The Average That Lies to Everyone Who Lives It

Picture two ways of describing the very same gamble, both of which sound like “the average outcome,” and both of which the word average was invented to name. In the first, a thousand people each play the game once, all at the same moment, and we average across the crowd. In the second, one person plays the game a thousand times in a row, and we watch how their money grows along the way. Common sense — and every statistics course that didn’t warn you otherwise — quietly promises these two numbers are the same.

For the tame risks your intuition was trained on, they are. And for the risks that actually decide your life — the ones that compound, where this round’s winnings become next round’s stake — they are not. They can point in opposite directions. One can say “get rich”; the other can say “go broke”; and they can be describing the identical bet. Learning to tell which average you are actually living is the whole course, and it starts with a coin.

A “sure thing” that ruins almost everyone

Here is the coin. It is fair — a genuine 50/50. On heads, your wealth grows by 50% (multiply by 1.5). On tails, your wealth shrinks by 40% (multiply by 0.6). Nothing is hidden; the odds are honest; the edge is real.

Compute its expected value the way you were taught. Half the time you end a round at 1.5× your money, half the time at 0.6×, so the average multiplier per round is:

$\tfrac{1}{2}(1.5) + \tfrac{1}{2}(0.6) = 1.05$

A flat, delicious +5% per round. Expected value says: play this forever and your money grows without bound. It is the kind of edge a casino would burn down its own building to get on the other side of.

Now live the sequence instead of averaging the crowd. You don’t get a thousand parallel copies of yourself; you get one run, and each round multiplies what the last round left you. Watch what multiplication does. A heads then a tails leaves you at $1.5 \times 0.6 = 0.9$ — you are down 10% after one win and one loss. Order doesn’t save you: $0.6 \times 1.5 = 0.9$ too. Every balanced pair of flips quietly shaves off a tenth of everything you have. Play long enough and, with near-certainty, you spiral toward zero — while the crowd’s average sails to the moon on the backs of a lucky few.

The two averages, named

The contradiction dissolves the moment you admit there are two different averages hiding under one word.

The +5% is the ensemble average — honest about the thousand-person crowd, where a few explosive winners haul the mean upward even as most players sink. The time average is what happens to you, compounding one sequence, and here it is about −5% per round (the geometric mean of 1.5 and 0.6 is $\sqrt{0.9} \approx 0.949$). Same coin. One average promises wealth; the other delivers ruin. Neither is wrong — they are answers to different questions, and catastrophe is what happens when you take the answer to the crowd’s question as the answer to yours.

Watch the two averages tear apart

Here is the coin made draggable. Set the gamble — the win move, the loss move, the odds, how many rounds — and run it across 400 parallel players at once. One line is the ensemble average: the mean wealth of the whole crowd. The other is the typical player: the median person, living the sequence over time. Watch them split.

With the default coin the ensemble average rockets upward while the typical player is quietly ground toward zero — the +5% is real, but it lives in a vanishing sliver of astronomically lucky paths, and almost everyone else busts. Then reach for the one knob that changes the story: drag the bet fraction down. Stake a sliver of your wealth each round instead of all of it, and the typical player’s path bends from a slide into ruin into a genuine climb. That is a preview of Kelly betting — and of the whole survival logic this course is built to give you.

Two things to carry forward. First, the ensemble average is not lying — it is answering a different question than the one you asked. It faithfully reports the fate of the crowd; you were the one who assumed the crowd’s fate was yours. Second, nothing here is rigged. The coin is fair, the edge is genuine, the arithmetic is clean. The ruin comes entirely from the difference between averaging across parallel worlds you will never inhabit and living the one path you actually walk.

Why this earns a place in the latticework

Because “on average, it works out” is one of the most confident and most ruinous sentences in all of decision-making — and this is the model that tells you precisely when it’s false. Expected value is a genuinely great tool, computed by someone who quietly assumes they survive every round to collect the average. The instant a loss can be irreversible — an absorbing barrier at zero, from which nothing rebounds — that assumption shatters, and the glittering positive average becomes a mirage that only the early-lucky ever touch. Blown-up traders, bankrupt “high-EV” gamblers, over-leveraged funds, the person who bet everything on a near-sure thing: all invisible to someone holding only the ensemble average, all obvious to someone who has internalised the time average.

This is an expert-tier synthesis, and it stands on three models you’ve already built:

• From fat tails, the shape of the danger: rare, enormous outcomes dominate a mean, so one unrepeatable outlier can prop up an ensemble average no individual will ever reach.
• From compounding, the engine: growth is multiplicative, so path and order matter, one bad factor poisons the whole product, and a zero anywhere is forever.
• From asymmetry & optionality, the survival logic: an irreversible, capped downside isn’t merely a big loss — it’s an absorbing barrier that ends the game, and no upside pays out after it.

The map of the course

Six teaching lessons build the model from its core out to its limits, then one exam locks it in:

1. Ergodic vs Non-Ergodic — the central distinction made plain: when the time average equals the ensemble average (trust expected value) and when it doesn’t (don’t).
2. The Flagship Coin Flip — the +50%/−40% bet worked out in full, and why a vanishing fraction of lucky paths drags the mean up while almost everyone busts.
3. Additive vs Multiplicative — the fork that decides everything: why averaging is safe for a small, bounded side-bet and fatal for your whole bankroll.
4. The Kelly Fix — optimise the average you live: expected log-return, the geometric mean, and betting to survive.
5. Ruin & the Real World — absorbing barriers, gambler’s ruin, position sizing, and insurance as buying back ergodicity.
6. Where the Model Lies — not everything is non-ergodic, it’s about the dynamics not a mood of doom, and it isn’t the same as plain risk-aversion.

Then a Final Exam — graded, one question at a time, one-way: once you answer, it locks. No back button, 70% to pass — fittingly irreversible, like the barrier at zero the whole course is about.

How to use this course

One habit does most of the work: before trusting any average, ask whether you are the crowd or the path. If outcomes merely add up and no single step can wipe you out, the two averages agree and expected value is your friend. If they compound and a step can be ruinous, stop — the ensemble average is quietly computing someone else’s future. Keep returning to the engine above; drag the bet fraction across the point where the typical path flips from ruin to growth until that split feels like a fact about the world, not a technicality.

Next up: lesson 1, Ergodic vs Non-Ergodic — the one distinction that decides whether the average you were handed maps your future or somebody else’s.