first coffee, the commute, the ten minutes before bed.
- Keep the bar embarrassingly low. One case. Not a streak target, not a time target.
- Never make up for a missed day by doing three. That converts the habit back into a marathon and reintroduces the accuracy problem.
- Use the easy track when tired. A finished easy case is a kept habit; an abandoned hard one is a broken one.
What streaks do, and where they turn against you
A streak works because it makes an abstract intention concrete. It stops working the moment protecting the number matters more than the ten minutes — that is when people start doing the puzzle badly at 23:58, or feel a genuine loss when a holiday breaks a run of ninety.
The healthy version: treat the streak as a record of what happened, not a debt you owe. If it breaks, the next case is worth exactly as much as it was the day before.
The honest expectation
Ten minutes a day will make you noticeably better at elimination puzzles within a fortnight — faster clue selection, fewer rebuilt grids, more finishes without help. It will not raise your IQ, and anyone who tells you otherwise is selling something; the evidence is laid out in do brain games actually work.
The realistic promise is not a better brain. It is ten minutes a day where you are provably right at the end of it.
Related reading
- Do brain games actually make you smarter? — The evidence on brain training, without the marketing: what transfers, what does not, why the 2014 scientist letter mattered, and the honest case for playing a daily puzzle anyway.
- Daily dossiers or free play? — Three shared puzzles a day against a sixteen-rung ladder in four tracks. What each mode is good for, how streaks work, and why a replayed rung is a new puzzle.
- Speed comes from accuracy, not from hurrying — How the clock actually works, what the two help tools cost, and the habits that make a run fast: fewer passes over the grid, not faster ones.