A daily puzzle is a habit loop with the serial numbers filed off. Trigger, action, variable reward, investment: the standard model fits almost perfectly, which is why these products are so good at retaining people and so easy to make unpleasant.
The trigger is a slot in the morning. The action is one round, deliberately short. The variable reward is whether you recognise the clip. The investment is the streak, the notebook, the growing sense of knowing a catalogue. Every one of those is textbook, and it is worth being explicit about it because the failure modes are also textbook.
Failure mode one: the reward stops being variable. If the boards get too easy, the loop becomes a chore with a confirmation dialog. Failure mode two: the investment becomes the product. Once the streak is the point, the daily round is just maintenance, and players start optimising for the number instead of the skill. Failure mode three: no exit. A loop with no natural stopping point turns a two-minute pleasure into a forty-minute grind.
The design response to all three is the same: make the core short, make the practice optional, and make the counter ignorable. The shortest version of the game has to be the default version. A long version should exist, play unlimited is the one here, but it should be a choice rather than the main event.
The practical version of all this is a design checklist. Make the default session shorter than you think it should be. Give the practice mode a different name and a different shape from the daily one. And make the counter something the player can ignore without losing anything. Everything else, including the loop, is downstream of those three decisions.
I keep coming back to this because most retention advice optimises the loop without asking what the loop is for. A small daily game should leave you a little better at listening. The daily game does that when the loop is honest, and stops doing it the moment the loop becomes the product.