A Decision You Never See the End Of Teaches You Nothing

Getting better at anything takes three things: making a decision, seeing what it did, and connecting the two. Without all three, repetition produces nothing but more repetition.

A chess player improves quickly because a move is paid for within thirty seconds. A cook learns the same evening whether the plate came back. Whether a hiring decision was right, on the other hand, takes a year. Whether a pricing change worked takes two. What a career choice was worth might take ten.

When feedback arrives late enough, it stops being learning and becomes remembering — and we usually remember it wrong.

What the delay does

When the result is slow, too much gets in the way. The market moves, the team changes, you change. By the time a number lands you cannot say which decision produced it. What usually happens instead is this: a good outcome gets credited to our judgement and a bad one to circumstances. Neither teaches anything.

This is not a personal failing. It is the same problem in every field with a long feedback loop, which is why pilots fly simulators, surgeons work on models, and firefighters drill in buildings that are not on fire. The point is not to imitate reality. The point is to turn feedback that takes months into feedback that takes minutes.

Compressed time

A business simulation plays ten years of a career in twenty minutes. That is not living ten years in twenty minutes — the claim would be silly. What it does is narrower and more useful: it keeps the decision and its result close enough together that the link between them is still visible.

Raise your spending by one notch and the cash-flow line moves on the same screen. Break an investment early and watch the compounding you gave up. Hire two people and notice that the payroll starts tomorrow while their output arrives two weeks later. None of this is new information; all of it is already known. The difference is between knowing something and having watched it happen.

What it does not teach

To be straight about it: a simulation will not teach you to manage people. Not the hard conversation, not delivering bad news to someone’s face, not how a partnership comes apart. None of that is modelled, because none of it can be.

What it teaches is more modest: how the numbers are wired to each other. Which three things twitch when you move one. What chases what. That is not the whole of running anything, but it is the skeleton underneath — and learning the rest without having seen the skeleton is far more expensive.

Where to start

Wherever the subject already interests you. If it is how a company runs out of money, try the startup or the store; if it is your own budget, personal finance; if you lean macro, economic policy. All of them run in the browser and none of them ask you to sign up.

Knowing how long things take

The least discussed part of experience is a sense of timing: how long a piece of work really takes, when an investment starts to return, when a team settles into a new arrangement. None of it comes from a book, because a book gives you the average, and what you need is to know which way your own estimates are wrong.

Almost everyone is wrong in the same direction: we assume things will be faster. In a simulation you notice that bias within a few runs, because your forecast and the outcome sit next to each other. What you called three months takes five — repeatedly, and even once you have started planning around it.

How you know you learned something

The sign that you learned something is not a better result; results owe something to luck. The sign is this: on the next decision you already know which number to look at. In the first run you are discovering what the screens contain. By the third you know which two figures you will check before the month ends.

That is the part that transfers to real work. What a simulation gives you is not the right answer but the right question. Knowing which screen to look at outlasts knowing what was on it.