Looking for a second opinion on whether over fitting is being used in this scenario

You take some thing that happens in markets, we'll call it "S", you collect a significant number of occurrences of S over time. Once you have a large sample size of S, you being analysing the results of these occurrences.

Would it be curve fitting in this scenario if one were to look at the way markets react when S happens and analyze patterns within the pattern...such as : when S happens price moves within 1 ATR against before moving 2 ATR in favor and this occurs 70% of the time. Otherwise price moves more than ATR against.

Hope this made sense. Thanks