The chances that change a trajectory do not arrive when you are ready. They arrive when they arrive, stay open for a moment, and close on whoever was still getting ready.
I grew up thinking every hard choice was a gamble. It sounded cynical then. I still think it, and now it just sounds like a description: the information you would need to be certain does not exist, and will not exist before the decision has to be made.
I can date my own window. Two calls in August 2022, and a yes to a software job after 3D and architecture work that began in 2014. What made the yes possible had been built from 2016, when I started writing code to drive geometry in Rhino and Grasshopper. I was not preparing for anything. I was trying to generate forms faster.
Plan for the worst, hope for the best
Two separate instructions, working on different parts of the problem, and most people run only one.
Plan for the worst is about the downside. Assume the project fails, the market turns. Build so that outcome is survivable: keep reserves, take no position you cannot hold through a bad year.
Hope for the best is about the upside, and it is the half people quietly drop because it sounds unserious. Plan only for the worst and you build a position optimised for survival and incapable of capturing anything. You get through the bad year in perfect shape and have nothing in place when the good one arrives.
Holding both means structuring so a bad outcome does not remove you, while keeping the capacity to move fast when a good one appears. Those are compatible. They are rarely held at once, because each is a complete philosophy on its own.
Why the window is narrow
An opportunity is usually a temporary mismatch, a gap between what exists and what is possible, and it stays open only until enough people notice. The gap closes. The only question is whether you were positioned to move while it was open.
So the useful preparation is never about the specific opportunity. You cannot prepare for a thing you have not seen. You accumulate the general capacity to act.
The grand version is a trajectory changing in a week. The ordinary version is three things built on days when nothing is happening:
- Portable skill – Learn the part that survives a change of context, because a skill held as an identity resets to zero when the context changes and a skill held as a capability does not.
- Bought months – Keep enough money to fund a stretch of reduced income, because most windows demand one and that alone disqualifies most people.
- Returned message – Do work that makes someone answer when you write, because a reputation is the slowest of the three to build and the one the window actually tests.
The fourth thing is harder to list, because it decides whether you see the window at all: understanding deep enough to recognise the gap while looking straight at it. Gaps are hardest to see from inside. Three generations of my family each worked under a constraint they never named, and each one only became visible after it dissolved.
In my case the ordinary days were the ones between 2016 and 2022. Tens of online courses, then a bootcamp that turned out to be easy because of them, all alongside 3D work that paid unevenly: some months nothing, some months over four thousand dollars. None of it looked like a plan. It was readiness, and I could not have told you what for.
When it opens, size the move
Readiness gets you to the window; sizing decides whether you survive walking through it. I think the two get confused because from outside, readiness and recklessness look the same.
Gambling ruins people through two mechanisms, and the losses themselves are neither. A loss is a normal outcome of any bet at any skill level. The damage comes from bets sized so that losing removes you from the game, and from bets placed without information that was available and not gathered, because gathering it was slower than acting.
Everything useful follows from those two, in the order I apply it.
Never bet what you cannot survive losing. Surviving keeps you in a game where the good windows are rare and unscheduled. Being unable to act when one appears is the real cost of a ruinous loss, paid long after the loss itself.
Keep bets small until you have a clear lean. If I could give someone one sentence, this is it. A clear lean is the signal that you have absorbed enough of the situation for your judgment to have something to work with. Before it arrives you are guessing with a larger stake.
Buy information before you buy the position. Most big decisions have a cheap partial version. Talk to people who did it, and to one who regretted it. Do the thing for a month at low commitment. Some things you can only learn by starting, which is an argument for starting small rather than skipping the learning.
Then move properly. Once the lean is clear and the information is in, hesitating becomes its own error. Windows close, and half-committing to something that needed commitment collects the downside of the bet without the upside.
Prefer reversible. A decision you can unwind is worth more than an equivalent one you cannot, and people under-price this. When two options look close, take the one you can exit.
I never bet the income on the code. It ran alongside the 3D work until a salary replaced it, and the last paid 3D job came in 2023, after the software job had started. That overlap was the whole of my sizing.
What “survivable” actually means
The sizing rule is easy to agree with and hard to apply, because the question gets answered optimistically by default. Three questions make it concrete. If this goes badly, how long until you can make another bet this size: years means it was too big, next quarter is the right region. What are you assuming stays true for the downside to be survivable: continued income, a relationship holding, your health, each uncertain on its own and dangerous stacked. And would you still be able to say yes to something good next year, because the cost of a bad bet is the opportunity you could not take afterwards, and you never see that one to grieve it.
Reversibility deserves the same honesty, because it is claimed more often than it is true. A decision is genuinely reversible when you can unwind it without permission from someone else, without a cost that scales with how long you waited, and without having burned something that does not regrow: a relationship, a reputation, a specific window.
Never hold back effort you could have spent
The cost of effort you did not spend is invisible, and therefore systematically underestimated. You will never see the version where you tried harder. So the mind fills the gap with a comfortable assumption that it would have turned out much the same. Sometimes it would. Sometimes it would not, and you cannot know which, so the assumption does no work except making you feel better.
Given that asymmetry, and given that the windows are narrow and unscheduled, the reasonable policy is to spend the effort you genuinely have rather than saving it for an occasion that does not announce itself.
The limit matters, because the advice is dangerous without it. Effort spent by damaging your health or your relationships is not saved for later either. It is spent, on something with a negative return, and the bill arrives at the worst possible moment. Spend what is actually there. Do not manufacture it out of things you cannot afford to lose.
And spend it unevenly. Most of the return comes from a small fraction of what you do, so the skill is sequencing work so that progress cannot be undone, which is how you find the fraction instead of believing it exists somewhere.
Good decision, good outcome
A good decision and a good outcome are different things, and confusing them is how people learn the wrong lesson from experience. You can bet well and lose. You can bet badly and win, which is worse, because it teaches you the process was sound and you will run it again with a bigger stake.
So judge the process. Did you gather what was gettable, size it so a loss was survivable, move once you had conviction? Yes and it went badly is variance. No and it went well is luck, and repeating it will eventually be expensive. Outcomes are loud and processes are quiet, and the only fix I have found is writing down what I expected before the outcome arrives, which is also the difference between having a considered position and feeling like you have one.
Waiting is a bet too. Information has a shelf life, and past a point the delay is itself a position, one that assumes the option stays open while you deliberate.
The part that is uncomfortable
You can do all of it and have no window open. Preparation raises the probability of capturing one. It creates nothing, and anyone promising otherwise is selling something.
I find that easier to hold than the alternatives. The version where effort reliably produces outcomes is false, and believing it makes every bad outcome a personal failure. The version where nothing matters is false too, and believing it removes the only lever there is.
Years on, the code I wrote in 2016 to move geometry looks like the first line of a plan. It was not a plan. It was an ordinary day, and there were a few thousand of them, and one of them turned out to be the one the window was waiting for.
You cannot schedule the window; you can only be someone who fits through it. Which of your ordinary days this month is building something you cannot yet name?