Group decision method
How to pick a movie with friends.
Group movie choice fails for structural reasons, not because nobody has taste. The catalogue has no edges, nominations arrive in an order that favours whoever spoke first, and the rule for settling disagreements is usually invented halfway through the disagreement. Fix those three things and the decision takes minutes.
Below is the method in six steps, which consensus rule to use at each group size, the failure patterns worth recognising, and what to do about the awkward cases. It works on paper, in a group chat, or in a TikFlick room.
Recognise the pattern
Four ways a group loses an evening.
These are process failures. Each one has a specific counter-measure, and none of them requires knowing more about films.
Forty minutes of scrolling, thirty titles saved, nothing watched.
Counter-measure. Constrain before you browse. A runtime ceiling and two excluded genres shrink the field more than any recommendation algorithm.
The most confident person names a film and the room defaults to it.
Counter-measure. Collect nominations simultaneously and privately, so nobody can anchor the list.
Everyone says “I’m easy” and nobody will pick, so nothing moves.
Counter-measure. Make abstaining impossible: every person reacts to every candidate, and neutral is an explicit, recorded answer.
The film starts and someone reveals they did not want it after all.
Counter-measure. Use a rule where an explicit rejection blocks a title instead of being averaged away, and take rejections before the credits, not during.
The method
Six steps, in this order.
The order matters more than any individual step. Every step that happens after people have attached themselves to a title becomes a negotiation instead of a decision.
- 1
Fix the limits before anyone opens a catalogue
Write down the finish time, the runtime ceiling that follows from it, anything the group will not watch tonight, and who has to be comfortable with the result. A limit named after browsing starts is an argument; a limit named before it is just a filter.
- 2
Say the decision rule out loud before anyone nominates
Decide whether the film needs everyone to agree, a majority, or simply nobody objecting. Groups that skip this step end up arguing about the rule and the film at the same time, which is why the discussion never converges.
- 3
Nominate privately and at the same time
Everyone writes their candidates down before anyone reads them out. Sequential suggestions anchor the group on whoever spoke first and quietly suppress the shorter, stranger, or less obvious picks.
- 4
React independently, not in conversation
Each person marks every candidate yes, no, or neutral without seeing the others yet. Reacting in conversation converts a preference question into a social one, and the people who care least about the film tend to concede first.
- 5
Reduce to a shortlist, not straight to a winner
Apply the stated rule and keep every title that survives it. Two or three viable options is a healthier result than one, because it gives the group a real choice and a fallback if the first pick will not play.
- 6
Commit out loud, then start it
Name the chosen title, confirm it, and press play. A group that agrees implicitly has not finished deciding, and the evening leaks away in a second round of browsing.
Step two, in detail
Which consensus rule to use.
Choose by what you want to prevent, not by which sounds fairest. Every rule below is fair; they simply fail in different directions.
| Rule | Best for | Protects against | How it fails |
|---|---|---|---|
| Unanimous | Two or three people | Anyone ending the night watching something they disliked. | The group is large or tastes barely overlap — you can exhaust the candidates without a single match. |
| Veto-safe | Four to eight with mixed tastes | A strong objection being buried under mild approval from everyone else. | Someone rejects broadly. One person vetoing most candidates stalls the room and needs a conversation, not a rule. |
| Majority | Five or more, or a group on a clock | The decision stalling because one person will not commit. | There is no real quorum. Without one, the fastest voters decide for everybody. |
| Weighted | A host, a birthday, a guest of honour | The occasion. The person the night is for gets more say, openly. | The weighting is not announced. Undisclosed influence reads as rigging once it changes the outcome. |
| One wildcard each | Groups that watch together often | Adventurous picks that would never survive a consensus rule. | Wildcards are unlimited or unspent. Cap it at one per person per night. |
TikFlick implements all five as explicit room settings, shown before anyone votes, because a rule that can change silently mid-room is worse than having no rule at all.
Scale changes everything
Group size changes the answer.
The same rule that feels generous with two people becomes unworkable with eight. As the group grows, tighten the constraints rather than loosening the rule.
- Two people
- Require both yes votes. With two people a majority rule is meaningless and a veto rule is identical to unanimity, so use the strict version and keep the candidate pool wide instead.
- Three or four
- Veto-safe is usually right. Unanimity is still achievable but starts running out of candidates; a plain majority lets a two-against-one outcome happen every week to the same person.
- Five to eight
- Majority with a stated quorum, or veto-safe if the group has one member with narrow taste. Tighten the constraints hard first — a large group needs a smaller candidate pool, not a bigger one.
- Nine or more
- Stop trying to satisfy everyone and pick for the room: a broad certification, a familiar genre, and a runtime that fits. Then use majority with a quorum, and accept that some people are along for the company.
Edge cases
The awkward cases.
These are the situations that break an otherwise sound process, usually because they arrive after the rule has been agreed.
- Someone joins halfway through
- Let them react to the surviving shortlist rather than the whole original list. Re-running every nomination punishes the people who showed up on time, and re-opening the rule mid-decision is how a settled group starts arguing again.
- The group is split between a film and a series
- Decide the format before the title, and treat it as a hard limit, not a preference. A two-hour film and a six-episode series are different commitments, and comparing them candidate-by-candidate never resolves.
- There is a real finish time
- Convert it into a runtime ceiling, subtract the time you will actually spend deciding and pausing, and refuse candidates above it. Do this at the start; a runtime limit introduced after a favourite appears looks like a manoeuvre.
- One person has not seen the earlier films
- Treat prerequisites as an exclusion rather than something to talk around, and say so while nominating. It is a constraint, not a taste, and it is the one thing nobody discovers until the plot stops making sense.
- Nothing survives the rule
- Do not quietly relax the rule — that discredits the whole process. Loosen one named constraint instead, in the open, and run the same reactions again on the wider pool.
Be honest about these
Shortcuts, and when they actually fail.
Not every shortcut is wrong. Two of these are perfectly good under the right conditions — the problem is using them by default.
The host just picks
Fast, and fine occasionally. Used every week it stops being a group decision, and quiet resentment shows up as disengagement rather than complaint.
First suggestion wins
Rewards reaction speed and confidence, nothing else. It is the loudest-voice problem with a rule attached.
Scroll until someone caves
The person who cares least concedes, so the group optimises for whoever is most tired. It also spends the evening on the decision instead of the film.
Take turns choosing by week
Genuinely good for a standing group, and worth keeping. It only fails on the one-off night with people who will not all be there next time.
Run the method, not the argument
A room does steps two to six for you.
Set the limits, choose the rule, share a six-character code. Everyone reacts privately in their own browser, and the room ends when one title is deliberately confirmed. No account, no install.