Why
Acts are not a template. They are an allocation, and allocating is the decision — the shape of a film is mostly the answer to where does the runtime go.
Each act does a different job, and each job has a different relationship to time:
- Act one buys legibility. The audience needs the world, the person, and the question. Its cost is fixed by how strange your world is, not by convention.
- Act two buys weight. Consequence accumulates here, and accumulation genuinely needs duration — you cannot compress the feeling of things getting worse.
- Act three buys resolution. It is almost always shorter than writers want, because after the climax the audience is finished and the film is still talking.
The conventional 25/50/25 is not a rule; it is the median of a distribution. What matters is that the proportion is chosen and that the choice serves the specific film. A long first act is correct when the world is unfamiliar. A long third act is correct almost nowhere, and is usually a sign that the climax landed too early.
The trap is that proportion is decided by accident. A writer writes until the first act feels done, then writes until the second feels done, and the resulting division is a record of their attention rather than a design. The film that results has no pacing problem you can point to and a pacing problem you can feel.
Two consequences worth holding:
Act length sets scene budget. A twelve-minute short with a 25% first act has three minutes, which is three or four scenes. Deciding proportion decides how many scenes you can afford, before you write any.
Moving the act break moves the meaning. The break is where the audience stops asking what is this and starts asking how does it end. Placing it later makes the film feel more like a study; earlier, more like a machine.
How
- Write the runtime target in minutes. Everything below is arithmetic against it, and the arithmetic is the point.
- Choose the proportion explicitly and write one sentence of reason. "25/50/25 because the world is ordinary and the pressure is the interest" is a decision; leaving it unstated is not.
- Convert each act to minutes, then to a scene count at your format's average scene length. This number is a budget, and you will want to exceed it.
- Place the act-one break at the moment the dramatic question becomes unavoidable — not at the inciting incident, which is usually earlier than writers think.
- Place the act-two break at the lowest point of the charge curve. If the lowest point is not near the second break, either the proportion or the curve is wrong.
- Check the third act against the climax: count the minutes after the climax lands. More than 12% of runtime and you are writing a second ending.
- Re-run the arithmetic after the beat board exists. If the beats do not fit the budget, cut beats — do not quietly extend the act, which is how a twelve-minute short becomes nineteen.
Notes for AI film
Proportion is where generated film breaks from shot film most sharply, because the third act is the most expensive part to produce and the least forgiving of drift. Every minute after the climax is a minute of shots that must hold continuity for a payoff the audience has already received.
In short and vertical formats, the useful reframing is that act one is often a single shot. The proportion still exists — it is 8/70/22 rather than 25/50/25 — and stating it still prevents the most common failure, which is a piece that spends half its runtime establishing and then rushes the turn.
For the 15- and 30-second commercial the three-act reading is genuinely unhelpful; use the problem/solution/proof shape those formats seed instead.