Probability is not a promise
A 90% result still leaves a 10% miss path unless your budget reaches the modeled hard-pity boundary. Use percentiles as planning targets, not guarantees.
Enter your target rate, current pity and Trait Crystal budget to see success chance, expected rolls, median, custom confidence and worst case.
Trait Crystal lab
Choose a reported trait preset or enter your own rate. Presets are community-listed and can be overridden if the in-game UI shows something different.
A 90% result still leaves a 10% miss path unless your budget reaches the modeled hard-pity boundary. Use percentiles as planning targets, not guarantees.
Damage and SPA traits can be tested in the raw DPS matrix. Economy, range and control value need role context from the trait and tier pages.
It uses one minus the probability of missing every roll, then caps expected and percentile budgets at the remaining hard-pity threshold when you enable pity.
At the community-listed 0.10% rate with a 1,500-roll modeled guarantee and zero pity progress, the capped expectation is about 777 rolls.
Current pity changes how close you are to the modeled guarantee. Before that guarantee, the natural per-roll chance remains the rate you entered.
No. Launch presets are publicly reported and can be disputed. Forsaken and Draconic are explicitly flagged, and every chance or pity field is editable.