There was an interesting discussion thread started yesterday by readers Jay and Trav concerning 4th downs. I wanted to write a post on this because this is something I often get worked up about! And I also see a (very) gradual evolution in football that may eventually lead to NFL coaches coming around to my view on this. Anyway, here are the posts that got this started with my commentary beneath. Let us know what you think.
Read Jay wrote: Oh hey… Question for you. I am sure you saw the stories on Pulaski HS in Arkansas. What do you think of the statistics based strategy? Think it could work in the NFL? I really enjoyed seeing Rodgers convert though 4th downs today!
Reader Trav responded: Jay – Read that same article some time ago and it is an interesting use of the data to make game management decisions. We talk about this at work all the time with our business partners: “the data doesn’t lie”. I think it’s a matter of who will have the guts to step up and embrace a data driven decision model like this and be willing to take the heat for it when they don’t make a 4th down conversion and force a short field for the D.
I think what hurts NFL and College Head Coaches to the same extent, is that they have the precedent of “this is always how it was done” that rules their judgment along with, generally, being way too conservative in their management style.
Using the NBA as a counterpoint, it is such a statistics driven league to the point of how to defend a player to push them to a particular point on the court because their shooting percentage is X percentage points lower than at a different spot. Same with MLB.
You don’t read that much about the NFL using similar statistics, or I should say I haven’t seen cover stories pointing this out like the SI article you referenced. They may do it and it would be interesting to see how often the “gut” of the coach would override the data. I have to believe MM has a chart about when to go for 2 along with Asst Coaches with him and in this case went with his gut to kick the PAT rather than try it earlier in the 4th.
My Comment: Good points Trav and Jay. Trav, I agree there is really an element of “this is how it’s always been done” – something that I find so unfortunate about the approach to 4th downs. And I also agree Trav that it often comes down to coaches not wanting to take the heat for bucking what has unfortunately become the conventional approach to 4th downs. While coaches may understandably fear the immediate and seemingly relentless scrutiny of the now ubiquitous NFL media and the wrath of the fans, that’s no excuse in my book because on a fundamental level, their overly conservative decisions are not supported by the available analytics.
Jay, I have read those articles and a few others about teams that have done this. The best guy to read on this stuff is Gregg Easterbrook – ESPN’s Tuesday Morning QB. Like me and the coaches of these teams, he can’t stand the weak coaching decisions to punt at times when it really doesn’t make sense and every week in his article he calls out the coaches that made poor 4th down decisions. Now, I’m not sure I’d go as far as some of these coaches go when they virtually never punt as there are actually some situations where I would punt. But overall, the numbers really do support being far more “aggressive” than coaches are on 4th downs and it’s really a wonder that more coaches don’t take more “risks”. And I don’t find it coincidental that the coaches that are usually more willing to take “risks” are simply better coaches (Belichick, Sean Payton, now Chip Kelly, etc). As you note Trav, I think MLB is a good example of how numbers can help make decisions (though I have to say, there are times when I think some MLB managers go perhaps too far the other way relying solely on numbers and discounting other factors).
Admittedly, there are a number of other factors to consider when going for it on 4th down like: how far do we have to go (stats obviously say the shorter the distance better chance of converting)? where on the field am I? how good is my offense? how good is their defense (or how well has their defense been playing lately)? how good is my defense (if I have a stifling D, playing a field position game makes a tiny bit more sense)? how good is their offense? is it a home or away game? how good is my punter/kicker? is it early or late in the game? could my team use a momentum jolt? will not converting cause a major confidence meltdown for my team? past short-yardage conversion percentage? do I have good short-yardage conversion plays available? A good coach should be able to consider all these questions quickly in the moment when making a 4th down decision. And these factors do matter – and can and should at times cause a coach to perhaps disregard analytics.
But one other key factor that isn’t talked about much in discussions of all of this is that 4th downs are often made more difficult because of the play calls made on 3rd downs. In fact, if I were a coach, I would make the decision to go for it on 4th down, most of the time, before it is even 4th down. I would decide to go for it when making the play call for 3rd down. For example, if it’s 3rd and 7 at the 50 yard line, most teams try to throw it 7 yards to get the first. If they don’t get it they punt. That’s conventional wisdom (yet that’s not supported by the widely available/proven analytics). If I were a coach, I would tell the team on 3rd down that this is 4 down territory and I would call a run play or throw a short pass for 3-4 yards or so, leaving a much shorter 4th down conversion (of course still hoping my guy can get those 7 yards). Going about it this way would greatly increase the overall odds of converting then on 4th down and short. And defenses would likely be caught off guard as they’d be playing for the 7 yard pass and giving up pretty much anything less than that. Interestingly, this is how coaches should approach down/distance on 1st and 2nd downs too in order to increase the odds of a 3rd down conversion – part of why I find the Dallas Cowboys games so unwatchable (they face more 3rd and long plays than anyone it seems…). Anyway, if coaches embraced this approach (again, generally deciding on 3rd down that they will be going for it on 4th down if need be, and then going after shorter/higher percentage yards on 3rd down), the numbers that already support going for it on 4th down would be even harder to ignore because the yardage needed to convert on 4th down will have been decreased most likely by the 3rd down play. And the shorter the yardage needed to convert, the higher the chance of converting. Of course. (And I should add, that if the 4th down play call could be immediately ready for the team after 3rd down – i.e. no huddle – the 4th down conversion attempt would be considerably more difficult for a defense to defend as they likely wouldn’t have the right personnel on the field.)
Bottom line is – if I were a coach, I would go for it on 4th down quite a bit more than NFL coaches do. And that’s why I state often during Packer game live blogs that “this is 4 down territory”. While I can admit that I’d be concerned about the aftermath of a decision that didn’t quite work out if I were a coach, I wouldn’t let that guide my decisions making. In most cases after considering other factors, I would most likely let the numbers guide it.
Support for this viewpoint can be found in this fantastic article at the New York Times. Be sure to check out the chart that breaks down what NFL coaches typically do vs what they should do. Just fantastic.