I thought I'd do a quick comparison of how teams in the MAC have done versus how you'd expect them to do, based on the recruiting services. Three services rank the MAC in order, Rivals, 247, and Scout (ESPN does not). I took their ranking for each of the years 2008-2015 and averaged those, to get the ranking for each class for each team. Next, I took a weighted average. The 2102 team, for example, will include members of the recruiting classes from 2008-2012. The classes which impact the quality of the team the most would be the 2008-2009 classes, as those would be Seniors/4-th year Juniors, and the least impact would be the class of 2012, which would be true Freshmen.

Next I projected what the order of finish would have been for 2012-2014 if things played out as you would expect from the recruiting rankings, and then compared that to the actual finish. First, here is 2012, the first number is the projected finish, followed by the actual finish, and then the difference (with minus numbers meaning a worse than expected actual finish):

EAST
Miami 1, 4.5, -3.5
Kent 2.5, 1, +1.5
Akron 2.5, 7, -4.5
BG 4, 2, +2
Ohio 5, 3, +2
Buffalo 6, 4.5, +1.5
UMass 7, 6, +1

WEST
Toledo 1, 2.5, -1.5 (Projected MAC winner)
WMU 2, 5, -3
CMU 3, 4, -1
Ball St 4, 2.5, +1.5
NIU 5, 1, +4
EMU 6, 6, 0

Next, the same thing for 2013:
EAST
BG 1, 1, 0
Miami 2, 7, -5
Kent 3, 5, -2
Akron 4, 3.5, +.5
Ohio, 5, 3.5, +1.5
Buffalo 6, 2, +4
UMass 7, 6, +1

WEST
Toledo 1, 3.5, -2.5 Projected MAC Winner
WMU 2, 5.5, -3.5
CMU 3, 3.5, -.5
NIU 4, 1, +3
Ball St 5, 2, +3
EMU 6, 5.5, +.5

And for 2014:
EAST
BG 1, 1, 0
Miami 2, 6, -4
Kent 3, 7, -4
Ohio 4, 2, +2
Akron 5, 4.5, +.5
Buffalo 6, 3, +3
UMass 7, 4.5, +2.5

WEST
Toledo 1, 1.5, -.5 Projected MAC Winner
WMU 2, 3, -1
CMU 3,4, -1
NIU 4, 1.5, +2.5
Ball St 5, 5, 0
EMU 6, 6, 0

Now, looking only at the differences for the three years, you notice that some teams are consistently positive, others are consistently negative, and others are close to zero each year. Sorting the teams by the total differential:

NIU +9.5 (+4, +3, +2.5)
Buffalo +8.5 (+1.5, +4, +3)
Ohio +5.5 (+2, +1.5, +2)
Ball St +4.5 (+1.5, +3, +0)
UMass +4.5 (+1, +1, +2.5)
BG +2 (+2, 0, 0)
EMU +.5 (0, +.5, 0)
CMU -2.5 (-1, -.5, -1)
Akron -3.5 (-4.5, +.5, +.5)
Kent -4.5 (+1.5, -2, -4)
Toledo -4.5 (-1.5, -2.5, -.5)
WMU -7.5 (-3, -3.5, -1)
Miami -12.5 (-3.5, -5, -4)

Given that the errors do not seem to be random, and given that there is a horizontal consistency (some teams always positive, some teams always negative), it seems to me that there are several possible explanations for the positive/negative numbers:
1. The recruiting services consistently underestimate the classes from some schools, and overestimate them for other schools. This could happen for 3 reasons:
a. Some schools like NIU recruit areas that are not widely recruited (or covered, so they might be under-rated), such as the northern plains, while other schools, like Toledo, recruit in hotly contested areas (so they might be better covered, slightly over-rated).
b. Some schools may be better at finding hidden gems than others. Schools that uncover more hidden gems would perform better than the ratings might indicate.
c. All schools probably have some players that they announce as a part of their class, but who end up being no-shows, or who leave the program, but this may happen more at some schools than others. Schools with a lot of no-shows or departures would perform worse than expected.
2. Perhaps the coaches at some schools coach their players up (giving plus numbers), while other schools don't get the maximum potential from their players (giving negative numbers). If coaching is the explanation, you'd expect to see a shift in the numbers when the coaches change. You do see a shift upwards in Bowden's second year, and in Fleck's second year, plus a drop at Kent after Darrell Hazel left, and a jump in UMass with their new coach, so coaching seems to be at least part of the differential we see.

FWIW, here are the projections, using the same method for 2015:

EAST
BG 1
Miami 2
Ohio 3
Kent 4
Buffalo 5
Akron 6
UMass 7

WEST
Toledo 1 (Projected MAC Winner, as always)
CMU 2
WMU 3
NIU 4
Ball State 5
EMU 6

Now, for grins, let's apply the average error, since they seem pretty consistent:

EAST
BG .3 (Projected MAC Winner)
Ohio 1.2
Buffalo 2.2
UMass 5.5
Kent 5.5
Miami 6.2
Akron 7.2

WEST
NIU .8
Toledo 2.5
CMU 2.8
Ball State 3.5
WMU 5.5
EMU 5.8

This is a lot more reasonable order, I think, though Buffalo and CMU will have new coaches, so this won't predict them accurately, and WMU will certainly be higher than that.

Very interesting analysis, L.C. This confirms what I've always heard about the Nebraska program from which Frank comes -- masters at coaching up players who might not be sought by their peers. Of course there is an interesting interplay between coaching and recruiting here because you have to be able to identify the players who are coachable. It seems to me, with a few notable exceptions (particularly early on), Frank has been very good at that at OHIO.

I understand that, and I think you did an admirable job of making the case that player rankings from the rating services when viewed in the aggregate have some predictive validity and also that some schools perform better -- or worse -- than would be expected solely from analyzing the ratings. However, it is also true that if you look at certain specific coaches and programs you can see those who do a better job at "coaching up" lesser ranked players, those who recruit better than they coach, and those who both recruit well and coach well -- with gradations in between.

As an aside, there are also coaches who simply do better with lower ranked players who want to improve and have a good work ethic. Switching sports for a minute, it was a real irony back in the Jim Snyder basketball era that Coach Snyder coached a team to the Elite 8 in the mid-1960s and had multiple victories over B1G teams in the late 1960s with "diamonds in the rough" kinds of players but when he finally reached the point where he started to recruit what would today be 4- and 5-star rated athletes that he falter and could not coach the prima donna athlete.

Two who come to mind right away Casper71: Ulice Payne and Larry Slappy.

That's exactly my point, oldkatz. Snyder, I believe, retired when he did because of players like these. They were very talented, but they were also hard to coach and didn't take to his coaching style. On paper, his last few recruiting class were his best ever, but it didn't translate on the court because of a mismatch of coaching style to player ego. I know at the time that's what some "insiders" told me, and it seemed to make sense in terms of what I was able to observe.

Some of the schools in the MAC have issues. Miami has always turned its nose up at partial-qualifiers. Akron is a fly over for the good MAC recruits in Ohio and EMU the same in Michigan. Almost always do recruits prefer BG, Toledo, CMU or WMU over those schools. Buffalo and NIU have more lightly covered recruiting areas. Its really difficult to compare different states in recruiting side by side. Ohio thinks more about upside and development so the MAC level players they pick don't always correlate with the highest rated MAC level players. Mostly I believe its better to sit in the stands on Saturday and judge the talent by what you see on the field. Or tabulate the amount of players a school has put in the NFL. Ultimately though with Athens being one of the smaller college towns in college football and not real close to Florida like an Auburn or Clemson, nor land grant like Virginia Tech makes it difficult to build up talent. The same issue in basketball. There is nothing stopping the school from having great facilities and playing strong schedules to put its best foot forward which we currently don't have.