Quick answer: A drop rate is the chance per chest, not the chance per session. At 1%, about 63% of 100-chest sessions see at least one drop; at 0.03%, only about 3% do. Always compare cumulative chances, not just the headline percentage.
Per chest versus per session
Most rate tables show a per-chest number. Players experience sessions. The gap between the two is where most disappointment lives.
If a sprite has a 1% rate, a 100-chest session has a 63% chance of at least one drop. If the rate is 0.1%, the same session has only a 9.5% chance. The headline numbers look close; the session experience is completely different. That is why a calculator or a simulation is more useful than comparing raw percentages.
| Reported rate | 100 chests | 500 chests | 1,000 chests |
|---|---|---|---|
| 3.5% | 97.1% | 99.999% | ≈100% |
| 1.5% | 77.9% | 99.95% | ≈100% |
| 0.1% | 9.5% | 39.4% | 63.2% |
| 0.03% | 2.96% | 13.9% | 25.9% |
The table makes the practical difference obvious. A 3.5% sprite is a realistic target for a normal weekend. A 0.03% sprite is a long chase that most sessions will not complete.
Reported does not mean official
Community rate tables are estimates. They come from aggregated openings, datamined pools, or creator testing. Each method has error. A table can be right about the order of magnitude and wrong about the second decimal. Treat every percentage as a range, not a fact.
Hotfixes matter too. A rate can change without a patch note. If a route suddenly feels different, check the date on the table before you blame your luck.
Three questions before you trust a number
- When was it last updated?
- Which chest type does it describe?
- How large was the sample?
If the answer to any of those is missing, lower your confidence. A precise-looking number with no sample is just a guess in a nicer font.
Why averages hide the story
“Expected drops” is a useful summary and a terrible plan. At a rare rate, most sessions end with zero and a small number of sessions carry the average. The distribution is skewed, so the mean is not the typical experience.
A simulation shows the shape: the share of empty sessions, the median, and the unlucky tail. If 80% of sessions end with nothing, a single session is a lottery ticket. If 80% end with at least one, the grind is predictable enough to schedule.
What to track in your own runs
Log four things: chest type, route, session length, and result. After twenty sessions you will have a personal picture that no global table can give you. It will not prove the true rate, but it will show whether your route is producing the draws you expect.
Turning a rate into a plan
Pick the sprite you want. Look up the reported rate. Decide how many chests you can open in one session without ruining your normal play. Then use the simulator to see the zero-drop share for that session size. If the share is high, either extend the plan across many sessions or pick a more reachable target.
That is the whole method: rate, session size, distribution, decision. It is less exciting than a hype video and far more useful.
Sample size: why 100 openings prove little
A rate estimate is only as good as its sample. If a creator opens 100 chests and sees two rare drops, the observed rate is 2%. The uncertainty around that estimate is wide enough to include 0.5% and 4%. Reporting “2%” as a fact turns noise into a number.
Larger samples narrow the range. This is why a community table with 10,000 logged openings is more useful than a single video with 200. It is also why a precise-looking decimal without a sample size deserves suspicion.
A quick sanity check
If a rate is 1%, you expect about one drop per 100 chests. Seeing zero in 100 is normal. Seeing zero in 1,000 is unlikely but possible. Before you call a rate wrong, compare the result with the range the sample can support.
A worked example with two sprites
Suppose you want two sprites: one at 1.5% and one at 0.03%. You have time for 200 chests this week. The 1.5% sprite has a 95.1% chance of appearing somewhere in those 200 chests. The 0.03% sprite has a 5.8% chance. Same effort, wildly different odds.
That does not mean the rare sprite is not worth chasing. It means the two goals belong in different plans. Put the 1.5% sprite in this week's plan and the 0.03% sprite in a long-term collection plan. Mixing them creates the feeling that the game is unfair when the math is simply doing its job.
Common mistakes
- Comparing per-chest rates across different chest types. A normal chest and a Sprite Chest are not the same draw.
- Using the average as a schedule. The mean number of chests is not the number you will personally need.
- Ignoring the date. A rate from three patches ago may no longer apply.
- Changing three variables at once. If you change route, session length and chest type together, you learn nothing.
A five-minute checklist before you trust a table
- Does the table name the chest type? A mixed count is not comparable.
- Does it show a date? A rate without a date can be a season old.
- Does it show a sample size? A precise number with no sample is a guess.
- Does it separate confirmed and unconfirmed entries? Mixing them inflates confidence.
- Does it explain the method? Community logs and datamined pools have different errors.
- Does it match your own experience? If not, log twenty sessions before you decide who is wrong.
If a table passes all six, use it. If it fails two or more, treat it as a rumor with a decimal point.
What to do next
- Pick one sprite you actually want.
- Write down the reported rate and the date you checked it.
- Choose a session size you can finish.
- Run the simulator and note the zero-drop share.
- Decide before you queue whether the session is worth it.
That is the entire method. It takes five minutes and replaces a week of arguing about averages.
Frequently asked questions
Are the rates on this site official?
No. They are labeled as reported third-party estimates. Verify in-game after a hotfix.
Why do two sites show different rates?
Different sample sizes, chest types and update dates produce different estimates. Treat both as ranges.
Does the simulator predict my next session?
No. It shows the spread of many possible sessions at the rate you enter.
What is a good session size?
One you can finish without compromising your normal route and extraction. A finished session produces more draws than an ambitious one that ends early.