bable.pro
Probability is the most modest of the sciences. It does not promise certainty — it offers only the rough shape of what is likely. Yet in this shape we find a strange consolation: that randomness, observed at scale, behaves with the orderliness of a constitutional monarch. The roll of a single die is anarchy; the roll of ten thousand dice is law.
Pierre-Simon Laplace, writing in 1812, called this the Theorie analytique des probabilites. He was not the first to study chance — the chevalier de Mere had pestered Pascal with gambling questions a century and a half earlier — but he was the first to suspect that probability was not merely a tool for accounting wagers. It was, he argued, the proper grammar of any inquiry where ignorance and evidence both played a part. Which is to say: nearly all of them.
Read this page slowly. There is no menu, no dashboard, no sidebar of related articles. There is only descent — from the parchment-warm shallows of intuition into the slate-blue depths of mathematical certainty. The page itself is a small treatise. You are not browsing; you are reading.
Two hundred dots, falling without pattern.
Then, all at once, they remember a shape they were never told.
value
frequency
A short genealogy of the bell.
Before the curve was famous, it was furtive. Abraham de Moivre derived it in 1733 as a footnote to a problem about coin tosses, and let it sit unloved in an obscure pamphlet. It was Laplace, half a century later, who recognised its universal vocation; and Carl Friedrich Gauss who, working independently on the orbit of the asteroid Ceres, gave it the name we still mostly use.
The curve owes its persistence to a theorem that ought to surprise us more than it does: the central limit theorem. It says, roughly, that whenever you average enough independent things — almost any kinds of things — the average behaves as if drawn from this exact bell. Heights, errors, IQ scores, bullet holes on a battleship. The shape is universal because the act of averaging is universal. The bell curve is what averaging looks like.
A cloud of correlations.
A line, drawn through the cloud, that minimises every regret at once.
x
y
Random fragments, sorted by colour.
A pie. Playfair drew the first one in 1801.
- certain
- probable
- plausible
- doubtful
A meditation, before the finale.
We are now deep enough in this descent that the parchment has cooled to slate. The text has lightened to compensate. This is not arbitrary — the page is, in its small way, a thermometer of intellectual depth. The bell-curve intuitions of the upper sections have given way to the harder question: what does a probability mean?
Two answers contend. The frequentist holds that probability is the long-run frequency of an event in repeated trials — the proportion of fair-coin tosses that come up heads, in the limit. The Bayesian holds that probability is a degree of belief, updated as evidence accumulates — how confident a rational agent ought to be that the coin is fair, having seen a few flips. The two camps have been arguing since the eighteenth century. The argument has not ended. It will not end, because it is not really an argument about coins.
John Maynard Keynes, in his Treatise on Probability (1921), suggested a third path: probability is a logical relation between propositions, neither subjective nor frequentist but objective in the way that a valid syllogism is objective. The view never quite caught on, but it has the merit of taking probability seriously as a part of reason, not merely a tool for gamblers and insurers.
Whichever camp you join, the descent ends in the same place: a square, a circle, and the steady accumulation of dots. Below, you may witness the most famous demonstration in all of probability — the one in which mathematics itself is wrenched, by sheer chance, from the air.
Buffon's needles, recast as raindrops.
Click anywhere on the square to summon a sudden rain.
The dots fall, and pi emerges — not because we have computed it, but because we have agreed to count. Every approximation will wobble, none will arrive. This is the lesson of the descent: probability does not promise the truth. It only promises a beautiful approach.
— bable.pro —