What Movie Should I Watch Tonight?

The question is badly formed, which is why it takes forty minutes. “What movie should I watch tonight” asks you to name the best film out of everything ever made, from memory, while tired. Nobody can do that. What people can do easily is look at a film and say no.

So invert it. Five films arrive without you having proposed any of them, and your job is only to reject the ones you do not want. Rejecting is fast, it needs no justification, and it does not carry the small dread of having picked the thing that wasted everyone’s evening.

1. Press once, before you discuss anything. Debating what movie I should watch in the abstract is the part that takes forty minutes. Five concrete titles change the conversation from open-ended to a short list.
2. Strike the obvious noes. Usually three go instantly. Open any row you are unsure about — the poster and a two-line synopsis are normally enough, and if they are not, that hesitation is itself an answer.
3. Take the last one standing. If two survive, take the shorter one. If none do, press again; the round you just rejected drops into a strip below, so nothing is actually lost when the new five turn out worse.

The rule that makes this work is committing to the round before you see it. Decide in advance that the evening’s film comes from the next draw, and five titles is a decision. Keep the option of another draw permanently open and it is just a slower version of scrolling, with a button.

Nothing here is a recommendation. Every film matching your filters has exactly the same chance of coming up, which is the point — a suggestion you can argue with is another thing to evaluate, while a coin toss you agreed to abide by is over. That is the whole method for what movie should I watch tonight: commit to the round, draw, strike, press play.

If you would rather start from what everyone else is watching, the Netflix top 10 movies page has this week’s chart — a ranked list rather than a draw, and a different way to end the evening’s debate.

Every version of “what movie should I watch” is really two questions — what do I feel like, and how much evening is left. Start with the second. Of the four filters above, length is the one to set first, and the one most people never touch. Genre is a preference and preferences bend. The clock does not.

Work backwards. Decide when you want to be asleep, subtract the twenty minutes that go on getting settled and looking at your phone one last time, and whatever remains is the real budget. The median film in this library runs 102 minutes, so an ordinary evening film is closer to a two-hour commitment than the runtime suggests.

Under 90 minutes — 2,511 films

The weeknight band, and the most neglected one. Start at half past ten and you are still in bed before midnight. It is also where most of the animation and documentaries live.

90 to 130 minutes — 9,012 films

Two thirds of everything, and the default shape of a film. Leave the filter alone and this is broadly what you get.

Over 130 minutes — 1,630 films

A Friday or a weekend afternoon. Starting one of these on a Tuesday is how a film gets abandoned forty minutes in and never resumed.

Setting length before you draw does something the other filters do not: it removes the argument you would otherwise have after you have already fallen for a title. Nobody wants to be the person who says no to a film everyone just agreed on because it runs to three hours. Rule it out first and that conversation never happens.

The counter under the button shows how many films are in play, so you can see straight away what the constraint costs you. Under 90 minutes and a single genre is a much smaller pool than either filter alone — narrow, draw, and if the round is thin, give back whichever constraint you invented rather than the one the evening actually imposed. Length first is the shortest route from “what film should I watch” to actually watching one.

Alone, “what movie should I watch” is merely slow. Deciding with someone else fails differently, and the reason is not that your tastes differ. It is that proposing costs something and agreeing does not.

Whoever names a film has taken on responsibility for the next two hours. If it drags, that was their call, and both of you will remember it. So both of you stay vague — “I don’t mind”, “whatever you fancy” — which sounds accommodating and is actually two people refusing to hold the bag. Half an hour later there is still no film.

A random draw removes the proposer. Nobody chose those five films, so nobody has to defend them, and vetoing carries no cost at all — you are not turning down your partner’s idea, you are turning down the machine’s. That single change is what makes the shortlist move when the conversation would not.

In practice: draw once, agree out loud that tonight’s film comes from this round, then take turns striking one title each until one is left. Vetoes need no reason. If all five go, draw again — but agree to the same terms first, because a round nobody committed to is just five more things to have opinions about.

It works on a room as well as on a couple, and better than a vote does. Voting rewards the least objectionable film, which is how you end up with something nobody actively wanted. Striking rewards the opposite: whatever survives is the one film in five that nobody in the room could find a reason to refuse. It is the couple’s answer to what movie should I watch tonight, and neither of you had to win it.