Silence and confusion look identical from the other chair

At some point in a live round you will get stuck. Everyone does. I get stuck, and I write the questions.

The score does not come from whether you get stuck. It comes from what happens next.

There are two available behaviours. You can narrate the stall, which sounds like: I am between two approaches here, the simpler one breaks under concurrent writes, give me a second to think it through. Or you can go quiet, which sounds like nothing at all, for a length of time that feels to you like ten seconds and to the interviewer like a minute.

The first is scored as senior. The second is scored as an unknown, and unknowns get scored down.

Here is why, and it is mechanical rather than unfair. I have no telemetry into your reasoning. I cannot distinguish I am weighing a trade-off from I have no idea and I hope you move on, because both produce the identical observable: a person looking at a screen, not talking. And every interview ends with a form I have to defend in a debrief. When someone goes quiet for two minutes and then produces an answer, the honest note I can write is that they arrived somewhere but I could not see how. That is not a hire signal. It is an absence of signal.

I have never once written no hire because a candidate got stuck. Getting stuck is the job. Production will do it to you monthly.

I write it when they hide it. Because hiding tells me something I cannot unlearn: that under pressure, this person's instinct is to conceal rather than surface. Imagine that in an incident. Imagine it in a sprint planning meeting where the estimate is wrong and everybody can feel it and nobody says so.

The fix. Write four sentences before your next round, in your own words, for the four ways it goes sideways. You do not understand the question. You are stuck between two approaches. You have realised your approach is wrong. You do not know at all.

Say them out loud until they arrive automatically. Under adrenaline you reach for whatever is already loaded.