A training session I haven't used in months but think is really cool is what is called Yasso 800s. It's supposedly a killer way to estimate your marathon time, but the way it works is a bit odd.
Here's the logic, if there is any: you run 8-10 800 meter intervals on the track in some time, say 3:30 each, with a 3:30 jog between the intervals. Sounds okay. The strange part is that the thesis says if you can do this workout you should be able to run a 3 hour 30 minute (3:30) marathon. Where's the connection between minutes to run a series of half miles and the number of hours/minutes it takes to run a marathon? I don't see the logic. But what about the practice?
I put it to the test last year once, and it worked out fairly closely. I set out to do the 800s in 3:40, accomplished that without too much difficulty, and then "surprised" myself with a 3:38 marathon. Seems like the Yasso theory works.
I do have some caveats, though. It seems like it would work "down", but not "up". In other words, if you can do the workout the marathon time should take care of itself. But one may not have the sheer foot speed to make it work the other way: a 3:30 marathoner may not be able to run the Yasso intervals in the requisite time. No proof on my part, just thoughts.
Another caveat is that "faster" runners, those with simply more foot speed like middle distance runners as opposed to truer marathoners, have a built-in advantage. The Yasso 800s are built for a guy who can run fast but also run a marathon since it feels so much slower for so much of the race. I think I fall into this category, although my actual marathon results a few years ago seem to refute this.
I'm not convinced any of this matters, but it is fun to think about. Not much else in the brain at mile 9 now is there?
Today's workout was 30 minutes on the bike and 20 minutes of upper body resistance training. Ho hum. I doubt there will be much of anything tomorrow since I am working from home and the weather will be very cold and very wet. Bummer.