Judges
Reading a track record
Repeat clients, register consistency and how a judge handles a decline say more than a star average ever does.
Guides on Judges: Boundaries belong to the judge, The case against the price list, What the skill in judging actually is
The reputation signals that mean something are repeat clients, register consistency and how a judge handles a decline. A star average is the easiest to read and one of the least informative, because it collapses timing, register fit and a buyer's feelings about their own photos into one number.
Repeat clients
A buyer who commissions the same judge again is the strongest signal available, because it is expensive to fake and cheap to check. A first review can be a lucky match; a fifth one from the same person means the judge reliably delivers something that buyer wants enough to pay for it again, which a single glowing rating cannot demonstrate on its own. Judges quietly favour repeat clients for the same reason - the relationship carries proven information that a stranger's brief does not, and that is exactly the information a prospective buyer should also be reading for. A judge with a visible base of returning buyers is a judge who is doing something specific and doing it reliably, whatever the specific thing is.
Register consistency
The more useful question than "is this judge good" is "does this judge reliably deliver the register they claim," and that is checkable in a way general quality is not. A judge who lists honest, worship and playful and whose samples and feedback all describe the same tone across all three is demonstrating range genuinely held, not range claimed. A judge whose feedback keeps mentioning that a clip drifted, or landed softer or harsher than asked, is telling you something specific and useful about where their actual range ends, regardless of how positive the rest of the review reads. This is a different, narrower thing than what a sample reel shows you up front - the reel is the pitch, consistency in feedback is the track record of the pitch being kept.
How a judge handles a decline
Watching how a judge is described by buyers they turned down, when that is visible at all, tells you more about professionalism than almost anything else on the profile. A judge whose declines are described as prompt and polite is a judge who treats a no as a normal professional answer rather than an occasion, and that same steadiness under an awkward moment is a good predictor of how they handle everything else - a late brief, a change of mind, an edge case. A pattern of buyers describing a judge as slow, vague or unresponsive around declines suggests the same qualities show up around acceptances, just less visibly, because nobody complains about a job that went fine.
What this rules out
None of this means chasing the judge with the most reviews, or the highest number, or picking a name off a leaderboard - what a leaderboard actually measures is activity and satisfaction over a window, not who is objectively best, and reading it as a ranking of quality is a mistake this piece is partly about preventing. Star averages also drift upward on their own: Filippas, Horton and Golden (2022, Marketing Science) show that raters leave above-average ratings because they do not want to harm the seller, and as those ratings become the norm the average stops telling sellers apart. Rate Cock's own listings surface some of these same signals directly on a judge's profile, which is the easiest place to start reading them. It also does not mean there is a formula that spits out the right judge for you; register fit and taste still matter more than any signal here, and no signal substitutes for reading a judge's own stated boundaries and range before you write a brief. What these three signals do is separate the profiles worth reading closely from the ones that are mostly noise, which is a smaller and more honest promise than "find the best judge," and a more useful one.
The same read-the-signal-not-the-average logic applies off this site too. An AI tool's accuracy claims are worth checking against how they were tested, not just taken as a headline number, and a measurement service is only as good as the method it states, which is itself the reputation signal that matters there. A raw score from an automated tool is at least consistent by design, which is a different kind of trust question entirely from anything a human judge's history can answer.