After reading the article this past week about Bill Squires and his Boston group, I decided to buy his book to see what kind of training his group did.
I liked the structure of the program and would even consider trying it in the future, but I noticed something strange about the marathon training/workouts. There is only one day that specifies running marathon pace, and it is only 6 miles long during a 10 mile run about 10 days out from the race. Other than a few longer races (15k, half marathon) and surges during the 20 mile weekly long run, the workouts are mainly intervals and hill repeats. I have always believed that it wasn't good to run your goal pace into the ground, but in this training plan marathon pace is non-existent.
It looks as though this is accurate as I briefly skimmed through Bill Rodger's log (http://www.bunnhill.com/BobHodge/Rodgers/rodgers75log.htm) and he seems to have raced himself into shape and did not do any specific workouts at marathon pace.
This also goes along with Frank Shorter's training philosophy that he trained for the 5k all year round, focusing on interval workouts, hill repeats, and fartleks, and did not believe in tempo runs, just easy high mileage. He also believed in the surges during runs described in Speed with Endurance to teach the body to respond to moves in a race.
So the training method that seemed to be popular in the 70s and 80s is that you don't have to train at marathon pace to run a marathon, which seems to be very counter-intuitive to what many people do today. I was curious about what other people thought of this as I was a bit surprised when I read this. Maybe it makes sense that the longer workouts elite runners do now-a-days (e.g. Ritz's 20 miler, Hall's 15 miler) don't benefit a particular system and in the end just take away from their peak performances later on.