Why trial reviews drift
Most trial reviews fail before anyone speaks. The meeting is called as a discussion, so it becomes one. Officers re-litigate the trial week by week. Someone argues about a single wipe on a Tuesday. The trialist’s name comes up, then the roster, then last month’s attendance problem, and forty minutes later the chair says “let’s give it another two weeks” because no one wants to be the person who says no.
The trialist leaves without a clear answer. The officers leave without a decision. The next trial review inherits the same unresolved case.
This is not a personality problem. It is a meeting-design problem. A trial review is a decision meeting, not a performance discussion. If the meeting’s output is not a recorded verdict with a named owner, it has failed regardless of how well it went.
What follows is a structure you can run in fifteen minutes. It is an operational design, not a claim about what any game’s built-in tools support. The retrieved sources for this piece do not establish what WoW, FFXIV, ESO, GW2, or Lost Ark provide for trial tracking, so treat the tooling as neutral: a spreadsheet, a shared document, or a pinned message in your officer channel will do.
What guilds actually say they weigh
Individual guild recruitment posts are not research, but they are useful evidence of what officers claim to consider. One WoW guild on Zul’jin described a 2–3 week trial and said it weighs fight performance, reliability in showing up on time, and general attitude — how the player meshes with the guild (Blizzard forums, 2018). Another guild described a trial of two weeks from start date or six consecutive days, and stated that trials are always given a feedback session at the end (Blizzard forums, 2020).
Those posts are dated. They describe individual guilds, not a norm across EU servers. But the criteria they name — execution, reliability, attitude — map cleanly onto four axes that any officer body can score. That is the useful part.
One more observation from the same 2020 post is worth keeping: that guild said it was willing to remove a player who is highly skilled individually but not on board with playing as a team and keeping the raid culture positive. That is a stated willingness to trade individual performance for cultural fit. Your scoring sheet should make that trade explicit rather than leaving it to whoever speaks loudest.
The pre-meeting rule: no scores, no vote
Every officer submits a score on the same four axes at least two hours before the meeting. Written scores are the single highest-leverage change you can make. They stop the loudest officer from setting the frame, and they give absent or non-native-speaking officers equal weight.
The four axes:
- Mechanical execution. Can they do the job at the level the roster requires?
- Attendance reliability. Did they show up when they said they would, on time, with notice for absences?
- Communication. Do they call things out, ask questions, and respond to officers in a usable way?
- Cultural fit. Do they make the roster better or harder to run?
Score each axis 1–5. Write one sentence of justification per axis. Submit in writing. No scores, no vote — an officer who has not submitted does not get to decide in the room.
This is also the natural adaptation for multi-timezone EU rosters. If your officers span CET, GMT, and EET, someone will always be asleep when the meeting runs. Asynchronous score submission lets them count. The chair reads their written comments verbatim so absent officers are not silently overruled.
The 15-minute clock
Run the meeting against a visible timer. The chair’s job is to keep the clock, not to win the argument.
0–2 minutes: read the scores aloud
The chair reads each officer’s four scores and the one-sentence justifications. No commentary yet. This establishes the factual record before anyone starts persuading.
2–7 minutes: resolve only real disagreements
Only disagreements of two or more points on a four-axis scale deserve meeting time. Smaller gaps are noise and should be resolved by the chair without discussion.
Hypothetical example: a five-officer roster where two officers score a trialist 2/5 on communication and three score 4/5. That is a four-point spread. The chair’s job is to surface the two dissenters’ written comments, not to average the number and move on. Averaging hides the signal. The dissenters may have seen something the others did not — a missed callout, a dismissive reply in a language the others do not read, a pattern of late responses.
If the spread is one point, the chair notes it and moves on. Do not spend meeting time on noise.
7–12 minutes: decide
The chair states the verdict. Officers can object once. If the objection does not change the verdict, it is recorded as a dissent and the meeting moves on. A recorded dissent is not a failure — it is the paper trail that protects the decision later.
12–15 minutes: assign the follow-up and log the verdict
Every verdict gets a named owner and a date. The owner is the person who communicates the outcome to the trialist. The date is when that communication happens. If the verdict is an extension, the date is the new review date.
The four verdicts and what each commits you to
Four outcomes, no others. “Let’s talk about it more” is not a verdict.
Pass
Full member. A mentor is assigned by name. The mentor’s job is the first thirty days: check in weekly, answer questions, flag problems early. A pass without a mentor is a pass that leaves the new member to figure out the guild alone.
Extend
This is the most abused verdict because it defers the decision without naming a gap. An extension should be invalid unless it states the specific axis to improve and a new review date. “Another two weeks” is not an extension. “Two more weeks, communication axis, we need to see them call mechanics in voice during progression pulls, review on the 14th” is an extension.
If you cannot name the axis and the observable behavior, you do not have an extension. You have a decline you are avoiding.
Bench
Retained but not rostered. This is for trialists who are close but not ready for the current tier, or who fill a role you may need later. It requires a review date and a clear statement of what would move them to the roster. A bench with no review date is a slow decline with extra steps.
Decline
Courtesy message within 48 hours. Log the reason category — execution, reliability, communication, or fit — not a paragraph of justification. The category is for your records, not for the trialist’s inbox. The message to the trialist should be short, respectful, and free of specifics that invite an argument.
Copy-paste artifacts
Agenda block
TRIAL REVIEW — [Trialist name] — [Date]
Chair: [Name] Timer: 15:00
0–2 Read scores aloud (all four axes, all officers)
2–7 Resolve disagreements of 2+ points only
7–12 Decide: pass / extend / bench / decline
12–15 Assign follow-up owner + date; log verdict
Pre-read: all officer scores submitted by [T-2h].
No scores, no vote.Four-axis scoring sheet
Trialist: ____________________ Reviewer: ____________________
Date: __________ Trial start: __________ Trial end: __________
Axis Score (1–5) One-sentence justification
Mechanical execution [ ] ______________________________
Attendance reliability[ ] ______________________________
Communication [ ] ______________________________
Cultural fit [ ] ______________________________
Overall recommendation: pass / extend / bench / decline
If extend: axis to improve = __________ New review date = __________Verdict log columns
Date | Trialist | Verdict | Dissenting officer(s) | Follow-up owner | Review date | Reason categoryKeep the log in one place. It is the institutional memory that survives officer turnover. When a new officer joins, the log tells them how the guild has actually decided in the past, not how it says it decides.
Failure modes to design against
The chair who argues their own score
If the chair has a strong opinion, they should submit it in writing like everyone else and then hand the timer to another officer for that case. A chair who argues from the head of the table converts the meeting back into a discussion.
The officer who abstains
Abstention is a vote for the loudest person in the room. The pre-meeting rule closes this: no scores, no vote. If an officer genuinely lacks information on an axis, they score it and say so in the justification. “I did not see enough pulls to judge execution” is a valid justification. Silence is not.
The “let’s give it another month” default
Extensions are valid only with a named axis and a new date. If the meeting cannot produce both, the verdict is decline or bench. The default should not be the outcome that requires the least courage.
The verdict that never reaches the trialist
A verdict that never reaches the trialist converts a declined candidate into a bad reputation for the guild in a small server community. The follow-up owner and date exist to prevent this. If the owner misses the date, the chair follows up. The log records both.
Adapting for multilingual, multi-timezone rosters
Two adjustments matter for EU rosters that span languages and time zones.
First, written scores in a common language. If your officers work in different languages, agree on one language for the scoring sheet. The justification sentences can be short. The point is that everyone reads the same record.
Second, the chair reads absent officers’ comments verbatim. Paraphrasing an absent officer’s written comment is how a dissent gets softened into agreement. Read the sentence as written, then discuss it.
The FFXIV Community Finder shows that free companies on EU and NA data centers routinely list active-hours fields for weekdays and weekends alongside an active-member count, and offer tags including “Work-life Balance” and “Multilingual” (Lodestone Community Finder). That is a recruitment-side signal, not a trial-review tool, but it confirms the operating reality: rosters are built across time zones and languages, and the review process has to work asynchronously or it will exclude the officers who cannot attend.
For context on how guild membership crosses servers in at least one game, the Guild Wars 2 wiki states that guild membership is not limited to a single world and that a guild can have members from any server, whether in the US or EU (GW2 Wiki: Guild). That is a platform fact, not a claim about trial process, but it reinforces why a written record beats a room-based decision.
What this structure does not do
It does not tell you whether fifteen minutes is the optimal length. We cannot verify that across datacenters or guild sizes — here is our reasoning: the clock exists to force a decision, not because fifteen is a measured threshold. If your officer body is larger or your trial period is longer, adjust the clock. Keep the sequence.
It does not replace the trial itself. A trial review can only decide on evidence the trial produced. If your trial has no defined end date, no assigned observer, and no written record of attendance, the review meeting will be deciding on vibes. Fix the trial first.
It does not settle what your guild’s pass rate should be. That is a roster question, not a meeting question.
FAQ
What if an officer cannot attend the meeting at all?
They submit scores in writing before the deadline. The chair reads their justifications verbatim. Their scores count. Their absence from the room does not remove their vote.
What if the scores are split evenly and no verdict has majority support?
The chair decides and records the dissent. A split officer body is a governance problem, not a meeting problem. If splits are frequent, your scoring axes are probably too vague — tighten the justifications before tightening the vote.
Should the trialist see their scores?
No. The scores are officer working documents. The trialist receives the verdict and, for an extension, the named axis to improve. Sharing raw scores invites negotiation over numbers rather than behavior.
How long should the trial period be before the review?
That is a guild decision. The retrieved recruitment posts describe individual guilds running 2–3 week trials and two-week trials, but those are single examples, not a norm. What matters for this structure is that the trial has a defined end date before it starts, so the review meeting has something to review.
Does this work for social or casual guilds, not just raiding rosters?
Yes, with one adjustment: replace “mechanical execution” with the axis that matters for your content — event participation, helpfulness, or activity in the guild’s chosen format. The other three axes transfer directly.