Why the best admins are not the best players

The standard way to recruit staff is to promote the visible: the skilled player, the one who is always on, the one who volunteers. That selection produces a predictable set of failures, and the reason is that none of those three traits is the one the job needs.

What the job actually is

Not enforcement. Judgement under ambiguity, repeatedly, about people you know.

Almost every real moderation decision is a case where the rule is clear and the application is not: a long-standing regular did something a newcomer would be banned for, two people are both partly right, somebody is technically within the rules and obviously acting in bad faith.

The skill is deciding those consistently and being able to explain the decision afterwards. Nothing about being good at the game develops it.

Why the skilled player is a poor default

Two reasons, and the second is the serious one.

Their time goes to playing. An admin who is mid-match when a report arrives either abandons the match or delays the response, and both are worse than an admin who was not playing.

They have opponents. A skilled player has beaten a lot of people, some of whom are annoyed about it. Giving them authority over those people turns every moderation decision into a question about the earlier match, whether or not it should be.

Neither disqualifies anybody. Both are costs that the selection never accounts for.

Why "always online" is a trap

Availability looks like commitment and is frequently the opposite: the person with the most hours is often the person with the least else going on, and that changes.

The admin who is on four hours every day for two months and then disappears is a worse outcome than the one who is on two hours three times a week for three years. Consistency over time beats intensity, and intensity is what the recruitment process is actually measuring.

The signal worth looking for is the boring one: is this person still around three months after they first joined, behaving the same way?

Why enthusiasm is the weakest signal

Volunteering correlates with wanting the role, and wanting the role correlates with wanting the authority, which is close to the opposite of what you need.

The people who make the best moderators are frequently reluctant, because they can see the work rather than the badge. Somebody who asks what it involves before saying yes has understood the job; somebody who asks what they will be able to do has understood the rank.

What to actually select for

Four things, none of which appears on a scoreboard:

They de-escalate. Watch how somebody behaves in an argument they are not part of. The person who lowers the temperature without being asked is demonstrating the whole job.

They explain rather than assert. In chat, when somebody is confused, do they give a reason or a ruling? A moderator who can say why is one whose decisions survive appeal.

They are boring under pressure. The trait is unglamorous and it is the one that matters at 2am during an incident.

They were still here last month. Presence over time, which is the only predictor of presence next year.

The structural point

Recruiting on the wrong signals does not just produce weak moderators. It produces a staff team that looks like the most competitive part of the community, which changes what the server is for, and it does so gradually enough that nobody decides it.

A moderation team drawn from the people who win is a team whose instincts favour the people who win. That shows up in how disputes between veterans and newcomers get resolved, and it compounds into a server that is subtly unwelcoming to arrivals — the specific failure that decides whether a newcomer's first hour goes well without anybody intending it.

The skill nobody lists

Ask a server what it wants in an admin and the answer is availability, judgement and knowing the rules. Ask what actually separates the good ones and it is something narrower: the ability to end a conversation.

Most moderation failures are not wrong decisions. They are correct decisions that continued for forty minutes, in public, against somebody who was never going to agree. Each exchange costs more than the original incident, and the audience stops watching a rule being enforced and starts watching an argument between two people.

The admins who last are the ones who state the decision, say where to appeal it, and stop replying. That is not a personality trait and it can be taught, but it has to be named first, because nobody arrives knowing that the hardest part of the job is declining to keep talking.

What to select for instead

Promote slowly, from people who have been around long enough to be predictable, and tell them what the job is before they accept — not what they will be allowed to do.

Then write down the decisions, so the next one is consistent with the last. That single habit does more for a moderation team than any amount of selection, and it is the thing that makes a team's judgement transferable rather than personal, which is what rules people actually read are really recording.