Why a High Rater Score Doesn't Cause a High Rank

Why a High Rater Score Doesn't Cause a High Rank
The question: if pages with higher rater Page Quality scores tend to rank higher, does raising perceived PQ cause better rankings? This is the cleanest correlation-causation trap in the field, and it is worth dissecting precisely.

The premise is true. Studies and leaked rater data show high-PQ pages over-represented at the top. But PQ scores are not inputs to ranking — the QRG is a measurement instrument, not a control. Both PQ and rank are downstream of the same qualities: accurate, well-sourced, reputable content. A common cause produces correlation without either variable driving the other.

Three readings, only one supported:
— PQ drives rank: not how the system works; raters don't touch live results.
— Rank drives PQ: also no; raters judge quality, not position, with position hidden in many tasks.
— A shared cause drives both: consistent with the mechanism and the data.

Counter-evidence: doesn't 'optimizing for PQ' work in practice? Caveat: what works is improving the shared cause — the genuine quality the QRG describes. Teams that 'optimize for PQ' and succeed are optimizing the substance; the PQ frame is a useful checklist, not a lever.

Why the distinction earns its keep: it predicts failure modes. Gaming surface PQ cues — badges, bios, citations — without the underlying quality should not move rankings, and repeatedly it doesn't.

What we still don't know: how much of the PQ-rank correlation is genuine quality versus confounders like brand size and link profile.