How the probability works
Let p be the chance of a token from one eligible chest and n the number of eligible chests opened. It’s easier to calculate the chance of finding none, then subtract that from 1.
- One chest misses with probability
1 − p. - With independent rolls at the same rate, the probability that all n chests miss is the product of those miss probabilities:
(1 − p)n. - Finding at least one token is the opposite of missing every time, so the probability is
1 − (1 − p)n.
Why treat chests as independent?
The model treats each eligible opening as a fresh roll: a previous miss gives no reason to raise the rate for the next chest. That’s a modeling assumption, not a claim that we can inspect Epic’s server logic. A rate change during Treasure Hours is a separate issue: with different rates, multiply each chest’s own miss probability instead of using one exponent.
Why can a squad multiply the chest count by four?
If four teammates each open 20 eligible chests and a find can count for the whole team, there are 80 chances for a shared find. Multiply the number of rolls, not the per-chest drop rate. If people open different amounts, add the real counts: 20 + 15 + 10 + 5 is 50 rolls, not 80.
Two early community rates
| Per chest | Status | Where it came from |
|---|---|---|
| 1 in 700,000 | Estimate | An early community estimate with no verified sample or official rate behind it. The homepage and the examples below use this rate for consistency. |
| 1 in 2,000 | Disputed | A datamining-based community model that multiplies three assumptions: a 1% prize roll, a 10% chance of the top tier given a prize, and an even split between two top-tier rewards. Those weights were never verified as Epic’s server settings. |
Fortnite Status later published a normal-match rate of 1 in 137,500, also reported by Video Games Chronicle. Neither early figure represents the published in-game rate.
Chance per match by team size
Each column is the number of chests opened per player in one match. The percentages use the homepage’s 1-in-700,000 estimate and assume everyone on the team opens that many chests.
| Mode | 5 | 10 | 20 | 40 |
|---|---|---|---|---|
| Solo | 0.000714% | 0.001429% | 0.002857% | 0.005714% |
| Duos | 0.001429% | 0.002857% | 0.005714% | 0.011428% |
| Trios | 0.002143% | 0.004286% | 0.008571% | 0.017141% |
| Squads | 0.002857% | 0.005714% | 0.011428% | 0.022855% |
Crowdsourced stats
A useful sample includes sessions where nobody found a token as well as successful runs. Record the number of openings, team size, and whether Treasure Hours were active. Counting only the people who got a token would make drops look more common than they are.
Try your own numbers with the homepage chest counter.