Happy Friday, friends. I wanted to share some thoughts today about the coaching “carousel.” First, though, two questions –
1. What lame-ass reporter coined that term?
2. Does he get all proud every time somebody uses it, and tell his grandkids “Your clever old pop came up with that?”
Anyway, some vacancies have been filled, and others remain open, so I thought I’d weigh in generally on coaching hires, and how we tend to think and talk about them, which is to say, stupidly. As an accompaniment to this article, I suggest you re-read this article I wrote back in July about the nature of coaching football in the NFL. Peruse it, and then come back to this article, after the jump. We’ll wait.
Okay, welcome back. Something has been bothering me about the 2014 carousel, such as it is. It’s something that I’ve seen in the comments from smart and well-meaning people here, and I’ve seen it from guys like Mike Silver (a thin-skinned toolbag) too. Really, if you look hard enough, you can see it and hear it anywhere football is discussed. It's just about universal.
In short, what has been bothering me is the level of certainty that people have about their opinions of the quality of coaches. They’re sure that they have the measure of a coach, from past history, and from what he says in press conferences, and that they can predict the future from that measure.
Problematically, though, they always cite evidence which isn’t really evidence; at best, it’s a small part of a larger fact pattern. Most of the rest of those facts are unattainable to the average fan or reporter.
A notion somehow exists that people who aren’t in the team’s building, and who, in any case, don’t really know what they’re looking at, can adequately evaluate the performance of a coach. I’m here to tell you, you can’t. Media people and fans do it all the time, though, to the detriment of teams. They ratchet up the pressure on owners, and that causes some unnecessary and unproductive churn.
Here’s a thought exercise – you’re a fan, and I’ve asked you to make a judgment on whether a head coach should keep his job or be fired. What criteria do you employ to evaluate him? Pick your top three evaluation points, and we’ll go with them.
Most people are going to have a list of criteria like this:
1. Won-lost record
2. Feeling that team is heading in right direction
3. Satisfaction with play-calling and gameday decision-making
Do you know what those criteria have in common? They’re all observable by a fan watching on TV and reading articles, and an outside opinion that seems reasonable and defensible can be formed based upon them.
There are a couple of issues with that kind of evaluation, though. Number one is that what’s going on behind closed doors during the week is the bulk of the head coach’s job. You see the guy at work for three hours per week, but the other 75 (or more) hours that he’s working are almost certainly more important to the long-run success of the program than what happens on gameday.
The three gameday hours are when a team puts their best foot forward against another team trying to do the same thing. Everybody gets paid, and one team will win, and the other will lose. The inputs during the week, the month, and the season will tend to influence those outcomes, but they aren’t solely determinative.
Sometimes, the other team has better talent than you do. Should a coach get fired for losing to a clearly more talented team? In a vacuum, the answer is no. The team should address its talent shortage.
There’s also randomness to consider. In general, people struggle with the concept of randomness, and it’s unfortunately human nature to falsely affix blame or credit to people (or gods) for random events that happen in life.
For example, when a team loses more fumbles than it recovers, the natural inclination of a football writer tends to be to suggest that the team isn’t tough enough, or well-coached enough, or aware enough, or fundamentally sound enough. There has to be something or somebody to blame, when the reality is most likely that it’s random, and it’s all based on how an oblate spheroid (the shape of a football) bounces on flat ground. (Answer: Unpredictably.)
If a team loses three games because of bad fumble luck, and goes 7-9 instead of 10-6, should a coach be fired? Reporters, fans, and historical custom say yes. I say no, in a vacuum. The fumble luck should regress to the mean, and your outcome the next year (holding all other variables constant) should be about 10-6.
Oh yeah, we need to talk about holding all other variables constant. You can’t actually do it, because the NFL is an enormously complex and dynamic system of players, coaches, teams, divisions, conferences, and schedules. Within that system, players (and coaches and teams) are improving all the time, while other players (and coaches and teams) are getting worse. One year, a team may have a favorable schedule, and the next year, it can be unfavorable. A head coach might lose a key assistant coach to a head coaching job, and have a setback there that causes problems in the following season.
So many variables change that it’s impossible to intelligently evaluate based upon a won-lost record. The majority of the variables are outside the control of a head coach, and should be understood that way. You have to understand the context around the won-lost record. If you attempt to do that, you’re ahead of those who don’t.
Here’s one example of needing to understand context – the recent firing of Rob Chudzinski in Cleveland, after only one season as head coach. The statements put out by the team said the following:
"We appreciate Chud's passion for the Browns, and we have great respect for him both personally and professionally. We needed to see progress with this football team. We needed to see development and improvement as the season evolved and, unfortunately, we took a concerning step backward in the second half of the year.”
The emphasis on development and improvement is mine, and I’ll come back to that momentarily. First, understand this in the context of how the average fan and NFL writer understands that statement, and the rationale of the Cleveland leadership. Which is to say, simplistically.
The team did poorly in the second half, and lost a bunch of games in a row, so we held Chudzinski accountable and canned the guy.
Chudzinski didn’t get fired because the team went on a losing streak in the second half of the season. If you got them to be honest, the Browns’ leadership didn’t care that much about winning this year, because they knew they didn’t have a QB, and that they needed to position themselves to get one in the 2014 Draft. Losing helps with that.
I’m personally pretty sure that Chudzinski got fired because his performance during the off-camera 75 hours per week wasn’t what Joe Banner and Mike Lombardi were looking for. They didn’t care about winning in 2013 so much, but they deeply cared about the development and improvement of the players they’re trying to build around. That’s why I bolded “development and improvement.” If you read between the lines, they evaluated Chudzinski’s performance to be lacking in that area.
To some extent, that’s what you get for hiring a technocratic scheme guy, who doesn’t particularly have a reputation for coaching up players, and causing them to improve. His name was hot because his scheme looked good at Carolina.
The reaction to Chudzinski being fired was interesting. A lot of writers fell back to the default idea that if you hire a head coach, you've effectively married him. And like in all marriages, that means you have to give it at least three years, even if she quits putting out when you get back from the honeymoon, or something. I think that’s garbage, and I applaud the Browns for admitting that they thought they’d made a mistake, and moving on.
Among Browns fans, of whom I know many, the perception was “Woe is us! Same old Browns!”
When anything happens to that franchise, the default assumption is that it’s a disaster. Trade Trent Richardson for a first-round pick (a great deal, as I said at the time) it’s “Why aren’t they trying to win?”
Mostly, Browns fans are tired of churn, and their context is the totality of the last 15 years, during which there's been a tremendous amount of unproductive churn. For Banner and Lombardi (and owner Jimmy Haslam), though, their context is the last year only. To them, it doesn’t matter what happened in the previous 14 years, and they have to ignore fan sentiment, and do what they deem to be best for the team right now.
You want another example of context? How about one that hits close to home? How about the case of Josh McDaniels in Denver? Yeah, I’m going there.
McDaniels was hired on the strength of promising to completely tear down and rebuild the entire football operation in Denver. He said he was going to emulate a better and more coherent football operation, that of the Patriots, and that’s what Pat Bowlen and Joe Ellis enthusiastically signed up for.
McDaniels set about to do exactly what he said he was going to do. He tore the whole thing down, and started building it up the way that he had been raised to think it should be built. He switched from a historically terrible 4-3 defense to a two-gapping 3-4, and he fired Jeremy Bates and ran off Jay Cutler. There was no reason to think that success would come quickly, given the scope of the changes underway, which, again, McDaniels had told Bowlen and Ellis he was going to make. (Remember, it was Bowlen himself who ultimately ordered the football management to trade Cutler.)
A couple of things went wrong for McDaniels. First of all, the Broncos started 6-0, and expectations got ratcheted way up before they should have been. Second, he refused to play nice with the local media, particularly the asshole Woody Paige. Woody wanted to patronize McDaniels, and give the kid advice, and the kid (rightly) didn’t give a damn what Woody (an intellectual lightweight, and worse than that as a football thinker) thought about anything.
You can flip off the media when you’re winning, and they just have to take it. You can ask Bill Belichick about that, and I’m sure Josh learned that method of media relations during his time in New England. If you start losing, though, the media can really start causing some drama for you, even in a relatively tame place like Denver.
If McDaniels had consistently articulated the scope of the rebuilding effort, and kept up a messaging campaign that it would take some time, and if he were nicer and more deferential to the media, they would have probably helped provide him with some cover. He chose to do the other thing, and he paid the price. Next thing you knew, they were uncovering the overblown (in my opinion) videotaping thing.
Then it was all just drama. Pat Bowlen and Joe Ellis weren’t used to drama, and didn’t like it. I believe that they determined that going forward with McDaniels being such a lightning rod was untenable.
The interesting thing, to me, is that they didn’t abandon his program. A lot of his assistant coaches were retained, and there was no wholesale cleanout of his players, either. The Broncos basically brought in some (much) more media-savvy guys in John Elway and John Fox, and they more or less continued the rebuild, only deviating from the existing plan at the outset in small ways, such as what to call the defensive scheme.
If you ask a lot of people, some of whom are Mike Silver, and others of whom are commenters on this site, McDaniels proved that he’s a bad coach with his truncated tenure in Denver. These people are sure of this, but they’re wrong. Even if the next team he is the head coach for fails, they’re wrong.
There are a couple of key biases at work in making them/you wrong. The first is the granddaddy of them all, confirmation bias, which TJ has schooled IAOFM readers on in the past. The second is one that was unfamiliar to me until recently, before I read Rolf Dobelli’s The Art of Thinking Clearly.
Dobelli goes through 100 biases, and all the way in chapter 95, he covers Illusion of Skill, which was originally proposed by Daniel Kahneman. To paraphrase the two-page chapter, Dobelli makes the point that in considering success, it’s not easy to determine how much of it comes down to luck, and how much of it relates to hard work and talent.
In the narratives that people like to create, luck is de-emphasized, and we gravitate toward understanding success as being a function of hard work and talent. We like to make heroes out of our CEOs and football coaches, right? Of course, if you think about it, hard work and talent, while mostly necessary for success, aren’t sufficient causal factors for it.
Think about it like this. Imagine a technology entrepreneur who has a great idea, builds a business around it, and eventually sells out to Google, and walks away with $500 million. This happens, and it’s not hard to picture, right? How many of these guys go on to repeat that success in the future?
The answer is very, very few. They had some luck going for them the first time around, and they don’t get it again in their future endeavors. Just because you get one great idea doesn’t mean that you’re a fountain of future great ideas.
As Warren Buffett has said, “a good managerial record is far more a function of what business boat you get into than it is of how effectively you row.” Dobelli makes the point that while a plumber or a lawyer makes their living from their demonstrable skills, for leaders and entrepreneurs, skill is necessary but not critical.
With football coaches (or entrepreneurs), if it was all about skill, you could plug the same guy in repeatedly, and expect the same results. You would know that Mike Shanahan is going to win you Super Bowls, or that Josh McDaniels is going to tear down the team and piss off the media and fan base. That’s not the way it works though. Things are much more dynamic and random than that.
In an environment as luck-dependent as the NFL, particularly where every game is zero-sum, the Illusion of Skill is completely pervasive. It informs our own perceptions about every aspect of the game, and since you typically have not-very-smart people writing about and commenting on the game, the narratives that they create acquire the authority of undisputed truth, while coming from a biased and idiotic place.
Remember, sports reporter is to reporter as gym teacher (remember those?) is to teacher. These mostly aren’t people who think much about how to most effectively think. They’re not reading Nassim Taleb, or Daniel Kahneman, or Rolf Dobelli.
I want to give you a current example of the Illusion of Skill in a coaching hire, in the form of Jay Gruden. The football franchise representing Washington just hired this guy, which suggests that they think he’s an upgrade over Mike Shanahan. My question is, why would anybody think that?
In Washington, Gruden was uniquely helped by his last name, because the owner is fundamentally a marketing guy, and every decision he makes is through the prism of making a sale to his fan base. Jay Gruden’s name is compelling because of his brother Jon’s track record. Let’s look at that for a minute.
Jon had a couple 8-8 seasons in Oakland as a young head coach, and then experienced equal parts luck and skill in harnessing a late career renaissance in Rich Gannon. They won back-to-back division titles in 2000 and 2001, and had one playoff win in each season. At that point, Jon shuffled off to Tampa Bay and took over a good team from Tony Dungy.
In his first season in Florida, Jon won the Super Bowl with Dungy’s team (and against Gruden's old one). The following two years, the Bucs fell to 7-9 and 5-11. In 2005, they had a brief resurgence with Chris Simms at QB, and won their division at 11-5. In Jon’s final three years, they went 4-12, 9-7 and 9-7.
In Jon Gruden’s career, he’s 95-81 as a head coach, which is a .540 winning percentage. If you look at the above-linked PFR page, there’s a lot of season-to-season volatility there. My take is that Jon is a reasonably good coach who isn’t any better than most other coaches at overcoming the effects of randomness. Like nearly all coaches, he ended up getting fired.
Jon Gruden salvaged his name by becoming a media guy, and much like Bill Cowher (another overrated coach), the fact that Jon is reported to be sought-after every year keeps his coaching reputation strong, and actually strengthens it beyond what is necessarily deserved in hindsight. I remember the Navy much more fondly than my experiences at the time warrant.
That’s the deal with Jon Gruden and Bill Cowher, too. Both those guys are going to keep doing TV, and they’ll be remembered roughly like John Madden is – a good coach who bowed out to do TV.
So Jon Gruden is overrated, and Jay Gruden is trading on his name, to some degree. But what about Jay as a coach? Well, I claim to be somebody who knows something about offensive football, and my subjective judgment of his performance is that he’s pretty average as an offensive schemer.
Nothing the Bengals have been doing is innovative in any way. None of their offensive players have taken any huge leaps to where they’re overachieving their talent, and in fact, guys like Jermaine Gresham and Andre Smith have often underachieved.
The quarterback of the Bengals, Andy Dalton, is now the target of intense fan and media scorn. The reality is that he has a completely average skill set, and that Gruden got him to play like a completely average NFL QB. Basically, he turned two nickels into a dime. Good for him!
As for last weekend’s playoff loss to San Diego, I put that outcome far more on Gruden than I do on Dalton. Where was the adjustment to the protection scheme once it became clear that the Chargers were getting consistent pressure? It was completely obvious, and yet Gruden did nothing to help his QB, and that was the difference in the game.
I’m not saying that Jay Gruden is going to fail in DC. I don’t have any idea how the effects of randomness are going to treat him or the Washington franchise. I’m saying that I subjectively think that he’s pretty run-of-the-mill, and that he’s going to be riding the wave of factors which are beyond his control, like most coaches.
Coaching success is only somewhat about talent and hard work. There’s a hell of a lot more to it than that, though. Anybody who thinks they can predict success or failure in coaching with any degree of certainty is taking an inherently stupid position.
An owner isn’t buying a result when he hires a coach. He’s simply adding an element to a huge pile of randomness and dynamic variables, and the degree of success that will be realized is not at all predictable on the date of hire.
Addendum - keeping it close to home, if the Broncos don't win on Sunday, John Fox doesn't deserve to be fired. Anybody who suggests otherwise is most probably an idiot. I'm looking at you, Mark Kiszla and Zach Fogg. (Who?)
Wins and losses in football games are not a function of coaching competence. Much, much more is going on. Certainly, in a one-game sample, no fan or media person should ever presume to think that a coach proved or disproved his worth, amid all the randomness, and the fact that the other team is also good, and is also trying to win. Losing a zero-sum proposition doesn't mean that you're a loser.