... I had figured out a really elegant way to test an hypothesis, complete with two separate treatment groups and a control group. While the population under study was blog readers, I had come up with a reasonable plan to protect the human subjects, even mentally drafting the IRB short form.
I was very excited at how well it was all coming together. And then I woke up.
Which means I have no earthly recollection of either the hypothesis or the clever strategy for testing it.