Handling y-autorange outliers

[goodboy]

super annoying when the dark pool bois do this and you end up having a nasty awkward y-range..

I'm wondering if we can come up with some kinda sane heuristic for this?

• if large dispersion within a time step consider outlier?
• cross ref with whether the trade was reported as dark vlm?

[goodboy]

Ahh in this case it was a dark i think?