It would be unfair for me to comment on the frequency of ADs regarding striker-fired guns generally. I can only comment on our experience as it relates to Glocks. But that was our experience and I would caution against extrapolation beyond the 400 men and women of our agency.
In fairness to Glock, you must keep in mind that over 2/3s of the police agencies in the U S issue Glock pistols to their officers. That said, it would figure that if an AD occurs odds are that it happened with a Glock. The data on its face will be skewed against Glock unless the necessary mathematical calculations are made across all police officers carrying whatever striker-fired gun makes carried.
I can tell you that our records showed very few ADs when we carried revolvers.
I was assigned the investigations into both contact and non-contact police shootings all involving Glock pistols and after obtaining sworn statements as part of our shooting review board of those officers involved it became clear that some genuinely decided not to fire but finger on trigger with trigger press commenced caused a round to discharge that was unmeaning. Also, enter the phenomena of contagion firing. Here again, one shot fired, by bad guy or fellow officer, seemed to produce a fusillade of gun fire when it wasn't really meant.
Again, based upon my own anecdotal experience, I came to believe, and still do, that the striker-fired operation is not well-suited to police work. This could be caused by our own training failures, and any number of other intangibles. Clearly, more agencies (police chiefs) disagree with me. But that's just me. Also, since we transitioned to the DA/SA P226 the number of ADs plummeted dramatically. This, to include self-inflicted gunshot wounds to the leg, foot, and buttocks. Especially when re-holstering.