We Played 10,000 Financial Lives for Each Strategy. Diversification Had the Best Odds Every Time.

We built a financial-life game, then ran 10,000 seeded lives for each of thirteen strategies. The diversified one had the highest win rate in every starting scenario. The concentrated bets kept the biggest jackpots and the fastest deaths. Here is the full win-rate table and the honest caveats.

We built a game where you live an entire financial life, one year at a time, and try to reach financial independence before the clock runs out. Then we did something you cannot do in a real life: we replayed it 10,000 times for each strategy, under identical conditions, and counted who reached the finish.

The result is the reason the game exists. Across thirteen strategies and every starting scenario we tested, the diversified strategy had the highest win rate. It never won every life. It just won more often than anything else, in every setup, while the concentrated bets kept the biggest jackpots and the fastest deaths. This article is the receipt.

What the game is

The game is a financial-life roguelike. You start young with a paycheck, some debt, and a set of expenses, and each turn is a year: you get income, an event might hit (a layoff, a medical bill, a market crisis), and you decide what to buy and how to allocate. Reach financial independence before time runs out and you win. Run out of money and you go bankrupt. Reach the end still short and you time out. It is short, it has no login, and nothing leaves your browser.

We designed it around a rule we refused to break: the diversified strategy has to be the reliable winning line, and speculation has to be exciting but unreliable. A game that quietly teaches “just pick the right stock” would be worse than no game. So before the game could ship, it had to prove that property against its own math.

How the harness works

To test the rule, we wrote scripted players, one per strategy, and let each one play the same 10,000 lives.

Each scripted player changes exactly one thing: its allocation. The diversified player spreads across US stocks, international stocks, and bonds, and keeps an emergency fund. The single-stock player puts everything in one company. The crypto player goes all-in on crypto, the founder player quits the paycheck to build a startup, the cash player stuffs it under the mattress, and so on. Everything else, how they respond to events, whether they pay down debt, the skills they buy, is held identical, so any difference in outcome comes from the allocation and nothing else.

The lives are seeded and paired. Life number 4,281 deals the same market years, the same events, and the same opportunities to every strategy, so the diversified player and the single-stock player face the exact same history and we are comparing choices, not luck. We run 10,000 of these paired lives per strategy, in four different starting scenarios (an average earner, a graduate buried in debt, a late starter, and a low earner), and record how each strategy did.

This is the launch gate. The game is not allowed to ship unless, at 10,000 lives, the diversified strategy has the best win rate and the concentrated bets keep their spectacular tail while losing on odds. Saying that out loud is the honest version of the story: we did not hope diversification would win, we made shipping conditional on it, and then measured.

The win-rate table

Here is what the 10,000 lives returned on the baseline scenario, the average earner. This is every strategy we scripted, sorted by how often it reached financial independence.

Illustrative game outcomes, not advice. From the game's own certification harness: 10,000 paired, seeded lives per strategy on the baseline "average earner" scenario, at the certified engine (N=10,000). Win rate is the share of lives that reached financial independence before timing out or going bankrupt. Not a backtest of historical markets; see the caveats below.

Read the top of that table before the bottom. The three strategies closest to diversification are the ones that are themselves almost diversified: diversify-then-gamble-late (65.6%) and diversify-but-skip-the-emergency-fund (65.1%) trail the full diversified line by less than a point and a half. Diversification’s real margin is not over its near-twins; it is over the concentrated bets. A single stock reached financial independence in a quarter of its lives. Options and founding reached it in fewer than one in thirteen.

And this is not a baseline-only effect. In all four starting scenarios, the debt-loaded graduate, the late starter, and the low earner as well as the average earner, the diversified strategy had the highest win rate of any strategy. It was the single most reliable line no matter where the life started.

The variance story: big wins and fast deaths

If concentration loses on odds, why does anyone reach for it? Because it keeps the jackpot. Losing on average and winning huge occasionally are not contradictions; they are the same fact seen from two ends. Here are the same strategies again, but now with the top-1% outcome (the net worth of the luckiest one in a hundred lives) beside the number that went bankrupt.

Illustrative game outcomes, not advice. Same 10,000 paired seeded lives per strategy, baseline scenario, certified harness (N=10,000). Top-1% net worth is the 99th-percentile final net worth across the 10,000 lives. Read the columns together: the higher the ceiling, the lower the odds and the more lives ended in bankruptcy.

Look down the ceiling column and then down the win-rate column, and watch them move in opposite directions. The diversified strategy has the lowest jackpot in the table and by far the best odds. The founder has the highest jackpot of any strategy in the whole game, about 4.06 million dollars in its luckiest hundredth of lives, and the worst odds, with 9,220 of 10,000 lives ending in bankruptcy. The single-stock and crypto players keep a fatter tail than the diversified player (a top outcome above 2.3 million versus the diversified 1.69 million) and pay for it with a win rate cut in half or worse.

The mechanism is not mysterious. The concentrated players who died mostly died fast, of bankruptcy, when a bad year hit a portfolio with nothing to cushion it. The over-cautious players who died mostly timed out, never blowing up but never reaching the finish either. The diversified player is the one that most often did neither. Its spread across markets and its emergency fund meant a single bad year rarely ended the run, and its exposure to stocks meant it still grew enough to finish. It gives up the jackpot to keep the floor, which is the trade that wins a long game.

The honest caveats

This is where precision matters, because it would be easy to oversell what these numbers are.

This is a game economy, not a market backtest. Our retirement simulator does the opposite thing: it replays real bundled historical data, actual US and international market years, to estimate how a portfolio would have survived. The game does not replay history. It runs a procedurally generated economy, a synthetic world of returns, events, and opportunities, that is calibrated to match the statistical character of the real record. Its US stock returns average around 5 to 6 percent a year in real terms, its international returns land close to US rather than far behind, and its crashes and recoveries carry realistic magnitudes, all checked against the real datasets. But it is a designed system tuned to behave like reality, not a forecast of it, and not a claim about any specific historical decade.

The win rates describe that designed system. They are a property of the game’s economy and its scripted players. They are strong evidence that within a realistic-feeling world, spreading your bets beats concentrating them on odds. They are not a promise about your account.

The strategies are deliberate caricatures. A real diversified investor is not a script, and a real founder does not pay themselves nothing for two years by rule. The scripts isolate one variable each so the comparison is clean, which is exactly why they are extreme.

What survives all of that is the shape, and the shape is the lesson. The reliable winning line was the diversified one, in every scenario, at 10,000 lives each. The exciting lines kept the biggest jackpots and paid for them with the worst odds. That is not a moral, it is a table.

Where to go from here

If the game’s finding lines up with anything, it is the same story the historical record tells. A fair, same-years test of a US-only portfolio against a diversified one found the diversified mix survived more retirements. The reason single big bets are so hard to win with connects to why you cannot reliably time the market. And the danger that makes an early blowup so costly, a bad run of years arriving at the wrong moment, is sequence-of-returns risk, the same force the game’s bankruptcies are built from. The game and the data are two windows on one idea: spreading your bets does not raise the ceiling, it raises the floor, and the floor is what you are standing on when the bad year comes.

Frequently asked