If this ship stays on course, Michigan is due to collide with the 3-3-5 iceberg in the form of Jay Hopson for the 2009 season. For many, this is seen as a problem, and their argument is a legitimate one (see SCM's post immediately below this one). For starters, the 3-3-5 does indeed take away a defensive lineman, which for this year's team was certainly our strength. Then there was that Purdue debacle, which we probably would have won if we would have just stayed with our crappy cover-2 4-3. There's also the fact that this is a new defense that the majority of people say just won't work. This is by far the dumbest of the reasons, but there it is anyways.
This is why I'm not afraid of the 3-3-5:
Compare this to UM's defense over the same time period:
The majority of the time, West Virginia's 3-3-5 stack was statistically better than Michigan's defense. To add to that, they were doing it at West Virginia where 5 star linemen and cornerbacks don't play. I will admit that the Big East offenses don't exactly scare the daylights out of anyone, but um, which Big Ten offenses do (sit down, Penn State, we all freakin' see you).
This, to me, is the bottom line. The 3-3-5 defense will work and, given the time we're willing to give Rodriguez's offense to work, shouldn't we give him the fair shake on defense as well? (hint: yes)
This past year, our best unit was most certainly the defensive line, but what did that get us? A terrible defense because our linebackers and safeties couldn't be counted on. Sure, we were marginally better when we went to 4 down linemen, but it didn't exactly win us a ton of additional games. If, next year, we want to wholesale run the 3-3-5, how much worse could it be? If the linebackers have improved at all, my guess is it will be better. My point is this: no matter the scheme that you run, you'd better have the athletes to run it. If Rodriguez thinks he can get the athletes to run the 3-3-5 and that he will be most successful in that base defense, then he needs to do it. This scheme is new, but it is proven. It will work, provided the right personel run it.