I think there is some good signal in differentiating stocks based on how much volume is in the continuous book vs auctions.
We don't have this data. Do you?
If you have this data can you share a sample dataset? We are working on ways to make it very simple to use your own data. For example we are adding a Snowflake connector to seamlessly import your own stock factors.
No I don't, but I believe FactSet does. In their historical prices database I imagine the opening auction print has a special flag.