Average is not a promise
A single probability number hides the shape of luck. At a rare rate, most sessions end with nothing and a few sessions carry the average. The simulator shows that split. Set the number of chests you can realistically open in one session, the reported rate, and the number of sessions to simulate.
Reading the result
- Zero-drop share: how often a session ends empty.
- Median: the middle session; half do better, half do worse.
- Best and worst: the range across the runs.
If the zero-drop share is high, a short session is a lottery ticket, not a plan. If it is low, the grind is predictable enough to schedule.
What the model assumes
Each chest is independent and the rate stays constant. Real games can change chest pools, add pity systems, or rotate items. The simulator is a thinking tool, not a prediction of your next session.
Frequently asked questions
Is the simulation random?
Yes. Each virtual chest is an independent random draw at the rate you enter.
Does it predict my next run?
No. It shows the spread of many possible sessions so you can judge risk.
Which rate should I use?
Use a reported rate from the sprite database and remember that hotfixes can change it.
Why does the result change each time?
Each run uses new random draws. Large session counts produce more stable summaries.
Can I simulate a pity system?
No. The model assumes independent chests with a constant rate.