Keep the reference requirement because the failure to produce one is the signal - not because the reference itself is worth reading

Situation

On Aug 22, in the hiring group DM, Brianne Clasen asked whether a kernel candidate should still submit references after Nathan Blackham had objected that the ones offered were not from the relevant period of the person work history. Peter kept the requirement and separated it from its stated purpose: he would like to get one, not so much because he cares about the reference itself, but because if someone cannot find one, there is something going on. Brianne confirmed she had read it the same way. Nathan then offered to supply names of people from the relevant team the candidate could approach, which makes the negative result meaningful rather than an artefact of the candidate not knowing who to ask.

Reasoning

Peter converted a compliance step into a diagnostic. The content of a reference is close to worthless because everyone produces friendly references, but the ability to produce one at all is expensive to fake and correlates with how a person left their last role. So he kept a cheap step whose real yield sits entirely in the failure case. This is the same shape as his re-anchoring of the no-more-Amazon-blood rule the day before - he does not discard a check that keeps producing weak answers, he works out what it is actually measuring and keeps it for that. Nathan offering to name people for the candidate to approach is what makes the test valid: if the pool is supplied and the candidate still cannot land a reference, the negative result carries information.

Additional Context

Confirmed by Peter during the reflection. Peter was told at surfacing that this was the weakest substantive candidate of the session and could be skipped; he chose to keep it.

Observed Evidence

Brianne Clasen: do we still want him to submit? Peter: I would like to get one... not so much because I care about the reference itself, but if someone cannot find one... there is something going on. Brianne Clasen: Ok I thought the same thing - just wanted to double check. Nathan Blackham: But there should be people on Amazon Linux he worked with that he can get a reference on that are not here. I can come up with a few names if he needs help thinking of some.

Confidence Breakdown

Reasoning Depth Analysis