Your APM is a budget

Learner Competitor

The short version

• Actions per minute is not a score. It is an income, and like any income the interesting question is what you spend it on.
• At ninety actions a minute a fifteen minute game gives you about thirteen hundred actions. All of them.
• Spend them so that the last action you put into each category is worth about the same. When one category is obviously returning more, move actions to it.
• The best actions are the ones that stop you having to act later. Those are cheaper than free.
• Actions early are worth more than actions late, for the same reason minerals early are.

The situation

You watch a replay of a player two ranks above you and their APM is not much higher than yours. You had a hundred and ten. They had a hundred and thirty.

And yet they never missed a production cycle, they scouted three times, and they microed a fight you could not have microed. The twenty actions a minute did not buy that. What bought it was that almost none of their actions were wasted, and a good number of them were actions that removed the need for four more.

The budget

Start by making the number concrete, because it is smaller than it feels.

At ninety actions a minute, a fifteen minute game is about thirteen hundred and fifty actions. That is the whole game: every selection, every click, every hotkey, every camera move.

A single production cycle across four gateways, done with hotkeys, is maybe six actions. Done without, by clicking each building, it is more like sixteen. Over a fifteen minute game at one cycle every forty seconds, that difference alone is roughly two hundred and twenty actions, which is a sixth of everything you have.

This is why hotkeys are not a style preference. They are a discount on a fixed budget, and the discount compounds because the actions you save go back into the pot.

Three places to spend

Lesson 10 divides the budget into three categories. Here they are, with what each one returns.

Macro. Production cycles, supply, workers, expansions, upgrades. Characteristics: highly regular, therefore highly compressible with hotkeys and habit, and the returns compound. Almost every player at every level underspends here, and the reason is not that they do not know: it is that macro actions do not feel like they are doing anything at the moment you make them.

Micro. Fights, harassment, dodging, retreating. Characteristics: lumpy. Worth nothing for minutes at a time and then worth everything for eight seconds. Returns are bounded by the size of the fight, as T7's arithmetic shows, and they do not compound.

Information. Scouting, camera movement, looking at the minimap, checking what a unit is doing. Characteristics: cheap, chronically underspent, and the only category whose payoff is other categories spending better. An action that tells you the fight is happening at the ramp saves you the entire fight.

Equalise the margins

Here is the whole allocation rule in one sentence: move actions from the category where the last one did least to the category where the next one would do most.

That sounds obvious written down and almost nobody does it, because the feedback is invisible. So here are the three symptoms that tell you which way to move.

• You are floating money and your fights are fine. Move actions from micro to macro. This is the most common misallocation in the game by a distance.
• You are never surprised, your macro is smooth, and you lose fights you should win. Move actions from macro to micro, and specifically into the first five seconds of engagements.
• You are surprised. Something arrives and you did not know it existed. Move actions to information, which is almost always the cheapest fix available because the current spend is usually near zero.

The working

Four ways to buy back actions

None of these require faster hands.

Hotkey production, not just army. Most players hotkey their army first because fights feel important. Production buildings are the ones you touch every forty seconds for the whole game, and they are the bigger saving by a wide margin.

Rally properly. A correct rally point is one action that replaces one action per unit for the rest of the cycle.

Build so that the map does the work. A wall, a building placed to block a ramp, a turret where a drop always comes: each of these is a permanent substitute for attention. T1 calls this buying back a physical resource with minerals, and it is the best exchange rate in the game for a player whose hands are the limiting factor.

Batch. Doing all of your macro at once, on one camera position, costs fewer actions than doing it in four visits. The regularity is the saving: a habit you always do in the same order is a habit you stop having to think about, and thinking is not free either.

APM is not the target

Worth saying plainly, because the number is displayed and displayed numbers become goals.

Raw actions per minute counts repeated clicks that did nothing. Spam does not make you better and it does not make you faster; it makes the number bigger. A player at a hundred effective actions is meaningfully ahead of a player at a hundred and eighty raw actions, ninety of which were the same click.

What raising your real speed does buy is a bigger budget, and a bigger budget is genuinely worth having. But a bigger budget spent badly loses to a smaller one spent well, which is the only reason this guide exists.

See it in a real game

The lesson 11 lecture has the clearest example, at ►45:00. The lecturer introduces a replay as a case of converting the most APM you can have into work, and shows a player who spent actions early, in dead time, to buy himself a defence that cost almost nothing to use later. That is this guide's second half in one clip.

He then undercut it, characteristically, by pointing out that the opponent rebuilt his army in about two minutes and won the game anyway. The technique was right and it was not sufficient.

Still true today