You've killed the boss 47 times and still no legendary drop. The game says the item has a 5% drop rate, so mathematically you should have gotten it by now — right? This frustration is nearly universal among players of loot-based games, and it stems from a fundamental misunderstanding of how probability actually works. A percentage drop rate is not a guarantee per attempt; it is an independent probability applied fresh to every single kill, with no memory of what came before.
Why RNG Feels Unfair (And the Math Proving It)
Random Number Generation in games draws a new random number on every attempt. A 5% drop rate means each kill has a 95% chance of producing nothing. The previous 46 empty kills have zero influence on kill number 47. The universe does not owe you a drop.
This independence is precisely what makes streaks feel so painful. A player who kills a 1% drop-rate boss 100 times has not guaranteed themselves the item — they have experienced 100 separate 1% coin flips. The probability of getting at least one drop in 100 tries is not 100%. It is approximately 63.4%.
The math works through the complement: instead of calculating the chance you do get the item, calculate the chance you never get it, then subtract from 1.
The Drop Rate Formula: 1 − (1 − p)^n
The probability of obtaining at least one drop of an item with drop rate p across n attempts is:
P(at least one drop) = 1 − (1 − p)^n
Where:
p = drop rate as a decimal (e.g., 0.01 for 1%)
n = number of attempts
Example: 1% drop rate, 100 attempts
P = 1 − (1 − 0.01)^100
P = 1 − (0.99)^100
P = 1 − 0.366
P = 0.634 (63.4% chance)
To find how many runs n are needed for a target confidence level C:
n = log(1 − C) / log(1 − p)
Example: 1% drop, 90% confidence
n = log(1 − 0.90) / log(1 − 0.01)
n = log(0.10) / log(0.99)
n = (−2.3026) / (−0.01005)
n ≈ 229 runs
Probability Table: Runs Needed for 50%, 90%, 99% Confidence
This table shows how many attempts are needed to have a 50%, 90%, or 99% chance of seeing at least one drop, across different drop rates.
| Drop Rate | 50% Confidence | 90% Confidence | 99% Confidence |
|---|---|---|---|
| 0.1% (1 in 1,000) | 693 runs | 2,302 runs | 4,603 runs |
| 0.5% (1 in 200) | 139 runs | 460 runs | 920 runs |
| 1% (1 in 100) | 69 runs | 229 runs | 459 runs |
| 2% (1 in 50) | 35 runs | 114 runs | 228 runs |
| 5% (1 in 20) | 14 runs | 45 runs | 90 runs |
| 10% (1 in 10) | 7 runs | 22 runs | 44 runs |
| 25% (1 in 4) | 3 runs | 8 runs | 16 runs |
Notice that even at 50% confidence you need 69 attempts for a 1% drop — meaning roughly half of players who farm that boss 69 times will still leave empty-handed. At 99% confidence you need 459 attempts. The long tail of bad luck is very real and very long.
Pity Systems: How Games Fix Bad Luck
Pity systems are developer-side mechanisms that guarantee a reward after a certain number of failed attempts. They exist because pure RNG produces outliers who experience 10× the average bad luck, creating player frustration and churn.
Hard pity sets an absolute ceiling. In Genshin Impact, the hard pity for 5-star characters on the standard banner is 90 pulls — you are guaranteed a 5-star by pull 90 regardless of all previous pulls.
Soft pity increases the base rate incrementally before hard pity triggers. In Genshin Impact, the base 5-star rate is 0.6%. Starting at pull 74, the rate increases by approximately 6% per additional pull. By pull 89 the effective rate has climbed to over 90%, meaning most players naturally hit the 5-star between pulls 74 and 90 without reaching the hard cap.
Pseudo-pity (bad luck protection) in games like Path of Exile increases drop chance after a long drought without a formal guarantee. World of Warcraft uses a similar "bad luck protection" system for many rare drops where internal counters increase probability after consecutive failures.
Gacha Math: Expected Cost Per SSR
In gacha games, "SSR" or equivalent top-tier characters typically have a base rate of 0.3%–3% depending on the game. The expected number of pulls to obtain one SSR follows:
Expected pulls = 1 / p
At 0.6% base rate (Genshin): 1 / 0.006 = 167 pulls (without pity)
With hard pity at 90: Expected is approximately 62–75 pulls
At a typical pull cost of $1–2 per pull (varying by bundle), a single 5-star in Genshin Impact costs on average $75–150. To guarantee a specific character (50/50 banner system: 50% chance of the featured character, else a random 5-star), expected cost doubles to $150–300. Guaranteeing the featured character on a losing 50/50 streak can reach $300–600.
| Game | SSR/5-Star Rate | Hard Pity | Expected Pulls | Est. Cost (USD) |
|---|---|---|---|---|
| Genshin Impact | 0.6% | 90 | ~62–75 | $80–150 |
| Honkai: Star Rail | 0.6% | 90 | ~65 | $75–130 |
| Blue Archive | 3% (3-star) | None (soft) | ~34 | $30–60 |
| Fate/Grand Order | 1% (SSR) | None | ~200 | $200–400 |
| Arknights | 2% (6-star) | 50 (guarant.) | ~35 | $30–70 |
The Gambler's Fallacy in Gaming
The gambler's fallacy is the belief that past failures make future success more likely in a system with independent trials. "I've gone 150 runs without a drop, so it must drop soon" is false for any game without a pity system. Each attempt is genuinely independent. The probability of the item dropping on attempt 151 is exactly the same as it was on attempt 1.
This matters practically: players who believe in the fallacy often continue farming well past the point where they should weigh the time cost objectively. If you've spent 200 hours farming a 0.1% drop rate item, the expected time for the next 200 hours is identical to what you faced at the start — the game has no memory.
The correct framing is to decide up front how many attempts you're willing to invest, calculate what confidence level that represents, and treat any drop before that ceiling as favorable variance. If 229 attempts gives you 90% confidence at 1%, and you find 229 runs acceptable, farm with peace of mind knowing you're covering the likely case. If 229 runs is not acceptable, the item is simply too expensive for your time budget — no amount of hoping changes the math.
Understanding loot probability does not make RNG less random. It makes you a better decision-maker about how to spend your gaming time and, in the case of gacha, your real money.