Bill Polian, Hall of Fame GM turned ESPN analyst, apparently has a list that he uses to tell if a team is a Super Bowl Contender. There are a lot of ways a list like this could go, and while most of those ways are bad, some aspects of his list are so creatively, mind-boggingly bad that it's worth taking a closer look at. It's an ambitious combination of faux-statistical competency and selling out. It's like an Adam Sandler football movie died and was cremated and the ashes were inserted into a haunted etch-a-sketch that spontaneously created a list of winning football indicators.
Here's his list:
Let's dive into the subtle, aged wisdom of one of the NFL's most celebrated minds in bullet-point form:
1. Yards Per Pass Attempt
This is fine. Yards per Pass Attempt is actually a very good indicator of quarterback competency, and most statistically adept people are not afraid to mention it. In combination with completion percentage, it actually tells you most of the important information you might want to know about a quarterback, and doesn't overly complicate matters. It's much better than certain corporate-sponsored attempts to quantify NFL QB play in one number invented by certain "world wide leaders" to trot out on their yelling shows.
What's really striking about this list though, at the macro-scale, is just how clever it tries - and fails - to be. What makes an NFL team championship-caliber isn't exactly rocket science. Most of that information is captured by "points scored" and "points allowed". If you wanted to break it down a bit further you would want to see a good-to-great quarterback and a good-to-great defense, and if you wanted to be extra clever you could look for more complicated factors likely to regress to the mean to find hidden gems. Llike 3rd down completion percentage, for instance, often regresses to the mean established by your 1st and 2nd down completion percentage. If you wanted to be clever. We're talking about ESPN, so I understand.
2. Points Allowed per Game
You know what subtle indicator of greatness I look for in a champion?
THE SUM TOTAL OF THEIR ENTIRE DEFENSE. BET YOU NEVER THOUGHT OF THAT!
3. Turnover Margin
Having more possessions, and therefore more opportunities to score than your opponent is a good indicator of success. The problem with turnover margin is that turnovers are highly luck dependent. Yes, good defenses can and do produce more turnovers, and yes, some quarterbacks are going to reliably throw more interceptions than others, but fumbles, and especially fumble recoveries, are mostly random, and vary widely from season to season. Turnover margin is in general explanatory, but not predictive. At least not all that predictive. This is just another "duh" entry so let's move on to the meat.
4. Kicking Efficiency (Field Goals)
Screw you punters! And, I guess, kickoff specialists! Last year we finally got the Steve Gostkowski v. Steve Hauschka Super Bowl we've long waited for. It's really a shame that Janikowski has been trapped on the Raiders his entire career. Just one Super Bowl for the Dan Marino of kickers. I'll still roll with Adam Vinatieri though, just based on kicker winz.
5. Third Down Efficiency
We again vacillate between captain obvious and crazytown. Yes, teams who convert a lot of third downs will be successful. Those teams tend to have good offenses, and simply identifying a good offense will be more predictive of future success than looking at 3rd down efficiency, which can actually get out of whack from time to time. This is basically like turnover margin. It's great to be good at it, and you already know this. But we've been putting off the fun for far too long. Let's move on to the punchline.
Yes! QBR! ESPN's handmade nonsense statistic. I'll bet that if Bill Polian were sitting in the front office at Colts' HQ and some intern brought him in a binder with QBs listed by QBR that he would immediately throw that binder and intern into a Lucas Oil well. You can bet the league is dreading the Jets right now. Top ten in points allowed and turnover margin, and they have the number 7 quarterback by QBR. Watch out NFL playoffs, here comes Ryan Fitzpatrick, maybe, if they can catch a few teams! By gum, if Nick Folk were a bit better they might not be out of the current playoff picture!
Most of what you need to know about QBR can be gleaned from the fact that in this game, Charlie Batch posted the greatest QBR in history.
And in this gameTim Tebow posted the 2nd highest post-season QBR in history.
QBR also "clutch-weights" everything including running, and penalties! That 12-men call sure was clutch! Bonus points for that clutch penalty!
If anyone can tell me why Matt Flynn had a lower QBR in this game than Matt Stafford, please do chime in.
"Clutch" as a statistical concept is always a rough idea. Yes, some players play well in close and late situations, but they are generally the same players who play well all the time. Tom Brady is awesome in the 4th quarter...and the 1st quarter, and 2nd quarter, and 3rd quarter. I talked to some people on twitter who told me the clutch rating mainly serves to discount garbage time, but the other big problem with QBR is it's a black box. When does "garbage time" start? Isn't there value to running up the score to make a comeback impossible? Should you do a clutch adjustment for something as tangentially related to player performance as a penalty? Polian's first point on the list is Yards per Pass Attempt. Here are the top 5 in that, and in QBR:
The top five quarterbacks in Yards per Attempt are:
The top 5 by QBR are:
So what exactly is QBR giving you that you don't already have? Simple; the ESPN name. ESPN created this monstrosity, and if you're on ESPN you have to use the house stats. The fraud is on clear display by the inclusion of QBR and YPA on the same list. The house told Bill Polian that they ordered too much Shiraz and he's out there pushing it on every unsuspecting First Take-watching troglod...I mean, Insider. This was on something called NFL Insiders!
The fact is that QBR is a poorly conceived, entirely secret grab-bag of NFL talking points designed specifically to allow morons like Skip Bayless to "credibly" talk about Aaron Rodgers and Tim Tebow in the same sentence while simultaneously having a number on the screen. The next time they are seriously used in an NFL front office will almost certainly be the first, and I'm betting it would happen in either Cleveland or Detroit. They may as well call it "Subway Presents QBR".