> You also almost double your odds of success by not using violence.
Admittedly having not read the 400-page study, I don't think that's a causation that is necessarily supported by the correlation. It would be extremely surprising if the prior of "how likely is this movement to succeed" were not a determining factor in whether a movement tends to use violence, with the a priori less-promising movements being more likely to take violent action.
C.F. the difference between me demanding you give me an apple or your car.
Yea, causation here depends on a whole lot.
If you're movement is going to 100% cause a reaction of violence with the opposition regardless if you're violent or not, then there is zero reason for your movement not to use violence themselves. Simply put, you'd be rounded up and exterminated simply for existing.
> Admittedly having not read the 400-page study […]
It is not a 400-page study: it is a 400-page book that goes over the research available at the time and summarizes it. The book leans slightly academic, but it's a fairly easy read.
A movement's success is (partly?) determined by its size and how much of the general population gets on board with the original (presumably) small group that started it.
* https://en.wikipedia.org/wiki/3.5%25_rule
The/A thesis of the author is that people are turned off by the use of violence/force and are less likely to agree with, and/or get involved in, movements that use violence.
So if a movement wants to grow the 'coalition' of people that will help and/or join them, that growth is best achieved by eschewing violence as much as possible.
[Edited GP post to add some of this comment.]