A percentage without a precise event, deadline and model is not actionable evidence. “Profit,” “touch” and “hit my target” are different outcomes.
Define the event first
Probability of profit may mean positive value at expiration under a pricing model. Probability of touch concerns whether price reaches a level at any time before a deadline. Target-hit probability should refer to the chance that the whole position reaches a specified net P/L after costs under an explicit exit rule.
Those quantities differ because a price can touch a strike and reverse, or a stock can finish favorably while the option position loses after premium, time and volatility effects.
Model probability is not a win rate
A theoretical probability derived from current inputs is conditional on assumptions. A historical win rate summarizes a particular sample and rule set. Neither becomes a calibrated forecast merely because it is displayed to one decimal place. Calibration requires enough forward observations and comparison of predicted ranges with actual outcomes.
A platform should preserve the original prediction, timestamp, contracts, entry assumptions and outcome definition. Otherwise later edits can create hindsight bias.
Questions to ask when you see 70%
Ask: 70% of what, by which date, before or at expiration, before or after costs, using which price and volatility model, and validated on what independent sample? If those answers are missing, treat the number as a scenario-model output rather than a trading signal.