As the folks in Blitzortung forum subsection know, there are a lot of thunderous flashes that don't show up on the detectors, partly due to not enough stations sensing the stroke to validate it, but also only cloud to ground get counted I guess.
Which got me wondering. It seems that a lot of intracloud strokes happen while watching a storm, and fewer cloud to ground.
Is there a rough ratio, or is it extremely variable? I've watched some smaller cells drift by with constant flashing in the cloud, while no detectable ground strokes occur.