The Voltaire Lottery Syndicate of 1729 — The Original +EV Exploit

Three centuries before MIT students worked out Cash WinFall, a French government lottery contained a pricing error so large that a philosopher and a mathematician could simply buy their way through it. The structure of the mistake is worth understanding, because it is the same structure as every genuine lottery exploit since.

The scheme

In 1729 the French Controller-General, Le Pelletier-Desforts, needed to redeem a set of annuity bonds issued by the Hôtel de Ville in Paris, which had fallen in value. His solution was a lottery attached to the bonds: holders could surrender a bond for the right to enter a draw.

The design detail that mattered: the ticket price was proportionate to the size of the bond you held, but the prize was not.

The flaw

The mathematician Charles-Marie de La Condamine spotted it, reportedly in conversation with Voltaire at a dinner party. Stated plainly:

• Ticket price scaled with the value of the bond it was attached to.
• The chance of winning did not scale — each ticket was one entry.
• The prize attached to a winning ticket did not scale down for small bonds in proportion to how cheap the ticket had been.

So the cheapest tickets, attached to the smallest bonds, bought the same chance of a disproportionately large prize for a fraction of the price. A syndicate that cornered enough small-denomination tickets was not gambling at all. It was buying a positive expectation at scale.

This is exactly the shape of the Cash WinFall exploit 280 years later: the rules pay out more than the tickets cost, in a specific corner of the structure, and the only requirement is enough capital to occupy that corner.

The execution

Voltaire and La Condamine assembled a syndicate of investors and bought up a large share of the qualifying small bonds and their tickets. Winnings were pooled and divided among the members.

The results were substantial. La Condamine is recorded as holding thirteen winning tickets that had cost one livre each and entitled him to 13,000 livres. Across the operation the syndicate shared a sum reported at just over a million francs, with Voltaire's own take estimated at around half a million livres.

That money mattered historically as well as personally: it underwrote Voltaire's financial independence for the rest of his life, which is a relevant detail for a writer whose career consisted largely of annoying powerful people.

The legal outcome

The authorities investigated. The finding was that no law had been broken — the syndicate had identified and exploited a loophole in a scheme the government itself had designed and published.

The government's response was the only one available: it stopped running the lottery. The final draw was held in June 1730, and Le Pelletier-Desforts lost his position.

That is a template modern operators still follow. Massachusetts did not prosecute the Cash WinFall syndicates; it wound the game up. Ireland did not prosecute the 1992 syndicate; it changed the rules.

What the 1729 case teaches that the modern ones do not

Exploits live in the pricing, not the randomness. La Condamine never claimed to know which ticket would win. He noticed that the price of a chance varied while the value of that chance did not. Nothing about the draw needed to be predicted — a point worth holding onto whenever someone offers to sell you a number-prediction method (lottery prediction software tested).

The exploit needed capital, not genius. The insight was one sentence. Realising it required buying a large share of an entire market — the same requirement that made Cash WinFall a full-time operation for seven years and the 1992 Irish attempt a logistics exercise involving teams of buyers.

Small combination spaces are where this is possible. The 1729 lottery had a finite, purchasable set of qualifying tickets. So did the Irish Lotto in 1992, with 1,947,792 combinations. Modern jackpot games have hundreds of millions of combinations specifically so that no syndicate can corner them — see what it would cost to buy every combination and bulk-buying caps.

Every one of these holes was closed once found. There is no surviving 1729-style mispricing in a major modern lottery, because operators now model prize structures against exactly this attack before launch. The reason to read the history is to recognise the shape of a genuine opportunity — and to notice that nothing being marketed to you today has it.

For the general theory, see positive expected value in a lottery; for why even a real edge justifies almost no stake, the Kelly criterion.