Admittedly, I was a bit fuzzy before writing this on the specific timelines associated with each so-called age of the world. Fortunately, I was able to look it up, because the age we’re in currently is the “Information Age.” That much, at least, I already knew. I also know that the increased access we have to information in recent decades has allowed us to progress in many areas at an even more prodigious rate than generations prior.
The truly important progress we’ve made, of course, comes in areas such as medicine, information technology, and travel. For the less weighty purposes of this website about football, what’s more relevant are the statistical advancements we’ve seen in the world of sports, and how that knowledge informs decisions made by coaches, scouts, trainers, and athletes themselves. While I would consider myself a proponent of the movement toward advanced analytics, I’ll be among the first to say that they aren’t quite the be-all and end-all that some of their staunchest advocates or enemies make them out to be. There’s often context to which an algorithm can’t be adequately privy, just as even the sharpest coaches don’t precisely quantify every point of data as scientifically as a well-built program. The goal should be to use all of the information at one’s disposal to make the best-informed decisions possible, hence the situational decision guides that many head coaches now carry on the sidelines and the assistants on staff whose primary game day responsibilities are focused on advising in situations to ensure the consideration of available data. The benefits of the “Information Age” provide more knowledge off of which these decisions can be made, and this knowledge is valuable whether everybody likes the decisions it leads to or not.
Naturally, as any of us who have social media or merely watch a game with the volume on can attest, the decisions that are often disliked get heavily questioned and criticized when they don’t go according to plan, no matter how sound the information behind them may be. This has happened with several decisions already this season, including many this past week. The most panned of those took place a week ago tonight, in fact, when the Cowboys cut their deficit against the Buccaneers to eight with a touchdown and opted to attempt a two-point conversion, which notably failed. I’ll return to that specific decision and others like it later, but before I move in to the rest of the article, I’d like to make one point about my assessment of this topic as a whole very clear: coaching decisions, good and bad, are made based off of information. Furthermore, the results of each decision create more information, which has an unavoidable impact of future decisions.
Before getting into the much-maligned aggressiveness that many coaches have adopted in their post-touchdown decision-making, I’m going to talk about another sort of coaching decision that has been worthy of note in recent weeks, swinging two games. One of those instances also includes Brian Schottenheimer and the aforementioned Cowboys. In Week 3, having tied the game with :07 remaining, the Cowboys opted to send a kickoff through the end zone down in Rio de Janeiro. As many of you reading will likely remember, the Ravens proceeded to hit a deep pass over the middle and call a timeout with :01 left before kicking a 56-yard game-winner. Many, myself included, said that the Cowboys should have kicked the ball in the field of play and forced a return, thus whittling away at the time in which the Ravens could run an offensive play.
This Sunday, in London rather than Rio, the Eagles also lost on a game-winning field goal as time expired, made possible by a long Jacksonville kick return because Philadelphia did kick to the returner with :08 left. When many of us who criticized the choice by Dallas to provide a touchback also criticized Philadelphia for welcoming a return, aspersions of hypocrisy were cast, suggesting that only the results skewed perceptions of the decisions.
Here’s why that’s not an accurate depiction of reality. I’ll give you a hint: it has to do with information. Specifically, two important pieces of it.
- The Ravens had a timeout remaining, the Jaguars did not. As such, the chances of the Jaguars getting into field goal range without a defensive penalty were incredibly small, while the Ravens were going to be able to put themselves into field goal range or better Hail Mary position with a long completion and a timeout, as they proceeded to do.
- The Eagles had been vulnerable in kickoff coverage earlier in the game, while the Cowboys hadn’t shown any sort of susceptibility against the Baltimore return game. The Eagles knowingly and unnecessarily put the ball into the hands of someone who was liable to make them regret it. The Cowboys essentially did the same by giving the ball to Lamar Jackson with more time than he would’ve had following a kick return.
Each of those pieces of information given are seemingly small in a vacuum, but when factored in as a part of the bigger picture, they’re among the primary reasons that these two NFC East rivals are both 2-3 when both could be 3-2. While this point addresses one of the major controversies over coaching decisions currently being discussed in football spaces, it also serves to demonstrate how useful any amount of information can be.
Now, that brings us back around to the two-point conversation, though I’ll get to a different wrinkle to that discourse before I finally address Schottenheimer’s decision head-on. Most analytical models say that you should go for two when after scoring to go down by eight, as well as when you score to go down by nine. For as much criticism as the recent decision has spurred in certain circles, it’s going for two when down by nine that really seems to irk the masses. Simply put, the best reason to do that is so that you know what you need. Suppose you score a touchdown with 5:00 left to go down by nine, then fail on your two-point conversion attempt. You know at that point that you need two possessions to win the game. You know that it’s probably wisest to attempt your first onside kick at that point rather than hoping to get a stop and recover another onside kick later. Whereas if you take the extra point and fail on your necessary two-point attempt to tie the game later, say with :30 left, that doesn’t help you very much with time being that restricted. It’s better to have as much information as possible, as early as possible, to make the best decisions moving forward.
So, that leads us to last Thursday night, when Schottenheimer followed the recommendations of the many models and went for two after a George Pickens touchdown reduced the Tampa Bay lead to eight points, 24-16. (It should be noted that the Cowboys didn’t score again, so this was ultimately only consequential for Brandon Aubrey’s fantasy owners, not to whether the Cowboys added to their win or their loss column.) I’ve heard all of the arguments in the book against this approach. I imagine most of us have. Here’s the information I know: two-point conversions are successful about 53% of the time, extra points are successful about 97% of the time. In theory, you will score more points and win more games if you always go for two. Now, am I advocating for that extreme of an approach? Of course not. But it is relevant information. Things don’t simply work in real life because they work in theory.
Unless you choose to kick extra points on both touchdowns and simply accept overtime as your fate, you’ll have to go for two at least once near the end of the game. In theory, does it matter whether that comes after the first touchdown or the second? No, so theory falls short once again. There are analytical, data-based reasons why it does matter, and there are practical, common sense reasons, no matter how many people protest that there’s no common sense to this approach at all.
The data says that because your odds of converting the two-point try are above 50%, you give yourself a better chance to get the full eight points and be able to win on an extra point if you score a second touchdown. Even if you fail, the odds are that you’re getting one of the two-point conversions if you attempt two. As such, sending the game to overtime is the worst of the statistically likely scenarios. That logic holds up plenty well on its own, mathematically, although many factors can admittedly muddy the waters.
The practical reasons go deeper than the win probability calculated by these models. And those reasoned are based on the same principle I outlined earlier: It’s better to have as much information as possible, as early as possible, to make the best decisions moving forward. Along with being based on solid statistical information, going for two earlier gives you information.
First, as with the scenario of being nine points down, you know what you need. When you convert, you know that a stop and another touchdown almost guarantee a win, as all you need is an extra point. (For how much critics talk about a failed two-point conversion killing momentum, I think they’re underrating the momentum to be gained by converting it.) When you don’t, you know you need a touchdown and a two-point conversion to force overtime. It eliminates the tough decision in a one-point game of going for overtime or going for the win. It forces your hand. If you were taught about peer pressure as a kid, you probably heard some variation of the idea that if you choose not to (insert behavior here) before you’re ever asked, when you are asked you will have already made your choice. If you go for two earlier, you’ve already made your choice for the end of game scenario, whichever way that may go.
Second, if you don’t convert the first time, you’ve at least seen the defense in that situation. The offense has a lot more options from the two-yard line than the defense does, so if you have a better idea of what to expect from the defense, you can pick the right play call to counter it and have a better chance to even the score on your second try.
These are all important pieces of information, and it’s once again preferable to have as much information as possible, as early as possible. There’s little reason not to get that information too early rather than too late.
That being said, I did say earlier that there are situations where the presence of other information supercedes and affects the information provided above. Lest anyone think I’m entirely beholden to the conclusions of a formula, I’ll provide one example of an exception that I’ve used often in these discussions. Suppose you score to go down eight (or nine), and your opponent is incredibly hard to score on down at the goal line in the passing game. Now suppose that same opponent is extremely vulnerable to power running, and that’s the best way to get your two points. Finally, suppose that on the touchdown play, your best pulling guard had his helmet come off, and is ineligible for the next play, i.e. the two-point attempt. Do you take your chances, knowing that they are lower than normal? Or do you “take the points” and be ready to score on a power run the next time? I would take the points, with the information I have there.
Regardless of what we each might agree or disagree with as fans, these decisions are made using real, valuable information. Every decision exists exclusively within the context of available information. Every responsible decision-maker strives to use the breadth of that information as the heartbeat of their process. Information is paramount for success in any endeavor. Without information, you’re just guessing, and I assure you that guessing will bring far worse results than those we complain about as fans when measured decisions happen to go awry.
I hope when I go to the doctor that he or she is making the most of all available information while treating me. I hope my religious leaders are doing the same with the information they have, both spiritual and secular. I hope political leaders are using all of the information available to them in the decisions they’re required to make. And I hope the coaches of my favorite football teams are doing the same.
