To be fair, I don't think Mangini is a bad coach. I think he's at least as qualified as the average NFL head coach, and maybe he's even a little bit better. I only saw him actively mess up game management once (week 16 this year at Seattle). He seems to be a non-nonsense guy who doesn't play favorites. But his record also seems to be pretty closely tied to personnel changes (i.e., it's not clear that he's making the sum greater than its parts), and he's a member of the Belichick coaching tree, whose members are 71-90 as NFL head coaches. Plus, the thought of going with another former Belichick defensive coordinator as head coach (like Romeo Crennel) and another Ravens personnel guy as GM (like Phil Savage) has me feeling a little nauseous.
So I asked my friend Tim to talk me down off the ledge. Here was his resposne:
Here's the deal, of all available coaching candidates, we have ruled out college coaches (see Butch Davis) and hot NFL coordinators (see Romeo Crennel). That being said we are looking at retreads. I think Cowher is out of play this season, which leaves the most attractive candidates as Schottenheimer, Shanahan and Mangini. Schott doesn't even appear to be an option, I think he might really be retired (he hasn't been mentioned anywhere). Shanahan is on vacation and won't interview anywhere for the next couple of weeks. Lerner may wait for him to see where he stands but the guy has had total control in Denver for a while and he hasn't won a thing without John Elway. I don't think he is the slam dunk everyone is making him out to be. Which leads us to Mangini, a guy who has had two winning seasons in his first three years as a head coach.That got me to thinking. Unless an NFL owner intends to hire someone straight out of high school or the arena league, or someone who has never coached football at all before, then there are only three types of NFL head coaches, and Tim mentioned them all -- college head coaches, NFL assistants, and retreads. To categorically exclude two of the three strikes me as more than a little foolish.
Let's start with college head coaches. Only three current NFL head coaches were college head coaches -- Tony Sparano (New Haven (D-II), 1994-1999), Tom Cable (Idaho State, 2000-2003), and Tom Coughlin (Boston College, 1991-1993). Sparano and Cable spent significant a significant amount of time as NFL assistants following their departure from the college ranks. Coughlin went directly from BC to Jacksonville, where he coached for eight seasons, and then was out of football for one year before taking over the Giants. Based on this, it seems like a case could be made for staying away from college coaches. Sparano and Coughlin have had success, but Sparano was nearly a decade removed from his college job, and Coughlin was in a unique situation when he took over the expansion Jaguars. Cable is an interim this year, and perhaps shouldn't be in this discussion at all.
That leaves us with NFL assistants and, as Tim called them, NFL retreads. I have compiled some numbers on the retreads' records. Click on the picture below for a better look.
There are nine current retreads: Coughlin, Belichick, Dick Jauron, Wade Phillips, Tony Dungy, Jon Gruden, Herm Edwards, Norv Turner, and Jim Haslett. (Although Haslett is an interim, I've included him because he coached more than half of his team's games this year.) Phillips and Turner are on their third teams with full-time gigs. I did not include their interim terms because they were five games or less.
In their first stops as NFL head coaches, these guys were a combined 380-394, a .497 winning percentage. In their subsequent posts (i.e., including Phillips and Turner's third teams as well as everyone's second), they were a combined 408-292, a .583 winning percentage.
But wait a tick. Belichick and Dungy have been absolutely phenomenal in their second go-rounds as head coaches, combining for a 187-69 record, or .730 winning percentage, during their time with the Patriots and Colts. If you take these guys out of the equation, the group of retreads is 221-223 after their first jobs, a .498 winning percentage, almost exactly the same as the win rate for their first jobs. (To be fair, taking Belichick and Dungy out of the tally for first jobs lowers that winning percentage to .493, which means the other seven guys increased their win percentage by .005, an average of one additional win every 12.5 seasons.)
Looking at individual changes, besides Belichick and Dungy only Wade Phillips saw a large uptick in his winning percentage after his first job. Oddly enough, some people think Phillips may still lose his job before next year, and I think I'd get laughed out of town of I suggested that the Browns hire Phillips.
Here are the final breakdowns in terms of increased wins or losses for coaches after their first job:
Belichick - Extra 4.1 wins per season
Dungy - Extra 3.1 wins per season
Phillips - Extra 2.2 wins per season
Coughlin - Extra 0.9 wins per season
Jauron - No change
Turner - Extra 0.25 losses per sesaon
Gruden - Extra 1.3 losses per season
Edwards - Extra 2.8 losses per season
Haslett - Extra 4.8 losses per season
Average for everyone - Extra 1.4 wins per season
Average for everyone not named Belichick or Dungy - Extra 0.07 wins per season, or about one win every 12.5 years.
The point is this: with only Belichick and Dungy as exceptions, coaches in their next job perform at about the rate as they did in their first job. Mangini averaged just under eight wins a year. The Browns are in a division where everything Pittsburgh or Baltimore touches turns to gold. Eight wins a year, with a ninth win in 2021, isn't going to cut it.
By going with Mangini, Randy Lerner is betting that he's the next Belichick or Dungy, rather than the next everyone else. I hope to hell he's right, but I kind of doubt it.