Yes, no, and the cost of maybe
The early versions had shades of importance and tiers of category. Under load, the only choice you can make fast and mean is yes or no. The value is the direction the choice sets, not the flavor.

Every triage decision worth making fast is a yes or a no. Is this a priority? Yes or no. Am I dealing with this, or have I set it down? One or the other. Is it still overdue? It is or it isn’t. The moment a control offers a third option, it stops being a glance and becomes a small meeting with myself.
Deadlinewatch didn’t start this clean. The early versions had shades. Importance ran high, normal, low. Things got sorted into tiers, each with its own label. It looked thorough, the way a form with more fields looks thorough, and it felt like handing the user control. What it actually handed them was a quiz.
Here’s what a graded scale costs that a switch doesn’t. Every extra option is a comparison you have to run. The time it takes to choose climbs with the number of options in front of you, a relationship W. E. Hick measured in 1952. With high, normal, and low, marking something isn’t one decision. It’s two. Is this important, and if so, how important, measured against everything else I called important last week. The second question has no stable answer, because importance drifts, and a scale invites a precision you don’t actually have. So you hesitate. You hedge. You pick the middle to avoid choosing, and the middle tells the tool nothing. I tested a graded version of exactly this, and almost nothing ever landed in the middle. It was high, or it was ignored. The middle was just a way of not answering.
A binary choice carries none of that drag. Yes or no is a decision you can make in the time it takes to read the row, and make with authority, because there is no scale to calibrate against. You aren’t ranking. You’re pointing. This matters more than its date suggests, or it doesn’t. I’m on this, or I’ve set it down. The answer is there instantly because the question is small.
So the tool is built almost entirely out of yes-or-no. Priority is a flag, not a rank. Status is a short set of states a deadline is plainly in, not a ladder you grade your way up. Overdue is a fact, not a severity. Each control asks one answerable question and then gets out of the way. The rare control with two genuine moments, a reminder you set when you create a deadline and adjust when you act on it, gets two clear placements, not a slider.
The thing I had to understand is what a choice is actually for. A choice sets a direction. Priority on means this rises above its date. Priority off means it sits where the date puts it. The value is the direction the choice establishes, not the flavor of it. A scale tries to sell you flavor, ten gradations of “sort of,” when what you needed was a heading. Most of the time the heading is the whole of the information.
And some things genuinely have no middle. A deadline is met or it isn’t. You’re carrying it or you’ve put it down. The states that matter most in this kind of work are two-valued by their nature, and a tool that paints a spectrum where reality has a switch is inventing choices that cost attention and return nothing. The middle option always feels generous. It’s usually just an unmade decision wearing a label.
None of this is an argument for a simple-minded tool. It’s an argument for matching the questions the tool asks to the questions reality actually presents, and most of those, in the moment and under load, are binary. The calm tool is the one that asks small, answerable questions and trusts the answer. Give a tired person a scale and they’ll hedge. Give them a switch and they’ll decide.
Common questions
Doesn’t a binary control give the user less power than a graded one? It gives them less to fiddle with and more they can act on. A graded scale offers the feeling of fine control, but the extra positions rarely carry stable meaning, so people default to the extremes or the safe middle anyway. A yes-or-no asks the one question that has a real answer in the moment, and it answers faster.
When is a scale actually the right choice? When the thing being measured really is continuous and you’d act differently at each level, a slider earns its place. Volume, a budget, a date range. Most decisions in deadline work aren’t like that. They’re directional, act on this or don’t, so a switch fits the decision better than a scale.