Loading results…
Simulation Settings & How to Read the Results
Each score comes from simulating the full kit together against a matched reference. Higher mean bashin indicates a larger simulated gap. Event gain is the difference between the all-events and core-kit means; it can be negative.
The event toggle assumes every listed character-specific event skill is learned, including direct grants. Skills with a double-circle upgrade are evaluated at that level. Dimmed skills are inactive for this race or excluded from the selected kit; other skills retain their normal activation rolls and dynamic conditions. Evolution skills, support-card skills, and inherited uniques are excluded except for the Christmas Oguri recovery pool described below.
Native aptitudes determine the default list. Every simulation uses trained distance S, surface A, and style A, including rows revealed by the aptitude override. Training growth bonuses do not change the shared stats.
Christmas Oguri uses Ignited Spirit STA plus two recoveries from Global white skills and inherited U=ma2, Super-Duper Climax, or Superior Heal. Ignited Spirit STA can also be the intended trigger. Inherited skills with only a position requirement may compete as triggers; other triggers must have no situational requirements. The intended trigger must be third in at least 85% of successful activations, with each setup recovery activating after it at most 10% of the time when both activate. For each trigger, setup recoveries prioritize fewer situational requirements before mean bashin. The best qualifying triggers then compete on measured mean. The chosen trio replaces her innate/event recoveries; optional insurance does not change the ranking score. Insurance must have no situational requirements, rescue failed activations, improve paired mean with a nonnegative 95% confidence lower bound, and preempt at most 10% of otherwise successful activations.
For other Uma, skill-count prerequisites are assumed fulfilled, following the CM skill evaluator. Christmas Oguri's recovery count and halfway requirement are checked during the race, with independently sampled activation rolls and timing. Position-condition approximations remain and are labeled. Her recovery search models Soft Step in the first half of mid-race with normal failed rolls; the standalone Visualizer retains its original condition. Christmas Oguri's recovery optimization excludes opponent debuffs while retaining self-benefits. Other Uma currently retain the existing opponent-slowdown estimate; these scores are not CM win probabilities. Skill lane movement is disabled, so navigation-only skills may contribute zero.