THE FOLD / LOOT / THE DROP / THE COLLECTOR
THE COLLECTOR
collect them all — n*Hn draws, tail-heavy
1 WHAT IT IS · WHAT IT DOES · FACT OR FICTION
The coupon collector. There are n different prizes in the cereal boxes, one uniformly at random per box. How many boxes must you buy to get all n? The answer is the harmonic sum:
E = n·Hn = n·(1 + ½ + ⅓ + … + 1/n) ≈ n·ln n.
The cruelty is in the tail. The first coupons come easy — almost every draw is new. But the last one, when you already have n−1, appears with probability only 1/n, so it takes about n draws all by itself — as long as collecting the entire first half. Every gacha game, every ‘collect them all’, lives on this curve.
LIT verified live: for n = 5, 10, 20, 50 the empirical average number of draws matches n·Hn to within a fraction of a percent over tens of thousands of trials (window.__coupon.matchesFormula). FIG no framing; the n·Hn expectation, checked against simulation, is exact.
E = n·Hn = n·(1 + ½ + ⅓ + … + 1/n) ≈ n·ln n.
The cruelty is in the tail. The first coupons come easy — almost every draw is new. But the last one, when you already have n−1, appears with probability only 1/n, so it takes about n draws all by itself — as long as collecting the entire first half. Every gacha game, every ‘collect them all’, lives on this curve.
LIT verified live: for n = 5, 10, 20, 50 the empirical average number of draws matches n·Hn to within a fraction of a percent over tens of thousands of trials (window.__coupon.matchesFormula). FIG no framing; the n·Hn expectation, checked against simulation, is exact.
2 HOW IT WAS WEAVED · AI + HUMAN
David (human) seated this in THE DROP, beside THE RANDOM and THE NEEDLE — the loot domain of what falls when you keep pulling. The coupon collector is the mathematics of ‘collect them all’: the drops come, but the set completes far slower than it feels. AVAN (AI) built the instrument: the random draws, the harmonic expectation, the empirical match.
The weave: David names the seat (the endless pull for a full set); I make the diminishing returns visible and the formula checkable — the collection curve in 1D, the live draw in 2D, the filling ring in 3D. The sphere is the seam. Credit: classical probability (de Moivre, Laplace); asymptotics by Erdős & Rényi.
The weave: David names the seat (the endless pull for a full set); I make the diminishing returns visible and the formula checkable — the collection curve in 1D, the live draw in 2D, the filling ring in 3D. The sphere is the seam. Credit: classical probability (de Moivre, Laplace); asymptotics by Erdős & Rényi.
3 ONE DIMENSION
The collection curve: distinct coupons owned versus draws made. It rockets up at first — every pull a new prize — then bends and crawls, each remaining coupon rarer than the last, the final one an eternity away.
4 TWO DIMENSIONS · INTERACTIVE
Draw coupons and watch the set fill — fast, then agonizingly slow. The draw counter climbs toward the n·Hn prediction; the last few slots hold out for hundreds of pulls. Change n and feel the ln n grind.
5 THREE DIMENSIONS + AVAN’S INVERSE
The coupons as a turning ring, filling in as they arrive — green, the ones you already hold.
AVAN’s addition (the inverse-companion): the magenta slots are the still-missing coupons — and they are what the whole cost is about. Progress feels linear: n things, surely n-ish draws. The truth is inverted and back-loaded: the difficulty piles into the tail. The last coupon alone costs n draws — as much as the entire first half took — and the total is n·ln n, not n. Completion is not a sum of equal steps; it is a curve that flattens into near-impossibility, the reward for the final piece the same as for all the easy ones combined. The green is the easy majority; the magenta is the vanishing few that make ‘collect them all’ a long, harmonic grind.
LIT The coupon collector problem (classical probability; de Moivre, Laplace; Erdos-Renyi asymptotics). Verified live: for n = 5, 10, 20, 50 the empirical average number of draws to collect all coupons matches n*Hn to within a fraction of a percent over tens of thousands of trials (window.__coupon.matchesFormula === true). The n*Hn expectation and the tail-heavy structure (the last coupon costing ~n draws) are exact and simulation-confirmed.
FIG No metaphor is doing the work: the n*Hn expectation is a real theorem, checked against simulation. The 'collect them all' framing is literal — this is exactly the mathematics of gacha completion and diminishing returns.
FIG No metaphor is doing the work: the n*Hn expectation is a real theorem, checked against simulation. The 'collect them all' framing is literal — this is exactly the mathematics of gacha completion and diminishing returns.
◆ sealed .dlw.fold → folded to ROOT_0 · a sphere of THE DROP · David Lee Wise (ROOT0), with AVAN