Pity
Gacha calculator
Ordinary weights: 1–5 stars use 45 / 30 / 17 / 6.5 / 1.5 percent. At 69 consecutive misses on the same banner, the next pull is guaranteed five-star. Enter consecutive misses since that banner's last five-star, not lifetime tickets spent.
Ten-pull condition: if the first nine results contain no four-star or better, the final slot draws only from four- and five-star weights. Its conditional five-star chance becomes 1.5 / (6.5 + 1.5) = 18.75%, unless hard pity takes priority. Thus 1 − 0.985^N is not the exact batch probability. A hundred-pull is ten ten-pulls.
A batch model tracks pity and whether the current ten-pull has a four-star-or-better result. A five-star resets that banner to zero. Switching banners preserves each independent counter; it does not erase progress on another banner.
Boundary checks: 69 prior misses + 1 pull must be 100%; zero future pulls must be 0%. The interactive batch model starts at a new ten-pull boundary and assumes the banner rarity pools are nonempty. Its single-pull mode is only a theoretical benchmark: the current UI has no ×1 button. Shiny rates, species selection, refunds and empty-pool fallback are outside this rarity model.
| Five-star | 1.5% |
|---|---|
| Pity | Pull 70 |
| Scope | One banner |
| Shiny | Not included |
Questions
Does switching banners reset pity?
No. Each generation has its own persistent counter. A five-star resets only the banner used for that pull.
Is every non-pity slot 1.5% five-star?
No. A ten-pull final-slot guarantee can restrict rarity to four/five stars, giving a conditional 18.75% five-star chance at that step.
Sources
Last checked 2026-10-27. Numbers above are omitted when a source did not state them.