opinion piece on discretization and multimodel averaging in ecological statistics
the goal is to do some simulations to show that the standard ways of doing MMA don't give adequate coverage (see also @dormann_model_2018 ...)
at this point all I really want to do is recreate Dormann et al's figure 4,
Hartig's BES talk: https://www.dropbox.com/s/icf0ls2cjzvh31n/HartigEAB2017.pdf?dl=0
private repo forked from Hartig: https://github.com/bbolker/lowpowerregression
Dormann et al: http://onlinelibrary.wiley.com/doi/10.1002/ecm.1309/full