Reference · Acquisition
Randomised rewards and drop rates explained
A published drop rate is a precise statement about a single attempt and almost no statement at all about a session. The gap between those two things is where most of the disappointment in this genre comes from, and it closes entirely once the arithmetic is in front of you.
The short version
A one per cent rate means each attempt independently has a one in a hundred chance. Ten attempts give roughly a 9.6 per cent chance of at least one success, not ten per cent. A hundred attempts give roughly 63 per cent — still not a certainty. Where a title publishes a pity counter, that number bounds the worst case and is more useful for planning than the rate itself.
What a published rate means
A drop rate is a per-attempt probability. If a title states that a legendary character appears at one per cent, that describes each individual summon in isolation. The attempts are independent: an unsuccessful run does not make the next attempt more likely, and the system has no memory of what it has already given you unless the title has explicitly built one, which is what a pity counter is.
Two habits of thought make this harder than it needs to be. The first is treating probabilities as additive, so that a hundred attempts at one per cent feels like a guarantee. The second is the sense that a long unlucky run is owed a correction. Neither holds. The arithmetic below replaces both with figures you can check.
Published rates also vary in what they cover. A rate for a rarity tier is not a rate for a specific character within that tier, and the difference can be large where a tier holds dozens of characters. Where a title publishes a rate at all, read whether it applies to the tier, to the featured character, or to the pool as a whole.
The arithmetic, worked through
The chance of at least one success across several independent attempts is easier to calculate from the other direction: work out the chance of failing every time, then subtract it from one. For a rate r across n attempts, the chance of at least one success is 1 − (1 − r)n.
| Stated rate | Attempts | Chance of at least one success | Worth noticing |
|---|---|---|---|
| 1% | 10 | 9.6% | Fewer than one in ten sessions of this size succeed. |
| 1% | 69 | 50.0% | The halfway point sits far later than intuition suggests. |
| 1% | 100 | 63.4% | More than a third of players at this count still have nothing. |
| 1% | 230 | 90.0% | Even here, one player in ten is still waiting. |
| 0.5% | 200 | 63.3% | Halving the rate roughly doubles the attempts for the same odds. |
Worked example
Suppose a title publishes a one per cent rate for its top tier, and you have saved enough currency for forty summons. Your chance of at least one top-tier result is 1 − 0.9940, which is about 33 per cent. Two readings of that number are both correct and lead to different decisions.
The optimistic reading is that a third of players in your position will succeed. The planning reading is that two thirds will not, so any plan that depends on succeeding is a plan with a two-in-three chance of failing. If the character is the reason you are playing, the second reading is the one to budget against — and the honest conclusion is often to wait for a pity counter rather than to summon at forty.
Pity counters change the shape
A pity counter guarantees a given outcome after a stated number of unsuccessful attempts. Where one exists, it converts an open-ended gamble into a bounded purchase: the worst case is known, and it can be priced. A title with a published rate and no pity counter has an unbounded worst case, which is a materially different proposition even if the headline rate is identical.
Pity systems differ in ways worth checking. Some reset when they trigger and some carry across banners. Some guarantee only the rarity tier rather than the specific character, which means reaching pity can still deliver something you did not want. Some count only paid attempts. Each of those details changes the real cost, and each is usually documented somewhere in the game's own information screens.
Banners and timing
A banner is a time-limited pool with adjusted probabilities, usually promoting a new character. Banners concentrate spending because they add a deadline to a decision that would otherwise have no deadline, and they are the main reason spending in this genre is lumpy rather than steady.
The useful discipline is to decide the budget before the banner and not during it. A figure set in advance, in AUD, survives a run of unsuccessful summons; a figure set while summoning tends to move. This is not a claim about anyone's self-control, simply a note that the sequence in which you make the two decisions changes the outcome.
Turning probability into a budget
To compare two titles, or to decide whether a particular character is worth pursuing, reduce everything to one figure: the price in AUD of reaching the guaranteed outcome, if there is one, or of a number of attempts you choose in advance, if there is not.
- Find the price of the currency packs and work out the cost per summon in AUD. Larger packs usually carry a better unit rate, so use the pack you would actually buy.
- Find the pity threshold, if the title publishes one, and multiply. That is your bounded worst case.
- If there is no pity threshold, pick an attempt count you are willing to fund, and treat any result beyond it as not happening.
- Compare that figure to the monthly limit you set for yourself, not to the price of the pack. The pack price is designed to look small; the total is the decision.
Where this sits in Australia
Several Australian bodies have published roles that touch on randomised in-game rewards, and it is worth being precise about which does what rather than treating them as interchangeable.
- Australian Classification Board
- Classifies computer games under the national classification scheme and publishes the classification guidelines along with a searchable database of decisions. A title's classification and its consumer advice lines are recorded there, and that database is the authoritative source rather than a store listing.
- Australian Competition and Consumer Commission
- Administers the Australian Consumer Law, which prohibits misleading or deceptive conduct in trade or commerce and sets out the consumer guarantees. Claims a business makes about what a purchase delivers fall within that framework.
- eSafety Commissioner
- Australia's online safety regulator, which publishes guidance for families on games and online interaction and operates complaint schemes for certain categories of online harm.
- Scamwatch
- Run by the National Anti-Scam Centre, this is where scams are reported. Offers to sell accounts, currency or guaranteed summons outside official channels are a recurring category and are covered on the account security page.
If you believe a specific claim about a randomised reward has misled you, the ACCC's site sets out how consumer complaints are handled, and your state or territory fair trading office handles individual consumer disputes. This site does not give legal advice, and nothing above should be read as a statement about the legal status of any particular product.
Younger players
Randomised rewards bought with money share a structure with wagering: a stake is paid, the outcome is uncertain, and the result is revealed after the fact. That structural similarity is a reason for caution regardless of how any particular product is classified. A game whose economy includes purchasable randomised items should not be assumed suitable for a child or teenager simply because its setting looks harmless.
The practical steps are to check the title's entry in the national classification database and its consumer advice, to use the payment controls offered by the platform or the operating system so that purchases require an adult's authorisation, and to read the eSafety Commissioner's material on games. Those are the sources with standing on the question; this page has none.
Common questions
Does a long unlucky run improve my next chance?
Not unless the title has built a pity counter that says so. Independent attempts have no memory. The feeling that a correction is owed is a well-documented reasoning error rather than a property of the system.
Are drop rates always published?
No. Publication varies by title and by the requirements of the platform a title is distributed on. Where rates are not published, treat any figure circulating among players as an estimate of unknown quality rather than as a fact, and note the absence as a point against the title when you evaluate it.
Is a ten-pull better than ten single pulls?
Only if the title attaches something extra to the batch, such as a guaranteed minimum rarity or a discount. Where nothing extra is attached, ten attempts are ten attempts and the arithmetic above applies unchanged. Check what the batch guarantees before assuming it is better.
Related reading
The glossary's currency section defines the terms used here. Spending and consumer rights covers converting currency packs to AUD and what to do when a purchase goes wrong. The evaluation checklist includes rate publication and pity thresholds among the things to check before installing.