# BLACK, TWENTY-SIX TIMES IN A ROW *The Gambler's Fallacy in a Tailored Suit* Author: Elveander Welfendorff · Category: Risk · Date: 2026-10-29 · Place: Warsaw · EW Journal Source: https://ewbyew.com/black-twenty-six-times-in-a-row/ --- The story is told with three numbers. Twenty-six blacks fell in succession at Monte Carlo on 18 August 1913, the odds against it were one in sixty-seven million, and the players, certain that red was owed, lost millions of francs. The sixty-seven million is the arithmetic of a coin, two multiplied by itself twenty-six times. A wheel with a single zero makes a run of black about one in 137 million, and a run of either colour about one in 68 million, so the famous figure comes out nearly right by accident. No document comes with it. I could not find a retelling that cites a casino record of the night or a newspaper of the day. That is a pleasing state of affairs for a story about people who judge chance badly, since the story has been judged badly itself. It has the shape that a lecture needs, with a crowd, a streak and a punishment, and it has been repeated in that shape for so long that nobody asks for the ledger. The casino is easier to document. François Blanc, who ran the casino at Bad Homburg, took over the gaming at Monaco in 1863 on a lease of fifty years, and the resort was christened Monte Carlo in honour of Prince Charles. By one account the lease ran out in 1913, the year of the story, which is a tidy end for a lease and a suspiciously tidy date for a legend. Laplace described the crowd’s reasoning more than a century before the night in question. His men were eager for a son, and they watched the monthly register of births and reasoned that the boys already born must make a girl likelier for them, as though the register were keeping accounts. In 1971 Amos Tversky and Daniel Kahneman gave the habit a name, the belief in the law of small numbers, the conviction that a short run of anything must resemble the long average and will hurry to make up the difference. Their later paper put the roulette case plainly. After a long run of red, most people believe that black is due, because black would make the sequence look more like chance, and the wheel has no view of the sequence. It runs in both directions, which ought to have been a clue. In a lottery studied by Charles Clotfelter and Philip Cook, a number that had just been drawn was chosen by markedly fewer players over the following days, who reasoned that it was spent. Forty-one people had picked 244 on the day in April 1988 when it won, and three days later twenty-four did. A number that has not appeared is due, and a number that has appeared is finished, and the lottery was happy to take money on both opinions. Boethius gave the Middle Ages the wheel of Fortune in the second book of his Consolation of Philosophy, and it turned in manuscript miniatures and in the rose windows of cathedrals from Amiens to Basel. Four people ride it. The one on the left is climbing and says that he shall reign. The one at the top says that he reigns. The one on the right is falling and says that he has reigned. The one at the bottom, writhing, says that he has no kingdom. Fortune herself is drawn larger than her passengers. She is sometimes blindfolded, sometimes bald behind with one lock of hair at the front for the quick to seize, and sometimes she turns the wheel with an awkward crank. Fortune’s wheel really does owe something, because it is a sermon, and the sermon says that whoever is at the bottom is on the way up. The roulette wheel says nothing of the sort. Italy’s own game is Genoese. After Andrea Doria took power in 1528 the republic chose five of its councillors by lot from the senators, and the citizens began to bet on who would be picked. The list was cut to ninety names, and in 1620 a man named Benedetto Gentile is credited with detaching the draw from the election, so that the ninety names became ninety numbers and the senators were no longer needed. The republic made the lottery official in 1643, as a source of revenue, and Italian Lotto still draws five balls from ninety. Italians call such numbers ritardatari, the latecomers, and the newspapers print the list. On 19 April 2000 the 31 came up at Bari after 167 draws, with a crowd cheering outside the building and about 900 million euros in prizes. The most famous case is the 53, which stayed away, on the wheel kept at Venice, for 182 draws and finally came up on 9 February 2005. In a game that draws five numbers from ninety, a given number misses 182 draws in a row about once in thirty-three thousand attempts, and with ninety numbers on each of several wheels and decades of Saturdays something is bound to hold the record. The nation did not see it that way. A motion put to the Chamber of Deputies in December 2004 reckoned that more than six billion euros had been staked on the one number, against a payout it put at about 1.4 billion. A lottery-shop owner in Florence told a magazine that some clients played seventy to a hundred thousand euros on every draw, and families in Tuscany were reported to have mortgaged their homes. When the number finally appeared the prizes were reckoned at 513 million euros. Having learned nothing, the nation did it again. In the autumn of 2016 the 53 was 208 draws overdue on the national wheel, with about nine million euros a draw going on it, and by one count it eventually came up after 257. People in good suits do the same. In 2016 Daniel Chen, Tobias Moskowitz and Kelly Shue looked at three kinds of professional decision: asylum judges, loan officers and baseball umpires. In all three they found that a decision tended to be made in the opposite direction from the one before, whatever the merits of the case in front of the decider. A judge who had just granted two claims was more likely to refuse the third. The effect was stronger among the less experienced and where accuracy was less rewarded. Markets have a more courteous vocabulary. A fund committee says that a sector is oversold, or that a multiple has compressed to a level that cannot last, and it is sometimes right, for a reason. A price can fall far below the value of what it stands for and climb back because the value is there. A roulette ball has no value to return to. A rebound earns its place in a committee paper when someone can name the thing that will do the pulling. An argument that points to nothing but a long wait is the roulette player’s argument in a better coat. On 29 December 1989, the last trading day of that year, the Nikkei 225 closed at 38,915.87. In March 2009 it touched 7,054.98. It closed above the old record on 22 February 2024, at 39,098.68, thirty-four years and two months later. By the reasoning of the crowd at Monte Carlo the market had been due a rise for thirty years. It arrived on a morning after an American chip designer had reported record sales the evening before, with the Bank of Japan a decade into buying bonds and funds and the yen weak. When mathematicians needed a name for the method that simulates chance by running it again and again on a computer, they took the casino’s. Nicholas Metropolis coined the name soon after a letter from John von Neumann in 1947, and the Los Alamos account of the christening adds that Stanislaw Ulam had a Polish uncle who kept borrowing money from his relatives because he just had to go to Monte Carlo. History does not record which colour he backed. * * * --- © 2026 Elveander Welfendorff. Quotation with attribution and a link to the source is welcome.