One thing we have in common here is that we are both not letting results color our thoughts on process. Don't know where you got that for me.
Rcal already argued my point for me better. If that play works 35% of the time, it's a good send. You're leaving lot of expected value on the table by being so risk-averse in your process. 4th down game theory is a really good analogy for this, it's basically the same thing