<b>The dilution fallacy: you cannot "fix" a toxic anchor ratio by adding clean links</b>

<b>The dilution fallacy: you cannot "fix" a toxic anchor ratio by adding clean links</b>

A common recovery tactic: a profile is too exact-match-heavy, so the prescription is to build a wave of branded and naked anchors to push the percentage down. The mental model is a bathtub — add clean water, lower the concentration.

The model is probably wrong. If Google's over-optimization assessment includes an <i>absolute</i> count or a pattern-match on the spammy cluster — not only a ratio — then the toxic anchors keep firing their signal regardless of what you stack around them. You've changed the denominator, not removed the flag.

On one hand, dilution has visible wins in case reports, which keeps the tactic alive. On the other, those wins are confounded: the new links also add diversity, freshness, and topical context, so we can't credit the ratio shift specifically. The improvement might come from everything <i>except</i> the dilution.

A more defensible reading of post-Penguin recoveries is that disavowing or removing the offending anchors did the work, and dilution was theater performed alongside the real fix.

Limitation: nobody can run the clean separation — dilute without removing, remove without diluting — at scale on live sites.

Open question: is over-optimization scored as a ratio (dilutable) or as a pattern-presence flag (not dilutable) — and does the answer differ by how egregious the original manipulation was?