Tony Dungy, Lovie Smith, John Fox and others will never change. Dungy won a SB in spite of himself and he doesn't even realize it. You hit it 100% on the nose about the successful coaches. BB, Riverboat Ron and Tomlin are notorious risk takers as are Sean Peyton, Arians and Pete Carroll. It doesn't always work but they take that extra step to find ways to win.
But that's what's so confounding about it. If the coaches who are more aggressive are also more successful, even if it's not direct causation, why don't other coaches copy those techniques? Hell, you could use those guys as cover: "Coach, why did you go for it on 4th down?" "Well, as I look around this league, I see Belichick going for it all the time, and I figure if it works for him, why shouldn't I do it as well?" I mean, theoretically every coach should be concerned with keeping his job, and the best way to ensure that is by winning as much as possible, so why not do things that improve your chances?
One theory I've heard is that NFL coaches are NOT primarily concerned with preserving their current job. The fact is, 95% of head coaches will eventually get fired no matter what they do. What they're really concerned with is what happens AFTER they get fired. They want to get hired as coordinators, stick around, and hopefully get another shot as HC down the road. And the best way to do that is by being seen as a team player who doesn't rock the boat. (That would also explain why coaches like Belichick, Tomlin and Arians can afford to be more aggressive: there is basically zero chance that any of them will have to work as a coordinator again.)
As I stated upthread, my own theory, which is not mutually exclusive to the one above, is that there's essentially a "natural selection" that results in coaches who think conservatively. Football teams are inherently conservative, hierarchical organizations, and anyone who climbs the greasy pole to get to a HC position is probably going to be conservative. Think of it this way: I have no idea if Chris Kluwe would make a good special-teams coach, but I feel pretty confident in saying that if he ever did get a job, he would approach it in an unconventional way, because he's clearly the kind of guy who identifies as a free-thinker not beholden to tradition. But that's the exact reason no one would ever hire a Chris Kluwe as a coach (well, that plus the fact that he publicly accused his former team of bigotry on his way out the door and got them investigated by the league). The Kluwes of the world get weeded out like a short-necked giraffe.
So when a coach is facing 4th and 1 at the opponent's 40, or when he scores on a Hail Mary on the final play of regulation to pull within one, he doesn't consciously think, "I'm going to punt/kick the XP because, after carefully weighing all the options, that makes the most sense." Everything he's learned over the past two decades in football tells him that the conventional wisdom is to punt/kick in those scenarios, and it's also told him that above all he should follow the CW.
First of all, this is an excellent post.
To answer the bolded portion above: it's all about variance, or rather the perception of variance.
Doyle Brunson's
Super/System is often called "the Bible of poker". Brunson himself wrote the chapter on NL hold'em. What you'll notice about it if you read it is how aggressive an approach he advocates. He recommends betting and raising with good hands and average ones, with made hands and draws, and often with nothing at all. It's a style that would be right at home in a modern-day WSOP tournament.
Brunson published that book in 1979! Yet twenty years later (as you can read in accounts of the day by guys like McManus and Alvarez), people were still playing generally passive poker and Brunson and his ilk were still described as "reckless" and "gamblers". Why? Because even professional poker players of the '80s and '90s didn't have computerized analyses and online databases to draw on. They only knew what they'd seen and what they remembered. And because people remember losses far more vividly than wins, they remembered all those times they pushed all-in on a draw and busted out, even when it was the right play mathematically.
It wasn't until the 2000s - with the dawn of computerized poker programs, rigorous mathematical analysis, and eventually online multi-table poker - that players awoke to the realization that in most cases, the more aggressive, higher-variance play is the more profitable one. Many old pros, who'd honed their craft over literally decades at live tables, spent years getting their asses kicked by the 20-somethings of the Internet era while they learned to adapt to this new, more correctly aggressive game. Some never did, and went broke.
Football analytics are about 15 years behind poker at this stage. For the most part, the data are out there, but today's head coaches are mainly of a generation that weren't raised on it and so don't trust it. They've been standing on that sideline for 30 years now and, dadgummit, they know better than to go for it on 4th-and-2 from their own territory.
As with poker, it's going to take the next generation of analytics-savvy guys rising through the coaching ranks and consistently kicking the asses of the old guard before I expect that to change dramatically.
at Dungy's conclusion?
