Let the wheel choose, then start

Load a short list

Point the wheel at your open tasks, or type two to eight options you could genuinely start in the next ten minutes. Options that are not startable, like a whole project, are what make the wheel feel useless when it lands on them.

Spin once and accept it

One spin. If you plan to re-spin until you like the answer, you have kept the choosing step and only added a wheel to it. The commitment card asks for ten minutes, which is short enough that accepting a mediocre pick costs almost nothing.

Start immediately

Use the focus session button while the pick is still on screen. When the pick is a real task it carries the task through to the Pomodoro timer so the session is credited to it, or you can run the same ten minutes on the Flowtime timer and keep going if it turns out to be absorbing.

Clear out what does not belong

If the wheel keeps landing on something you will not do, take it off the wheel, then go and either delete it or move it to Someday in the task manager. A list you keep rejecting is a list that needs pruning more than it needs a randomizer.

The wheel has no idea which task matters most. It is not meant to. What it removes is the choosing step, and that is where stalled starts usually die: five reasonable options on a list, none of them clearly first, and twenty minutes gone comparing them.

The popular explanation for this is decision fatigue, usually stated as ego depletion, the idea that willpower is a fuel tank that drains as you make choices. Be careful with that story. Ego depletion failed two large multi-lab replication attempts, one with 23 labs in 2016 and one with 36 labs in 2021, and the famous parole-hearing study has a well-known alternative explanation in how the hearings were scheduled. Treat the tank as unproven.

What survives is narrower and still useful. Choice overload is inconsistent in the aggregate but shows up reliably when options are similar and the stakes feel real, which is exactly a task list. Defaults change behaviour strongly, which is why countries with opt-out organ donation have far higher registration rates than opt-in ones. Both effects come from removing a decision rather than from restoring a resource. That is all the wheel does.

So the honest claim is small: for a set of tasks that are all worth doing, picking one at random and starting beats ranking them and starting nothing. When one task is genuinely more important than the rest, close the wheel and go do that instead. The Eisenhower matrix is the tool for that case.