I imagine this has been covered somewhere already, but the search engine is down for me right now. Besides, I can just imaging trying to narrow down a topic with keywords "cFIREsim" and "4% rule" on these forums. :)
I have been running and re-running lots of FIRE calculations on my multiple spreadsheets, as well as playing around with various tweaks of my scenarios in the cFIREsim calculator. (BTW, I love that tool.) The observation I consistently make is that the 4% rule seems to be more conservative than running real simulations. For all of the hand-wringing that can happen here about how "this time is different" and "I need a fuzzy security blanket to sleep at night" and therefore only 3.5% or 3% will do, I am finding myself thinking that I would be comfortable pulling the plug when cFIREsim gives me an ~80% success rate, which is a smaller stash than the 4% rule would indicate. Has anyone else observed that and what are your thoughts?