Epic ended a two-week community argument on November 7 by publishing the drop rate for the most hunted item of the Treasure Hunters event: Pirate Shop Access, 1 in 137,500 chests. One number explained everything the community had been arguing about — including why the popular estimates were off by an order of magnitude.

Here is the math we ran, and the free calculator we built to make it reproducible.

The official numbers

Epic's announcement (via @FortniteStatus) priced every treasure in normal chests:

  • Weekly Gold Outfit: 1 in 1,562,500
  • Pirate Shop Access: 1 in 137,500
  • Weekly Vehicle & Decal: 1 in 7,813
  • Weekly Pickaxe or Backbling: 1 in 568
  • Weekly Jam Track: 1 in 391
  • Weekly Loading Screen: 1 in 272
  • Gizmo (any): 1 in 127

Treasure Hours boosted rates on a sliding scale: 1.5x the first hour (Access → 1 in 91,667), 2x in later hours (→ 1 in 68,750), with Epic noting future boosts could go higher. One fixed multiplier for "Treasure Hours" is already wrong — the boost is an input, not a constant.

The probability math

Each chest is an independent roll, so the chance of at least one token across n chests is the standard binomial tail: 1 − (1 − p)^n.

Three thresholds make the cost of a token concrete:

Target chance Official (1 in 137,500) TH 1.5x (1 in 91,667) Pre-reveal 0.05% (1 in 2,000)
50% 95,308 chests 63,539 chests 1,386 chests
90% 316,605 chests 211,070 chests 4,605 chests
99% 633,209 chests 422,139 chests 9,210 chests

The expected wait is simply 1/p = 137,500 chests, and the memoryless property means a dry streak changes nothing — your 100,000th chest rolls at exactly the same rate as your first.

For scale: a busy evening might open 25 chests (a 0.02% shot at the official rate). A hardcore two-week grind of 10,000 chests lands at roughly 7%. That is the arithmetic behind the players who farmed the entire event and still finished empty-handed.

How many players actually got one?

No census exists — Epic published the rate, not a redemption count. But the holder count is two multiplications away, and the difference matters:

  • Players who opened chests: ~2,000,000 to ~5,000,000
  • Chests per participant: ~20 to ~60
  • Chests opened event-wide: ~40,000,000 to ~300,000,000
  • Tokens dropped: 40M ÷ 137,500 ≈ 291, up to 300M ÷ 137,500 ≈ 2,182

Roughly 290 to 2,200 tokens across the whole event — but tokens are not people. One squad find equips the entire party, so the PEOPLE count runs about 2.5 to 4 times above the token count: roughly 700 to 8,800 people, minus duplicate drops from players who kept rolling after their first hit (repeats upgrade nothing — a holder's second token is pure waste).

The pre-reveal community math had said 20,000 to 150,000 people, based on a 0.05% per-chest estimate that was roughly 70 times too steep — about 2 to 30 times too generous on the people caliber. Two structural errors did the damage:

  1. Survivorship: the players who found a token posted about it; the thousands who found nothing scrolled quietly. Any estimate built from reported finds samples the winners of a 1-in-137,500 lottery.
  2. Denominator chaos: squads counted chests four ways, solo players counted once, and the population-level guess skipped per-chest counting entirely — three methods answering three different questions.

The calculator

We made the math reproducible: the Fortnite Pirate Shop Token Odds Calculator runs Epic's official rate, any Treasure Hours boost, the pre-reveal estimates, a target-probability reverse calc, and a dry-streak check ("I opened 500 chests and found nothing — is that normal?" → yes: 99.6% of runs look exactly like that).

The full holder-count breakdown — the complete formula with the Treasure Hours factor, party-size assumptions, and duplicate drops spelled out — lives at How many people have the Pirate Shop Token?.

The dev lesson

Hidden drop rates are a vacuum, and communities fill vacuums with confident guesses. Epic's reveal settled the argument in one line — and the line landed between the community's guesses: roughly 70 times steeper than the popular figure, five times friendlier than the steepest one. If you ship any kind of rarity system, publishing the rate (even a single number) is cheaper than two weeks of ReleaseTheOdds hashtags.

One deadline note for anyone still holding a token: the loot stops, but the reward does not. A paying chest hands over two access items — the 2026 one must be activated by December 15, 2026 (activating it starts 14 days of free claims on V-Bucks Item Shop items), and the 2027 one unlocks next year.