Sunday, November 21, 2010

RR v Harbaugh: Letting History Predict The Future

Though I don't believe RR will be fired at the end of the season, I keep hearing that if he is the program will take a huge step backward. I've never really believed this, so I decided to use historical numbers (very limited) and see what we can predict using the past. 

The data is very limited, of course, but it is what it is. I'm only using I-A numbers because a) It doesn't take as long and b) using div II for RR and I-AA for Jim Harbaugh weakens the argument (but not by too much)

RR has had two new HC jobs in I-A
-At WVU, he took a 7-5 team and made them 3-8, a winning % decrease of .208
-At M, he took a 9-4 team, and made them 3-9, a winning % decrease of .442
So, RRs average is to decrease a team's winning % by .325 (nearly 4 games in a 12 game season).
It is not surprising people are afraid of what a change could bring. RR has scared us.  

Let's take a look at what has happened with Harbaugh:
-At Stanford, he took a 1-11 team and made them 4-8, an increase of .242.
So, Harbaugh's average in his first season is to improve the team by .242 (just under 3 games in a 12 game season).

Adjusting for a 13 game season, and assuming a loss to OSU next week, but a win in the bowl, hiring Harbaugh would produce about a 11-2 season next year.

Another thing you have to consider, of course, is what happens to RR teams after he leaves.
-At WVU, in his last season they went 11-2 (I'm giving him credit for the bowl victory, even though he wasn't officially there when WVU beat ASU 48-28).  The next season, WVU went 9-4, a decrease of .153 (or 2 games).
If M fires RR and hires any random coach, and assuming the w/l as above, you can expect a 6-7 season next year.  

So 11-2 with Harbaugh, 6-7 with RR leaving, it gives us an average of 9-4-ish

But won't RR get that anyhow? Well, let's see how RR does between season 3 and 4 historically.
-At WVU, he went from 8-5 his 3rd season, to 8-4 his 4th season. That increase of .051 is worth about 2/3 of a game in a 13 game schedule. So if we keep RR, making the w/l assumptions above, M would go 9-4.  It is  a push for next year.

Now, what about the future.  How does Harbaugh do from season 1 to 2, and how does RR do from season 4 to 5.
For RR, season 4 to 5 is an increase from 8-4 to 11-1, that increase of .25 puts M at 12-1 in 2012.  I didn't put in a Big Ten Championship game, but it definitely could be possible.  BCS Bowl, too
For Harbaugh, season 1 to 2 is an increase from 4-8 to 5-7,  that increase of .083 put M at 10-3 in 2012

How about from season 5 to 6 for RR and 2-3 for Harbaugh
-RR 11-2 to 10-2, -.128 would put M between 9 and 10 wins.
-Harbaugh 5-7 to 8-5, .198 would put M between 12 and 13 wins in 2013.  Still using 13 game season, possible National Championship game, it looks like

So, it looks like M has the best shot at first getting a BCS bowl and maybe a conference championship in 2012 if RR is kept.  
But, the best shot at a National Championship is in 2013 if he is replaced by Harbaugh. 

I really expected it to look better for hiring Harbaugh.  Either way, I guess we are heading toward a BCS Bowl game in the next few years.  


  1. I had said before that 6-6, Michigan would have Fred Jackson coaching the bowl game. I think 7-5 probably saves RR's job.

    But that was before the way the Wisconsin game felt. He's potentially lucky that the OSU game is in Columbus this year.

    If Harbaugh would come, I'd hire him. But I wouldn't fire RR without having a sure thing ready to take over.

  2. I agree completely on your last part. Not having someone waiting in the wings would mean probably not getting anyone good.