Bill Gray said:

[talking about Orlando gun training]

Anyway, during the period after the training when the rapes dropped so
dramatically, none of the women ever fired their weapons. Moreover, none
of them shot a husband or boyfriend, nor were any children harmed with
these firearms.

I personally called the Orlando PD to confirm as much of this as I could,
and it checks out with their records department. There was an increase
in homicide the same year, but the base numbers were so low that even one
or two additional would come out to the 22%–in short, it wasn’t like
they dropped 122 compared to the previous 100, but more like 17 instead
of the previous 14. It was not considered statistically significant by
the statistician I spoke to at the Orlando PD anyway.

On the basis of what is known, it is hard to see a down side of the
Orlando experience. It is pretty clear–as clear as these things ever
get–that the well-publicized ability of a small percentage of the
potential victims to repel assault provided protection to nearly all
of that class of potential victims. It is also clear that there was no
societal cost–aside from the trivial expense to the police of providing
the training (which may have been offset by fees charged–I do not know
if the classes were gratis or not); certainly the increase in homicide
was not related to the women being armed.

To summarize your argument:

  1. the increase in homicides was not statistically significant — the
    actual number of homicides was small enough that increases of 20% or
    more were not uncommon.

  2. None of the women killed anyone.

Hence you conclude the increase in homicides was unrelated to the
training. Fair enough. I agree with you.

However,

  1. the decrease in reported rapes was not statistically significant –
    the rape rate fluctuated wildly in the years before the training
    program. It even decreased 100% one year.

  2. None of the women used her gun to defend against rape.

So why are you so sure that the gun training caused the reduction in
the rape rate?

Certainly, it is possible that some criminals were so scared by the news of
armed victims that they gave up rape for a year or more, so the
reduction in rapes, though not statistically significant might have
been caused by the training.

It is also possible that some unarmed criminals were so scared by the
news of armed victims that they armed themselves and ended up shooting
someone, so the increase in homicides, though not statistically
significant might have been caused by the training.