On one life and the decisions you cannot take back
24 July 2026
Imagine a bet that pays. A coin toss, heads adds half your wealth, tails takes forty percent of it away. Work out the ordinary average and it comes out positive, because fifty up beats forty down, so a reasonable person should play, and keep playing. And yet someone who plays without end drifts toward zero. The reason is quiet and countable. Loss and gain work through multiplication, not addition, and loss multiplies harder. When you lose forty percent, to make it back you no longer need forty, but sixty seven up. A win and a loss back to back do not return you to where you started, they leave you a touch lower, and so on around and around.
A group of a thousand players sees this differently, because among them there is always someone who caught a long run of heads, and his result pulls the average of the whole thousand up until it looks beautiful. Everyone is glad and no one counts the ones who lost everything. Except you are not a thousand. You are one. Averaging over a thousand copies, some of which go bankrupt while the average still rises, is something entirely different from walking a single path, yours, toss by toss. And no one can guarantee you that you are not that copy, the one that threw tails. There is a single number for this. The true rate for a single path is given not by the ordinary average of the outcomes, but by the average of their logarithms, because the logarithm turns multiplication into addition. For this coin it comes out to about five hundredths per toss downward, even though the ordinary average gives five hundredths upward. The first number is your fate. The second is the fate of the group.
Someone will say that by multiplying you never touch zero, you only go lower and lower, a hundred, seven, a half, and you still have something to play with. In theory, yes. But in the real world there is always a floor. A minimum stake, the smallest coin, and outside the casino a level below which you do not pay the rent, you lose your health, or the place from which you could still have bounced back. Sooner or later the drift pushes you through that floor, and erosion turns into a wall. On a single path, where such a fall means the end of the game, avoiding ruin stops being cowardice and becomes the arithmetic of survival.
This changes what you actually need to fear. It is not about how large the loss is. The risk of ruin is the product of three things. How large the loss is, how likely it is, and how much it costs to come back from it. A product, not a sum, because it is enough for one of these factors to be zero and the whole risk disappears. Something enormous and final, but impossible, threatens nothing. Something final and almost certain, but at a penny stake, threatens nothing either. Only all three at once make ruin. And we measure one thing, and not even its size, only its vividness. The alarm in the body answers to what is suggestive and loud, to an image that is easy to picture and that happens in front of people. That is why rejection and disgrace, though small, shout the loudest, while the quiet and final things pass without an echo.
Let us stop at this third factor, because it is the hardest. Irreversibility is not a zero or a one, it is the cost of return. For almost every decision that cost is finite, though we usually inflate it under the pull of emotion. A lost salary is large, and reversible, you will earn again. That is a puddle you can step into a hundred times and climb out of a hundred times, each time wiser. But there is a narrow class of things where the cost of return is not high, it stops existing. You killed someone, that person no longer exists. In anger you burned the only keepsake, there is nothing left to recreate. Here you do not ask how much it costs to come back, because there is nothing to count from. This is the wall, the edge of the scale, and only here is caution absolute. All the rest is a matter of price, not finality.
That cost is not fixed either, it changes over time, and in which direction depends on the matter. Coming back to medicine after ten years away gets harder and harder, because the skills and the papers fade, the cost rises. But after a sharp fight with someone close, time plays in your favor, because the other side forgets, the anger settles, no one is waiting for an apology anymore, and the return gets cheaper with every month. So the same decision can be a puddle or a wall, depending on when you think about it.
And the wall rarely stands in one place. More often you reach it piece by piece. Each single move is a puddle, a small debt, one week burned, one evening of the habit, out of each one alone you would come out with dry feet. But the cost of return is counted from the sum, not from the single step, and a sum of puddles can grow into a reservoir you no longer climb out of. It is the same coin, only spread over years. That is why with things that accumulate you do not only ask whether this one move will wipe me out, but whether their sum is drifting me toward the wall, and you set yourself a threshold in advance, while still cool. A threshold with a number, not with a feeling. Not "I will get out when I feel I am drowning," because that is never felt in time, but "if I drop below three months of reserve, I stop, however good the opportunity looks." A number cannot be talked down in the evening, a feeling always can.
There remains an honest question, how to recognize the wall before you walk into it, if instinct does not feel it. The answer is uncomfortable. There is never certainty. You will not get past the fact that you decide with incomplete information and limited abilities. You do the same thing as a player at the table who cannot see the others' cards. You put in the effort, you estimate the probabilities as best you can, you accept that a black swan may come out anyway, and you make the decision at the peak of your rationality, on the data you have. This is not a promise that you will recognize every wall. It is the only method there is. Whoever promises more is lying.
Not every move with no return is a mistake. Sometimes you cut off your own retreat on purpose, to change your own payoff matrix. Imagine a commander who orders his own ships burned, so that his men no longer count on going back, only on the conquest. As long as you have bridges behind you, part of you will want to cross back over them. Dropping out of university, leaving a secure job to go out on your own, is the same thing, deliberately taking moves away from yourself so as to focus all your force on one. The difference between that and a stupid move with no return is a single one. The stupid one throws you out of the game forever. The wise one you survive, and it shifts the board in your favor. You go through a one-way door when you survive the worst possible outcome on the other side, and the win on the upside is worth it.
With what is final, will alone is not enough. It is weak exactly when it is needed most, in the heated moment, under pressure, after alcohol, in anger, and with a move with no return a single weak moment is enough. That is why you do not fight such moves with will each time, you take them out of your own hands in advance, with a simple rule set while sober. I do not get behind the wheel if I have been drinking, full stop, no negotiating in the middle of the party over whether I will make it home this time. And such a rule is good not because you made it home this time, but because it was right before you tossed the coin. A good decision can be rewarded with a bad outcome, and a stupid one with a lucky outcome, and confusing the two is the simplest way to draw from one successful drunk drive the conclusion that you can. Will is left for what is reversible, where a mistake costs nothing but the lesson.
Life is not a game about who collects the most points, but about not touching the wall you do not come back from, and recognizing the few you should cross on purpose. And because you live once, one such move, in the bad or the good direction, weighs more than all the ordinary ones together.