Drop rate calculator
Type a loot table and read what it means for players: each item’s chance on one pull and within 10, 50 or 100 pulls, how many pulls until half, nine in ten or 99 in 100 players have it, what a pity rule changes, and how long collecting them all takes.
Free, no account. The table stays in your browser.
A hard pity: if the item has not dropped before that pull, it drops on it. Weights are any positive numbers and need not add up to 100.
Memory crystal
Per pull 1%
- Pulls for half of players
- 69
- Pulls for 9 in 10
- 230
- Pulls for 99 in 100
- 459
| Item | Per pull | At least one in 10 pulls | At least one in 50 pulls | At least one in 100 pulls | Pulls for half of players | Pulls for 9 in 10 | Pulls for 99 in 100 |
|---|---|---|---|---|---|---|---|
| Salt chunk | 60% | > 99.9% | > 99.9% | > 99.9% | 1 | 3 | 6 |
| Brine pearl | 24% | 93.6% | > 99.9% | > 99.9% | 3 | 9 | 17 |
| Tarred rope | 11% | 68.8% | 99.7% | > 99.9% | 6 | 20 | 40 |
| Tide compass | 4% | 33.5% | 87% | 98.3% | 17 | 57 | 113 |
| Memory crystal | 1% | 9.6% | 39.5% | 63.4% | 69 | 230 | 459 |
Collecting them all: half of players hold every item by pull 75, nine in ten by pull 230 and 99 in 100 by pull 459; 105.4 pulls on average.
How the numbers are worked out
These are the functions behind the drop-rate panel in Nodary’s spreadsheets, and the ones AI agents call over MCP. Exact wherever the maths allows.
- Chance per pull
An item’s weight divided by the sum of all weights.
p = weight / total- At least one within n pulls
One minus the chance of missing the item n times in a row. Pulls are independent, so a rare item is never certain after any fixed number of them.
1 − (1 − p)^n- Pulls until a share of players have it
The first pull count at which that chance reaches one half, nine in ten or 99 in 100: the quantiles of a geometric distribution. The average hides how long the unlucky tenth waits; these do not.
smallest n with 1 − (1 − p)^n ≥ q- Hard pity
If the item has not dropped before pull N, it drops on pull N. Nobody waits longer than N, and the average wait shortens.
average = (1 − (1 − p)^N) / p- Collecting them all
The coupon collector’s problem with unequal chances. Up to twelve items it is exact, by inclusion and exclusion over every subset of the table; past that, 4,000 runs simulated from a fixed seed.
A worked example
In the table above the Memory crystal weighs 10 out of 1,000: a 1% chance per pull. After 100 pulls, 63.4% of players have one. Half of them got it by pull 69 and nine in ten by pull 230, but one player in a hundred is still waiting at pull 459. A pity on pull 90 ends every wait there and brings the average from 100 pulls down to 59.5.
Keep the loot table next to the design it balances
In Nodary a loot table is a sheet inside the design document that explains it. Open the drop-rate panel from the sheet’s toolbar and these answers follow the weights as your team edits them.
Open Saltmarrow’s drop tables- Item and weight columns found by their titles, and changeable
- An AI agent connected over MCP reads the same answers with analyze_drops and changes a weight with set_cells
- Formulas for balancing in every sheet: POWER, SUMPRODUCT, PERCENTILE, COMBIN and more
Questions designers ask
Is a 1% drop guaranteed within 100 pulls?
No. Each pull is independent, so after 100 pulls 63.4% of players have the item and the rest are still waiting. Half of players get it by pull 69, nine in ten by pull 230, and 99 in 100 only by pull 459.
What is a pity system?
A rule that guarantees an item after a number of pulls without it. A hard pity guarantees it on that exact pull. It caps the longest wait, the one players remember, and shortens the average wait less than it shortens the worst one.
Do the weights need to add up to 100?
No. Any positive numbers work: each chance is the weight divided by the total, so 10 out of 1,000 and 1 out of 100 are the same 1%.
Does it model soft pity or rate-up banners?
No. It models a hard pity on one item. A soft pity, where the rate climbs over the last pulls, or a 50/50 rate-up needs its own model, which you can build with formulas in a Nodary sheet.
How many pulls does it take to collect every item?
That is the coupon collector’s problem. With unequal weights the rarest items dominate the answer. The calculator gives the pulls by which half, nine in ten and 99 in 100 players hold every item, exactly for up to twelve items.
Is my table sent anywhere?
No. The calculation runs in your browser as you edit. The table is kept in the page’s address, so a link you copy opens the same table for whoever you send it to, with a preview card that shows it. Download it as CSV to keep it, or put it in a Nodary sheet to work on it with your team.
Balance where the design lives.
Nodary is free for two members and ten documents, with no card.