Apropos of the Falcon Heavy test flight scheduled for February 6, 2018, what is the failure rate of first launches of new orbital rocket designs?
How has this rate changed over time since the 1950s?
Ideally, I'd like a breakdown by decades, but overall trends would be enough answer for me.
Defining rigidly what constitutes a "new" rocket design is, of course, impossible; I'd say that all the iterations of the Falcon 9 1.1 are "one design" for purposes of this question, but I can almost see a case for distinguishing between Falcon 9 1.0 and 1.1 because of the thrust structure redesign.
I'd define "failure" as (1) not achieving insertion into the desired orbit for any reason after liftoff or (2) catastrophic failure to lift off, but if you want to distinguish other cases, feel free. If the rocket ignites its engines, then immediately shuts down in a way that leaves it able to try again in a matter of days, that's not a failure.