Ed Thorp: The Mathematician Who Beat Blackjack, the Market, and the Problem of How Much to Bet

Most trading legends are known for a call — the big short, the currency trade, the bubble they saw coming. Ed Thorp is known for something rarer and more useful: he is the person who took the vaguest, most-guessed-at decision in all of trading — how much to bet — and turned it into mathematics. Before he ever touched a stock, he beat the casinos. Then he beat the market for two decades without a single losing year. And the thread connecting the blackjack table to the trading desk is the single idea every trader most needs and most neglects: sizing.

Key takeaways

• Thorp proved blackjack was beatable (Beat the Dealer, 1962), then priced options before Black-Scholes (Beat the Market, 1967), then ran a fund for ~two decades with no losing year.
• His enduring gift to traders isn’t a strategy — it’s the answer to the question almost everyone sizes by feel: how much should I bet?
• The Kelly criterion ties bet size to edge: bet more when the edge is bigger, less when smaller, nothing when you have none — to maximize long-run growth without risking ruin.
• The lesson beneath the math: an edge is necessary but not sufficient. Over-bet a real edge and a bad streak still bankrupts you, because ruin is permanent.

The man who broke blackjack

Thorp was a mathematics professor, not a gambler, when he became convinced that blackjack — unlike roulette — had a memory: the cards already played changed the odds on the cards to come. Using an early computer, he proved it, and in 1962 published Beat the Dealer, showing that a player who tracked the ratio of high to low cards remaining could tilt the odds in their own favor. It is one of the few books that has literally changed how a game is played; casinos altered their rules in response.

But the part that matters for traders is what Thorp realized after he had the edge. Knowing the odds were in his favor was not enough. If he bet too little when the count was good, he’d barely profit; if he bet too much, a normal run of bad luck could wipe him out before the edge paid off. He needed a rule for how much to stake as a function of how large his edge was. He found it in an obscure 1956 paper by a Bell Labs physicist named John Kelly.

The Kelly criterion: the math of how much to bet

Kelly’s insight was that to maximize the long-run growth of a bankroll, you should size each bet in proportion to your edge. Bet a fraction of your capital that scales with how favorable the odds are — aggressively when the edge is large, tentatively when it’s thin, and nothing at all when you have no edge. Bet more than Kelly says and you don’t grow faster; you grow slower and dramatically raise your risk of ruin. Bet less and you leave growth on the table but stay safe.

The deep lesson buried in the formula is about ruin. Kelly betting never risks the whole bankroll on any wager, because it always sizes as a fraction — which encodes the most important truth in all of risk-taking: you must never bet so much that a losing streak ends the game, because a permanent loss can’t be compounded back. This is the mathematical version of the survival-first principle we laid out in how successful traders manage risk. Thorp used it to size his blackjack bets: big when the count was rich, small when it wasn’t. Then he pointed the same machinery at Wall Street.

Beating the market before the tools existed

In 1967, with Sheen Kassouf, Thorp published Beat the Market, which laid out a systematic way to profit from mispriced warrants and convertible securities by hedging them against the underlying stock — an early, working form of the delta-hedging logic that the Black-Scholes-Merton model would formalize and win a Nobel Prize for years later. Thorp was, in effect, pricing and trading options before the world had the equation for it.

He turned the theory into a fund. Princeton-Newport Partners, which he ran from 1969, compounded at roughly 19–20% a year gross (around 15% net of fees) for about two decades — and, remarkably, did it with almost no volatility: by many accounts the fund had 227 winning months against only 3 losing ones, none worse than about 1%, and no losing year. A later fund extended the record into the 2000s. This is not a lucky streak that a few good calls can explain. It is the signature of an edge sized correctly, over and over, for a very long time — exactly what Kelly promises and almost nobody has the discipline to deliver.

Finding an edge and sizing it are two different skills. Thorp’s genius was seeing, before almost anyone, that the second one is where the money — and the ruin — actually lives.

The honest caveat: full Kelly is brutal

A responsible profile has to add the warning Thorp himself gives. The full Kelly bet, while growth-optimal in theory, produces gut-wrenching swings — drawdowns that most humans cannot psychologically tolerate, and that assume you know your edge precisely, which in real markets you never do. Overestimate your edge and full Kelly over-bets, tipping from optimal toward ruinous. This is why Thorp and virtually every serious practitioner use fractional Kelly — a half or a quarter of the formula’s amount — which sacrifices a little theoretical growth for a large reduction in volatility and a big margin of safety against overestimating your own edge. The humility is the point: bet less than the math says, because you are less certain than you feel.

Why Thorp belongs on a behavioral blog

Thorp is, at first glance, the least behavioral trader imaginable — a pure quant who removed emotion by replacing it with equations. But that is exactly why he belongs here. His whole career is a proof that the decision most traders make by feeling — how much to put on this trade — is the decision that most determines whether they survive, and that it can and should be made by rule instead. The retail trader who sizes up because they’re “sure,” who bets big on conviction and small when bored, is making the Kelly decision every day, badly, on vibes. Thorp made it on purpose, on math, and it bought him twenty years without a down year.

And notice how his discipline echoes the traders we’ve already profiled from completely different worlds: Paul Tudor Jones obsessing over reward-to-risk, Daljit Dhaliwal sizing up only on the setups his record proved, Michael Burry capping risk at a defined premium even at maximum conviction. Different centuries, different instruments, one shared truth: the edge tells you whether to bet; the sizing tells you whether you’ll be around to collect. As a bonus, Thorp’s clarity of thought also let him spot Bernie Madoff as a fraud in 1991, years before the collapse — the same refusal to accept a too-smooth story that runs through Fooled by Randomness.

From belief to behavior: are you sizing like Thorp or guessing?

You don’t need Kelly’s formula to test whether your sizing is disciplined or emotional. The pattern is in your trade history.

Resources and further reading

• The memoir: Thorp, E. O. (2017), A Man for All Markets: From Las Vegas to Wall Street — his own account of blackjack, the funds, and Kelly sizing.
• The blackjack proof: Thorp, E. O. (1962), Beat the Dealer; and Thorp & Kassouf (1967), Beat the Market, the pre-Black-Scholes hedging approach.
• The formula’s origin: Kelly, J. L. (1956), “A New Interpretation of Information Rate,” Bell System Technical Journal — the basis of the Kelly criterion.
• Thorp on Kelly: Thorp, E. O. (2006), “The Kelly Criterion in Blackjack, Sports Betting, and the Stock Market,” on applying edge-based sizing to markets and the case for fractional Kelly.
• The survival companion: Gecko, How Successful Traders Manage Risk — the practitioner version of never betting so much you risk ruin.

Frequently asked questions

Trader profile in Gecko’s trading psychology series. Biographical details and performance figures for Ed Thorp and Princeton-Newport Partners are drawn from Thorp’s A Man for All Markets and public accounts; figures are approximate and reflect specific historical periods, not a forecast or a typical result. The Kelly criterion is presented for education; full Kelly involves large drawdowns and assumes a known edge, and most practitioners use a fraction of it. Gecko is an educational and informational tool. Nothing here is financial, investment, or trading advice. Trading carries substantial risk of loss.