Gacha Probability Calculator
Enter a drop rate and the number of pulls to calculate the probability of getting at least one target item. Optionally add a cost per pull to see the expected total spending.
Gacha Probability Explained: How Pull Rates, Pity Systems, and the Gambler's Fallacy Work
Gacha games — mobile and browser titles where players spend virtual or real currency to receive randomized rewards — have become one of the dominant revenue models in the global gaming industry. Titles such as Genshin Impact, Fate/Grand Order, Arknights, and Blue Archive collectively generate billions of dollars annually, with a significant portion coming from players chasing rare characters and items through pull mechanics. Understanding the mathematics behind gacha systems can help players make more informed decisions about how they spend their time and money.
The Core Formula: At Least One Success
The fundamental question gacha players ask is: 'If I do N pulls at a drop rate of r%, what are my chances of getting the item I want?' The answer comes from the geometric distribution, a standard probability model for repeated independent trials.
The probability of getting at least one copy of the target item is: P(≥1) = 1 − (1 − r)^n, where r is the single-pull drop rate expressed as a decimal and n is the number of pulls. This formula works because the probability of failing every single pull is (1−r)^n, and by subtracting that from 1 we get the complement — success on at least one attempt.
For example, a banner with a 3% featured rate and 50 pulls yields P(≥1) = 1 − (0.97)^50 ≈ 78.0%. This means a player doing 50 pulls has roughly a 78% chance of obtaining the featured item at least once. The remaining 22% of players who reach 50 pulls will have received nothing, which is why pity systems exist.
Expected Pulls and the Geometric Distribution
The geometric distribution also gives us the expected number of pulls before the first success: E[pulls] = 1/r. At a 3% rate this is approximately 33.3 pulls; at a 0.6% rate (common for five-star characters in some games) it rises to around 167 pulls.
The expected value is an average across many players and many sessions. It does not predict any individual outcome. In practice the distribution has a long right tail: while half of all players will succeed within about 23 pulls at a 3% rate, a meaningful fraction will need far more. Roughly 5% of players would need over 98 pulls to get a single copy at that rate.
This variance is part of what makes gacha emotionally engaging — and financially risky. The possibility of getting the target item on the very first pull, combined with the possibility of going 100+ pulls dry, creates the same psychological profile as variable-ratio reinforcement schedules studied in behavioral psychology.
Pity Systems
Most modern gacha games include a pity system — a guarantee that the player will receive a rare item after a certain number of unsuccessful pulls. Common implementations include a soft pity (drop rate begins increasing after a threshold, such as pull 74 out of 90) and a hard pity (a guaranteed drop at a fixed ceiling, such as pull 90).
Pity systems fundamentally change the distribution of outcomes compared to a flat rate. Instead of a pure geometric distribution, the effective model becomes a mixture distribution that accounts for the increasing rate in the soft pity zone. Simulations of games like Genshin Impact suggest that the effective rate including soft and hard pity is roughly 1.6% per pull for five-star characters, compared to the stated 0.6% base rate.
Pity systems also raise the question of pity transfer — whether accumulated pulls carry over between banners. In some games they do; in others, starting a new banner resets the counter. This distinction can significantly affect player strategy and spending.
The Gambler's Fallacy in Gacha
One of the most common misconceptions in gacha is the gambler's fallacy: the belief that past failures make future success more likely. In a system with a flat, independent drop rate (no pity), every pull is statistically identical to the previous one. Having done 50 dry pulls at a 3% rate does not raise the probability of the 51st pull — it remains 3%.
Pity systems are specifically designed to counteract this frustration by introducing a genuine dependency between pulls. Once a player enters the soft pity zone, future pulls genuinely do have higher success probabilities. However, outside of an explicit pity mechanic, each pull is independent, and the history of previous pulls carries no predictive information about the next one.
A related misconception is the hot-hand fallacy in reverse: the belief that a recent success makes failure more likely ('the banner is in a dry spell now'). In a fair random system with no memory, success on one pull has no effect on the probability of the next.
Milestone Probabilities: 50%, 90%, and 99%
Rather than asking 'what is my probability after N pulls?', players sometimes prefer to ask 'how many pulls do I need to reach a 90% (or 50%, or 99%) chance of success?' This requires inverting the at-least-one formula.
The formula n_x = ⌈log(1−x) / log(1−r)⌉ gives the minimum number of pulls needed. At a 3% rate: 50% chance requires about 23 pulls, 90% requires about 75 pulls, and 99% requires about 151 pulls. At a 0.6% rate those numbers rise to approximately 115, 383, and 765 pulls respectively.
These milestone numbers are useful for budgeting. A player who wants at least a 90% chance of getting the featured character should prepare the number of pulls corresponding to their target threshold, not just the expected value. Using only the expected value understates the spending required for high-confidence acquisition.
Expected Cost and Responsible Spending
Multiplying expected pulls by cost per pull gives the expected spending to obtain a single copy of the target item. This is the average across many players — some will spend far less (lucky early pulls), and some will spend far more (unlucky long dry streaks).
When budgeting for gacha spending, the expected cost is a useful baseline but not a cap. A responsible approach is to decide in advance the maximum number of pulls one is willing to perform, and to treat that budget as a fixed limit regardless of outcome — rather than chasing losses after a dry streak.
Many countries and platforms now require gacha games to disclose drop rates, and several jurisdictions have classified loot boxes with real-money purchase options under gambling regulations. This calculator is provided as a transparency and educational tool to help players understand the mathematics of the systems they engage with.
Frequently Asked Questions
What does the 'at least one' probability mean?
It is the probability that, out of all the pulls you perform, at least one of them results in the target item. For example, a 78% probability means that if many players each did the same number of pulls at the same rate, about 78% of them would obtain the item at least once. The remaining 22% would receive it zero times across those pulls.
Why does the expected number of pulls differ from the 50% milestone?
The expected value (mean) of the geometric distribution and the median are different values. At a 3% rate the expected pulls is about 33.3, but the median — the point where half of players have succeeded — is about 23 pulls. The distribution has a long right tail, meaning a small fraction of players need very many pulls, which pulls the average higher than the median.
Does a pity system change how this calculator works?
This calculator models a flat, independent drop rate with no pity adjustment. If a game has soft or hard pity, the actual probabilities will differ from the results shown here — typically the real probability of success within the pity window is higher than the flat-rate calculation suggests. For pity-adjusted calculations, you would need the specific mechanics of that game.
Is the gambler's fallacy real in gacha games?
In gacha systems with a flat, independent rate and no pity, yes — each pull is independent and past failures do not increase the probability of a future success. However, in games with pity mechanics, accumulated pulls do genuinely increase the probability of the next pull once the soft pity threshold is crossed, which is a real dependency, not a fallacy.
How is the expected cost calculated?
The expected cost is calculated by multiplying the expected number of pulls (1 divided by the drop rate) by the cost per pull. This gives the average amount a player would spend to obtain one copy of the target item across many attempts. Some players will spend less (lucky early pulls) and some will spend more (long dry streaks), but the expected cost represents the long-run average.