For a listed item rate p and n independent openings, the chance of at least one target is 1 − (1 − p)ⁿ. The openings required for a selected confidence are rounded up from the logarithmic inverse of that formula.
A 0.1% target has a statistical mean of 1,000 openings, but 1,000 attempts only gives about a 63.2% chance of seeing at least one. Reaching 90% needs 2,302 attempts under the unchanged-rate assumption. This distinction is why the tool shows both mean and confidence threshold.
Sheckle totals multiply the displayed crate price by openings. They exclude the time required to wait for stock and do not value duplicates, Robux bundles or player trading.