RPG Loot Drop Probability Calculator
A 5 percent drop rate does not mean five kills in a hundred drop the item — it means every kill rolls the same dice. Enter the drop rate and your kills, and this RPG loot drop probability calculator returns your true chance of having seen the drop, plus how many kills confidence actually takes.
The Only Formula in Farming: One Minus Failure to the N
Every kill is an independent roll, so the probability mathematics of farming is a single line: the chance of no drop in n kills is (1 − p)ⁿ, and the chance of at least one drop is its complement, 1 − (1 − p)ⁿ. The worked default runs the genre's most farmed shape — a 5 percent drop over 50 kills: failure per kill is 0.95, raised to the fiftieth power is 0.0769, and the answer is 92.31 percent. Note what the formula is telling you: after fifty kills at a rate that "should" drop once every twenty, one farmer in thirteen is still empty-handed, and nothing has gone wrong. The complement structure also explains farming's emotional shape — early kills buy probability cheaply (ten kills already bank 40 percent), later kills buy it dearly (the last few percent toward certainty cost hundreds of kills), because each kill multiplies the remaining failure chance rather than adding to a progress bar. Games that show a percentage are showing you this curve; games that hide it are hiding this curve.
Expected Kills, Medians and the Confidence Ladder
Three landmarks on the curve deserve their names, and the calculator's note prints all of them. The expected kills per drop — 1 ÷ p, twenty at the default rate — is the long-run average across all farmers, not a schedule for any one of them: half of all farmers finish before it. That halfway point is the median, the kills-for-50-percent figure (14 at the default: 0.95¹⁴ first dips the failure chance below one half). The confidence ladder extends the same inversion — kills for 90 percent (45) and for 99 percent (90) — and its spacing is the whole psychology of the grind: going from a coin-flip's confidence to near-certainty costs more than six times the kills, because certainty is bought in the curve's flat tail. When a guide says an item "takes about 50 kills," it is usually quoting the 90-percent rung without saying so; when a player reports 200 dry kills at a 5 percent rate, they are reporting a 0.0035 percent event — rare enough to suspect the rate, the loot table, or a bug, in that order of charity.
Pity Timers, Bad-Luck Protection and Rigged Dice
Many modern games quietly break the independence this page assumes, and the section exists so the formula's limits are as clear as its results. Pity timers guarantee a drop by kill N (or raise the rate with each failure), converting the curve into a ramp with a ceiling — under a hard pity at 50, the 50-kill chance is 100 percent, not 92.31. Bad-luck protection inflates the rate after dry streaks in ways developers rarely publish exactly. Personal-loot and shared-loot systems change whose rolls count; magic find and drop-rate bonuses multiply p, which this calculator accepts directly — enter your effective rate, gear included. The way to tell which universe you are farming in is the data itself: a long dry streak that ends in suspiciously round numbers of kills across many players advertises a pity system; the wikis of record (the Terraria Wiki and its peers document rates and mechanics per item) are where the fine print lives. Absent such systems, the pure formula rules — and its indifference to your history is not cruelty but arithmetic.
The Gambler's Fallacy, Wearing a Loot Filter
The fallacy this site meets at the coin flipper returns here in plate armour: after 80 dry kills, the drop is not "due." Each kill's chance is the printed rate, memoryless, and the 92.31 percent at 50 kills was never a promise being kept late — it was a proportion across thousands of farmers, most of whom finished earlier. What long streaks do license is suspicion of the inputs, in a strict order: is the rate what the patch notes claim (rates change between versions), is the item on this enemy's table at this difficulty, is a quest state or a lockout interfering, and is the sample being remembered selectively (the 30-kill dry streak that ends the session is remembered; the two early drops are not). The calculator's final use is therefore diagnostic: enter the kills and the claimed rate, and if the resulting chance reads 99.99 percent against your empty bags, believe the arithmetic and go read the loot table — something else is wrong, and it is findable.
Farming as Budgeting: Pricing Your Evenings
The confidence ladder converts directly into a schedule, and scheduling is how adults farm. Read the 90-percent rung as the honest session plan: at the default's 5 percent that is 45 kills — at three minutes a kill, two and a quarter hours — with a one-in-ten chance the evening still ends empty and a plan needed for that outcome that is not "rage-quit the game." Players who farm well pre-commit to the median instead (half of all attempts end by 14 kills, so most nights are short) and treat the tail as the cost of the hobby averaged over months, not as tonight's debt. Streamers and guide writers invert the same numbers into content: a drop's rarity is its story, and the 99-percent rung is the episode count of the unlucky. However the evening is framed, the formula has already priced it — kills, hours and confidence are one multiplication apart, and this page performs it before the first mob is pulled.
Related Tools
More from the Other Calculators hub and the statistics shelf: the Coin Flipper Calculator — the same independence, flipped — the Odds Calculator, and the D&D Character Stat Calculator for the tabletop version of the same dice.
Frequently Asked Questions
If an item drops 1 in 100, am I guaranteed it in 100 kills?
No — you are 63.4 percent likely to have it (1 − 0.99¹⁰⁰), the formula's most quoted surprise. A hundred kills at a one-percent rate leaves more than a third of farmers empty-handed, and reaching 99 percent confidence takes 459 kills. The "1 in 100" phrasing names the per-kill rate, and the expectation (100 kills per drop, averaged over everyone) coexists with huge individual variance. This gap between the rate's plain reading and its true behaviour is the entire reason the calculator exists.
How many kills for a 90 percent chance at a 2 percent drop?
One hundred and fourteen: the inversion the note automates — ln(0.10) ÷ ln(0.98) = 113.97, rounded up. Halving the drop rate from the default's 5 percent to 2 more than doubles the 90-percent rung (45 to 114), because the ladder's rungs scale with 1/p and then some. Enter your own rate and read the ladder before you commit an evening to a drop: the difference between farming at 90 percent confidence and farming at "expected" is the difference between a plan and a hope.
Do drop rates stack across a party?
Only if the loot is personal (each player rolls independently — then the party's chance of someone seeing it follows this same formula with n = total kills across the group, while your chance is unchanged). Under shared loot with one drop per kill, the party's kills are simply the group's farm rate. The subtle case is tradable loot in a coordinated group farming for one member: there the group total is the honest n, and the calculator prices the team's evening correctly.
What is the expected number of drops in 200 kills at 5 percent?
Ten — the expectation multiplies linearly (n × p) even though the at-least-one probability saturates. Expectation counts multiplicity (two drops in one evening are two), while this page's headline probability asks only whether the count reached one. Both views are printed here in spirit: the note's "expected kills per drop" is the same linearity read backwards. For planning consumables and inventory, expectation is the budget number; for "will I finish the set tonight," the probability is.
What does a 1 percent drop rate really feel like?
Run the ladder: half of farmers have it by 69 kills, nine in ten by 230, and the unluckiest one in a hundred is still farming past 458. The average is 100 kills, but the distribution's long right tail is the lived experience everyone quotes — the median farmer calls it easy, the tail farmer calls it rigged, and both are describing the same curve from different seats on it. That spread is not a defect of the system; it is the system, and this calculator exists so you can see your seat in advance.
Why do I keep getting duplicates instead of the one item I need?
Because the table rolls over its whole list, and duplicates are the modal outcome once you own most of it: needing one specific item from a ten-item table with equal rates is a 10 percent roll per drop event, and the completion problem (the coupon collector's) prices the last item at an average of ten further drop events after the second-to-last. Games answer with tokens, target-farming and duplicate protection; the mathematics answers with this page — enter the effective rate for the specific item (drop chance × its share of the table) and the ladder tells you the true length of the road.