Why dying is often hard, but sometimes easy

You are sitting across from your doctor. You came in for a routine follow up appointment that is supposed to end with "everything looks fine, see you next year." But it does not end that way. The doctor uses words you only half hear. Terminal. Months, possibly. We can talk about what to do next.

You will replay this moment for the rest of your life, however long that is. The fluorescent light over the desk. A child laughing somewhere down the hallway. The faint smell of disinfectant. A text message from a friend you forgot to answer this morning. None of it has changed in the last minutes. You walk out of the office. The same faces sitting in the waiting room. The same barista at the corner. The same streets on your way home.

Nothing in the world has changed, but everything in your world has.

The number you never wrote down

Most people would say the new information of that moment is that you found out that you will die. It is not. You already knew you would die. You have known since you were five or six. The developmental research is fairly settled (Speece & Brent, 1984). By the time you can ride a bike, you have absorbed the fact that everything alive eventually stops. Death is not new information. You have been living with it for thirty years or fifty or seventy. ​And yes, knowing death as an abstract category is not the same as confronting one's own finitude, but for this argument it is beside the point.

The new information is the time horizon. Until that appointment, you were maybe operating with an implicit horizon of decades. Probably never have written it down anywhere or made it explicit in some other form. But every decision you made was being weighed against that hidden number. The job you took because it would compound. The skills you were building. The savings you were accumulating. The trip you were going to take "next year." All of that planning rests on a horizon you never named but always assumed. The diagnosis is not the introduction of death into the model, but the rewriting of the time variable.

Expected against actual

Two markers on one timeline. Expected is the horizon you assumed and lived toward. Actual is the one you turned out to have. The hatched span between them is the planning that no longer fits.

Now two things have to happen at once. The first is grief. You have to mourn the futures that will not happen. The version of you who would have grown old, seen the graduation of your children, held grandchildren. Worse, you have to grieve the time you spent optimizing for a horizon that turned out to be wrong. Every sacrifice you made for the long term was made in a world that no longer exists. The regret of that misallocation is its own kind of pain, separate from the fear of dying.

The second is recalibration. The horizon is now different, therefore the strategy has to change. The decisions you make from this point forward have to be optimized for a new and much shorter timeframe. You cannot do this slowly. The whole point of having less time is that you have less time. Somewhere in the back of your mind a clock has started to tick louder.

The geometry of every decision

The terminal illness example is the most extreme version of a problem that quietly governs every decision you make.

The Strategy Horizon Problem

The horizon you assume shapes the strategy you choose. The strategy you choose shapes the horizon you actually get.

Every decision you make is calibrated against a time horizon, whether you have named that horizon or not. And the horizon is not one number. It is two. The expected horizon is what you think the timeframe is. The actual horizon is what the timeframe turns out to be and they are almost never the same.

Your expectation about how long the game lasts informs what strategy you choose. If you think you have forty years, you compound. If you think you have three, you concentrate. The expected horizon is the input. The strategy is the output. But the strategy you pick does not just sit downstream of the horizon, as the choices you make can (but don't have to) change how long the game actually lasts. A reckless strategy can shorten it. A protective strategy can extend it. Some strategies run a small chance of cutting the horizon to zeroNon-ErgodicityThe Room with the Revolver and other Non-Ergodic SystemsRead the essay↗.

So the expected horizon shapes the strategy, the strategy shapes the actual horizon and the actual horizon is almost guaranteed to be different from the expected one.

The founder, the retailer, the career

A founder raises a Series A and quietly assumes the next round is eighteen months out. The strategy that follows is calibrated to that horizon. Hire ahead of revenue. Spend on growth. Push the gross margin question to next year. Then the market turns. The next round does not come. The runway that was supposed to last eighteen months runs out at fourteen. The strategy was correct for the expected horizon, although it got the company killed under the actual one.

The reverse plays out just as often. A retailer assumes the shift to e-commerce is decades away and keeps the long-arc strategy. Refurbish the stores. Deepen the supplier relationships. Train the floor staff. The expected horizon was thirty years of slow adaptation. The actual horizon was eight. The patient compounding work that would have paid off across thirty years became dead weight at year nine. Everything that was a strength inside the expected horizon turned into a liability under the actual one. Christensen (1997) gave this its canonical form. Incumbents do not fail because they manage badly. They fail because they manage correctly for a horizon that disruption has already shortened.

You see it in careers. Someone bets a decade on a slow-compounding craft, expecting twenty more years of upside. The industry collapses in five. The compounding never gets to do its work. Or the opposite. Someone abandons a craft early because they expect a short window of relevance. They watch the people who stayed make a thirty-year arc out of the same starting position.

In every one of these cases, the people involved optimized correctly for their expected horizon - the failure point was not reasoning. You always have to pick a strategy without knowing the horizon and the outcome will be largely determined by the inescapable structure of making decisions when the horizon is not knowable in advance.

A life that doesn't need to know

Every decision is trying to optimize for a specific time horizon. So the underlying problem is the uncertainty about the time horizon.

One way to deal with that is to stop betting on a single horizon and deliberately play two at once. Do something that pays this week and something that pays when you are eighty, as Luca Dellanna put it on my podcast Entertaining Ideas.

Two horizons at once

One payoff cashes this week, the other only at the far end, with nothing staked on the single horizon in between.

The walk taken, the meal with people you love, the call made, the emotional risk that returns its value the moment you take it - none of that can be revoked by a horizon that turns out short. In parallel the slow things keep running. The savings compounding, the craft deepening, the children being raised, the book being written and the friendships built over decades. If the long horizon arrives, those pay. If it does not, you were already paid some days along the way. ​The technical term for such an approach is a bimodal strategy.

There is a second class of decisions whose value does not depend on how long you have. Most decisions are horizon-specific, as they pay off only if the timeframe cooperates. A horizon-independent decision as I call them is one whose worth survives being wrong about that number. It is good if you have three years and good if you have forty.

Worth that survives the number

The horizon-independent one holds its worth at three years and at forty.

Such decisions are great independent of the time horizon they have to play out or your time horizon to experience the outcome of such made decisions. Planting a tree can be meaningful the day before your execution or if you have 60 years to live. You stop asking how long the game lasts and start asking whether the move is worth making in any version of it. ​There is much more to cover. Future essay will dive into more detail.

When dying comes easy

This is not a recipe for ease at the doctor's office. There is no recipe. Also not a recipe on how to live a life - there is definitely no recipe for that. What this essay tries to illustrate is why some lives feel finished even when they end early, while others, ended on "schedule", feel cheated.

None of us gets to know the actual time horizon of our lives. So... Dying is hard when you optimized for one horizon and got another. It is easier when you did not.