It was 4th and goal with little time left in the first half at yesterday's Super Bowl. A field goal would put the team from Philly up 18-12, and the safe, conventional call would not have surprised anyone.
We all know that's not what happened. The Eagles went for it.
Coach Doug Pederson and staff didn't much care that if the play failed, and they lost the game, the media would've had a conniption fit. They went for it because the numbers said they should.
As Ben Shpigel noted in the New York Times (written before the victory), the Eagles have been doing this all season long.
As we noted in a similar post last year with play calling, there are two main biases at play.
There's a flight to safety, or what we may call the bird in the hand decision making tree. This is the 'we're here so we have to get points' phenomenon the media and some fans love. And what we read about in school when we studied prospect theory.
From the NYT article: "Humans put an undue amount of weight on something that they’re giving away versus something that they could acquire. But the problem is, that’s counterintuitive..."
The second part, as I see it, is survivorship bias. When someone does something unconventional and it fails - say Pete Carroll calling a positive expected value pass on second down from the one that gets picked off - they are strung up to dry because it's very visible. When it works - like Bill Belichick calling a pass from the one in overtime in last year's Super Bowl against the Falcons to win the game - it's not visible. One man's genius play call is a goat for another man.
The above tends to make play calling and taking chances a form of heresy in NFL coaching circles. If they don't "take the points" and lose, fans and media go nuts because they gave away something they had. If they call a play that's mathematically sound, but unconventional, it's a billboard.
Both instances can get coaches fired, so risk avoidance, despite the sub-optimal math, is commonplace.
Flipping over to horse racing. Is there a business or sport that's slave to the conventional as it is?
Track executives (some, not all of course), come to work each day with the specific goal to not rock a boat. If you do try something mathematically sound - like a Canterbury Park takeout reduction and through no fault of your own it doesn't deliver right away - you might end up looking for work.
The response is not worth the risk. Track executives tend to kick field goals from the one yard line; they look at what other execs do and copy them.
This is why, in my view, that we so often see takeout hikes in this sport, despite the data and numbers showing the exact opposite should be tried. Joe did it and even though it looks like it didn't work, they've got my back; we've been doing it for generations, so we'll do it again.
Doug Pederson coached a great game. Racing - and I don't think we're being too harsh - has not coached a great game; it has coached a safe game. And that, in my view, has always been one of the sport's biggest problems.
I was digging through some old electronics recently and came across my Slingbox . For those who don't know, a Slingbox attached to your...
I thought I’d share with you all my Eclipse Award Ballot. Who am I kidding, Vladimr Putin has a better chance of getting an Eclipse Ballot...
I saw a tweet from Craig Bernick the other day. Horse business currently has $10 Billion in handle with 20% takeout split by various indu...
It all started with a note on a dark web forum populated by several horse racing people called MIDAS – “messages the industry doesn’t actual...
During this whole troubling spectacle - summarized and updated, story by story here by Jessica Chapel - I've had a devil of a time figu...
In Everybody Lies - an interesting look at big data in the Internet age - the author shares a data study from one of his mentors who looked...