A Sporting War, pt. 1
Honestly, I don’t actually read much sports commentary at all. But in my online perusals of regular news websites, I occasionally find a sports discussion that raises parallels between sports culture and military culture. Thus we have another edition of odd parallels between sports and war. This time it’s not so much about gender, as the reasoning for choosing one strategy over another.
What piqued my interest recently were two articles that relate to the unwritten conventions of American football, and how those conventions also seem to apply to warfare, generalship in particular. I’ll talk about the first article today, and then do a follow-on post on the second.
The first was an article from Slate (it’s difficult to figure out who is the actual author of this piece) which discussed whether or not it’s a stupid idea for football coaches to choose to punt on 4th down, rather than go for it. Those familiar with the American game can skip the rest of this paragraph, but here’s some background: the rules say an offense gets 4 downs (four attempts or plays) in which to gain 10 yards, barring penalties. But 95% of the time, it’s really only three downs, because on 4th down if you don’t get the ball past that first down marker, the other team’s offense takes over at that spot. So instead, most coaches usually decide to have the offense punt (kick) the ball to the other team on 4th down, in order to exchange ball possession for field position – you give the other side’s offense the ball, but much further away from your own goal line than if you don’t make the 4th down attempt. On rare occasions, usually towards the end of the game as a last-ditch effort, or when there’s only a short distance to go – say, just a couple of yards, coaches might take a risk and go for it on 4th down. It’s a no-brainer if your team is behind late in a game – you go for it on 4th down to keep the possibility of victory alive. But more challenging is what to do if you’re ahead, but not ahead by enough that another opponent’s score wouldn’t beat you. In other words, up by an uncomfortably-narrow margin with a 4th and short decision to make. That’s where the real question comes in: do you choose an aggressive strategy to try to put the game away by converting the 4th down into a 1st – with the concomitant risk that a failed attempt will make it a lot easier for the other side to score if you fail – or do you play it safe, rely on your defense and punt? There are other wrinkles to complicate matters further: if you don’t go for it conventionally, you could still try a trick play: a fake punt, where you line up to punt the ball but then try to gain the first down yardage with a run or pass. Or maybe you’ll pretend to line up to go for it, but instead try to trick the defense into jumping offsides, which would give you a free 1st down. You can be sneaky (but not too often if you expect it to work), yet the old vets in the broadcast booth prefer the ‘big boy’ game – line up and smash it down their throat, or do the smart and safe thing and punt for field position. Most coaches, most of the time – especially if it’s early in a game – take the safe route and punt.
So you have interesting combinations of risk-versus-reward calculations in a physical and mental competition, where subterfuge (fake punts) and misdirection (lining up like you’re going to rush right up the middle for the needed two yards, and then instead throw a pass to a wide-open receiver) are always a distinct possibility. And, as in wartime, you have to worry about the impact of command decisions on the morale of the players: the play’s result might have a psychological impact on the team (and certainly does on the fans, whether they are in the stands or back home watching the latest war news on TV). According to the announcers at least, if you don’t go for it on fourth-and-short, you might be sending the message that you don’t trust your offense – “You mean you can’t gain a single yard? What do we pay you players for?!?” But if you do go for it and don’t make it, you’re putting your defense in a really bad situation (backs against the proverbial wall) that they don’t expect you to put them in, and perhaps you’ve actually shaken your offense’s morale even worse than if you didn’t let them go for it – “You mean you really couldn’t gain one yard? Wow, you really must suck.” Given that football commentators, like many military historians, frequently interpret the game in moral terms – e.g. the winning team “wanted it more” – it’s no surprise that the coach’s decision is seen as having a big impact on the psyche of the team. Sometimes you can find other voices coming from past players who take a professional tack, and argue that players’ psyches aren’t really affected by such matters, since they are professionals after all. But everybody likes to judge a player’s psyche by their body language, and the dominant sense is that the results of games depend on critical turning points, and these inflection points are fraught with psychological importance.
Enter the Slate article, dissecting the Carolina Panther coach’s decision to punt on 4th-and-short rather than go for it, with 1:44 left in the game. The Panthers downed the ball at the 1 yard line, yet they ended up losing the game when the Atlanta Falcons marched down the field to score the winning points. What the Slate article focused on was questioning this unwritten convention of ‘conservative’ play-calling on 4th down, using a statistical approach to the game – in essence a type of sabermetrics, if you are familiar with Moneyball and the statistical analysis of baseball. I won’t get into the details of the math, but basically the author (and commenters) argued that four attempts to gain 10 yards will always give you better odds than only three attempts, and that since most of Carolina’s short-yard plays in general were successful, therefore the coach should have gone for it since his baseline odds were good that he would make it. The comments provide a bit more discussion, mostly in support of the more ‘brave’ decision to go for it on fourth down (insert gender here). Some of the commenters also mention a similar scenario where conservative strategy is seen as a losing proposition – playing a safe, ‘prevent defense’ where you allow the trailing opponent’s offense to gain short yardage in order to prevent a really long pass play. As the wags have it, ‘”prevent defense” only prevents you from winning the game.’ These two scenarios – fourth-and-short and prevent defense – are perennially debated by sports fans, just like the viability of an option offense (turning the quarterback into a potential runner on most plays) discussed in our previous post.
Personally, I could care less about whether the Panthers or Falcons won the game – no betting or fantasy football for me. What I did find interesting, however, was the parallel between what a coach should do and what a commander should do. I’m currently researching late 17C English definitions of what a Great Captain should (and shouldn’t do), an interest derived from the dominance of Great Captain biographies in military history. (I’ll let you guess which Great Captain I tend to focus my attention on.) The terms of the above sports debate follows many of the same assumptions and issues that we find discussed in military history, and that we find explicitly discussed in the prescriptive literature of the period as well. To mention a few:
- The assumption that offense wins football games. Trying to avoid defeat only leads to defeat.
- There’s the explanation that coaches play it safe not because they don’t want to win, but because they fear losing more than they look forward to winning (loss aversion).
- There’s the common explanation that coaches should try to put off the point of decision as long as possible. This is why, for example, the vast majority of coaches don’t go for two-point conversions (rather than kick an extra point) early in a game, and why, when a team is behind by two scores (say, a field goal and a touchdown, or 9+ points) and they have a 4th down within field goal range, they almost always kick the three-point field goal rather than attempt a more risky 4th-down conversion, saving the less-sure seven-point touchdown for their next drive.
- There’s the explanation that coaches play it safe because a risky loss is more likely to get them fired.
Do any of those arguments sound familiar to military historians? They should, because they have all been used to explain why so many commanders refused to fight battle in the open field, and instead relied on other strategies – attacking fortifications or hiding behind them, maneuvering rather than fighting, or outright battle avoidance (through field fortifications, retreating before an enemy, etc.). Even on the offensive, the preference in the early modern period is said to consist of ‘safer’ attacking and capturing fortifications, threatening your opponent’s line of communications, or otherwise maneuvering your enemy out of position. If a particular battle opportunity presents itself, only accept the fight if you have a clear advantage. In other words, kick the field goal (or punt) on 4th down and save the game-winning touchdown for the next possession. Hence, when the occasional military commander manages to fight and win field battles, it is clearly attributable to his will-to-battle – just as Carolina coach Ron Rivera clearly lacked a “decisive” mindset. (See my “Ramillies 1706” article for my take on this.)
What’s also interesting is that there’s a similar cultural clash both in sports and war: aggressive-minded vs. conservative-minded, which often seems to boil down to ‘professionals’ (coaches) vs. fans and the commentariat (Monday morning quarterbacks). Each of the above assumptions is contested by coaches to varying degrees: defense wins championships (until the rules change so defensive pass interference is called more often), the odd saying that ‘winning is a lot harder in the NFL than losing’, that you don’t risk a good game plan on a chancy play call, etc. In fact, it’s not a surprise to note that 95% of what fans and commentators focus on in football are field battle equivalents, those few ‘turning point’ plays during the game that appear to dictate its end result – hence the perennial fan complaint of a bad referee call costing them the game. Less attention is focused on all the other aspects of the game – training, preparation, the overall game plan, what’s happening away from the ball, and all the other things that football professionals spend 90% of their time on day-in and day-out.
Just as most sports teams tend to play conservatively, most military institutions tend to be inherently conservative as well – whether it’s their slow adoption of a new-fangled dagger inserted into the muzzle of the barrel or a new giant armored vehicle to counteract IEDs in Iraq, or whether they tend to overestimate the number of men and supplies needed for a particular operation. There are obviously counter-examples that one can point to, but their mixed record indicates that it’s not easy winning a war. In short, coaches face some of the same types of strategic decisions that commanders do, and their decisions are as likely to be dissected after the fact. Which strategy should one choose in a given situation? How do we judge the command decisions when things go wrong?
The Slate article prompted other thoughts, all of which reinforce the idea that sports strategies and military campaign strategies share some common elements – most notably human psychology and probability:
- To what extent is football a game that should be played according to the statistics? In baseball or basketball it might make sense when you play 80+ games in a season and every playoff is a series. But anybody who’s watched the New England Patriots lose to the wild-card New York Giants in two Superbowls surely understands that football, with its limited number of games and one-and-out playoff system, is one of the most random of sports out there. When we have a very small population of situation X to derive our statistics from, does it make sense to pay attention to the resulting statistics? So too in war. Field battles are indeed quite a bit like the flukey results of the single football game (whose chance element is summed up in the phrase “On any given Sunday…”): injuries, inexplicable failures of normally-consistent players, musketballs missing the commander by just a few inches… A sense of caution seems quite plausible in this context, and that’s not even when people’s actual lives, or kingdoms, are on the line.
- To what extent are statistics useful when they don’t take into account the specific situation? I’m far from a mathematical expert, but it seems like looking just at the success percentage for short-yardage plays for team X, as many of the Slate commentators seemed to be doing, misses the point. Is a 4th-down short yardage play just like a 2nd-down short yardage play? Does a team’s 4th-down success percentage matter if they only go for it in certain situations and not in others? Does it matter where the ball is on the field? Do situations influence participant psychology and performance (NBA star LeBron James, formerly known as the choker, or Tony Romo)? The Slate commentators explicitly dismiss these factors as being irrelevant – there are apparently no connections between events, each play is a draw with replacement (as the probability-types like to say). Instead of a simple baseline comparison, it seems like coaches are applying a slightly more sophisticated mental calculation – the expected value of going for it vs. punting – situational judgments. Which is not dissimilar to what early modern contemporary treatises did when, for example, they noted that there were different kinds of war, and that what would work in one type of war (say, a war of conquest) wouldn’t necessarily work in another context (say, a defensive war). If you have a larger resource base than your enemy, does it make sense to risk battle instead of patiently eroding away at the enemy’s territory through siege and keeping your army in the field?
- How long will the statistics provide a clear signal to choose strategy X over strategy Y? Let’s say teams did increasingly go for it on 4th down – what would be the overall result over time? Would that change defensive strategies, and as a result change the underlying statistics? Has sabermetrics led to the dominance of the team that first implemented it in baseball? Has information technology given a long-term advantage to any company when their competitors (inevitably) adopt similar processes? Military historians constantly stress that the enemy is a thinking opponent, which means that they respond to situations, reaction and counter-reaction, and all that. Do constant attempts to force the enemy to battle lead to more, the same, or less military success for the aggressor than if they took a more balanced strategic approach?
- One of the main meta points the Slate article makes is that you need to judge a coach’s decision based on the process they use (i.e. do they do the statistically smart thing or not?), not based off the results of the decision. In other words, you can’t say a coach made the right decision even if he won. Two points come to mind. First, this is, interestingly, the same debate early modern English contemporaries had regarding generalship. One manual argued that a commander should not be judged by his results, but by his decisions, since so much in war was outside of his control. Yet such advice was recognized to be theoretical rather than practical: commanders were also warned that they needed to be cautious because even if it wasn’t their fault when something went wrong, they would get the blame regardless – they should be judged by their decisions, but they won’t be. Given the cutthroat nature of competition for command, and the risks involved in wartime, a commander shouldn’t press his luck, and instead be content with his short-term gains. Fortuna is a bitch, to paraphrase Machiavelli. And that Italian political philosopher was a proponent of aggressively beating her down (literally, see The Prince, chapter 25); Vegetius and the prudent neostoics, on the other hand, took more of a kid gloves approach.
- A second point regarding judging process-vs.-result is more direct: why should we base our judgment of a coach (or a commander) on the process they use, and not the results? From a modern professional perspective, focusing on process might make sense: if a leader tends to make good decisions and follow a good process, he should, more often than not, do better than the average, or at least better than those without a good process. But it seems there are four problems with this. The first general problem is that this process-privileging logic uses results as the ultimate metric for success, so why not just use the results? (Perhaps there’s another explanation I’m missing though.) The second issue is that that the importance of a coach’s good process is only relevant if one assumes that leadership is a key determinant of the results – a big assumption IMHO. The adage that “it’s sometimes better to be lucky than good” is likely home-spun recognition of the influence of chance – those medievals and their ever-spinning wheel of Fate also come to mind. The two remaining problems relate to applying this to the military realm. Methodologically, it’s hard to judge commanders by their process. Most generals don’t have press conferences after the battle or war where they answer journalists’ questions about why they decided to go for it on 4th down – and when they do (usually in letters back to their boss), it’s hard to distinguish the self-justification from the reality of the situation, and conflicting reports proliferate. As a result, the war-waging process is hidden from view – as Clausewitz noted, we rarely have all the information available to commanders at the time, nor do we necessarily know what their process even was. And the final problem? The reality is that most military commanders don’t necessarily have lots of opportunities for statistics to become meaningful – there’s a reason we refer to probability as the “law of very large numbers.” One bad result (regardless of the soundness of the process) could be the endgame for a particular commander, regardless of how long the war lasts, or could even be the endgame for one side if it’s a truly decisive defeat. Even in football, the 2007 Patriots went 18-1 while the New York Giants went 13-6, yet the Giants won the most important game, the last one. In this respect, denying a battle might be seen as a way a prudent commander with a good process can increase his number of draws (in probability-speak) by not committing it all to one roll of the dice (to mix the gambling metaphors a bit).
- This debate among contemporaries over process-versus-result is also important because it influences how we rate our coaches and commanders. This is evident when we look at the criteria required for entrance into the pantheon of Great Captains. On the one hand, it makes sense to say that, in the real world, politicians and societies focus on the results of a war regardless of how good the strategic process may have been. “Just win baby!” And winning commanders definitely are foremost in our minds (Alexander and Caesar, Marlborough and Frederick II and Napoleon…), just like the successful coaches (Lombardi, Bear Bryant…). But what exactly is won is key: the war as a whole? specific battles? specific campaigns? In this respect, military historians tend to function more like the NFL Hall of Fame (battles won, campaigns won) than a simple count of Superbowl rings (wars won): Dan Marino is still an all-time great quarterback, with his (fewer and fewer) passing records, even though he never won a Superbowl. Followers of football are also familiar with the debate over whether Peyton Manning (one Superbowl ring) is a better quarterback than his brother Eli (two Superbowl rings); you hear the same debate in the NBA comparing LeBron James to other legends like Michael Jordan – how many championships does LeBron need before he can be compared with Jordan? Similarly, many of the greatest military commanders are judged based off of their peak performance, their greatest (usually battlefield) hits and personal stats (‘won every battle and siege he fought’), rather than the ultimate end result of the war they were fighting. The privileging of peak performance is illustrated by John Lynn’s favorite line about Napoleon: though he is almost unanimously ajudged one of the greatest military commanders in history, the only thing he left France with was a lot of dead Frenchmen and Avignon – and he lost a whole lot of territory that had been gained by the French Revolutionary armies. But those battles of Austerlitz and Jena-Auerstadt, and the martial glory won in them, his conquering of most of Europe, now that is his true military legacy, even though he couldn’t sustain it for ten years. And that seems to be the way military historians judge things: what was your top chart hit? How high did you climb? Caesar did well, until he pissed off a few too many senators, and of course Alexander died on a bit of a downward trajectory after crossing an Afghan bridge too far and planning a new bridge to Arabia. Several other famous commanders have similarly retained their fame even if their ultimate results were losing efforts. Most clearly would be the two Swedes Gustavus Adolphus and Charles XII – both winners, until they lost big. We could even mention the declining results that Marlborough was able to achieve as the Spanish Succession lingered on, and Frederick the Great’s late life similarly failed to live up to his earlier achievements. War is a young man’s game and Death ultimately wins to be sure, but military historians are still extremely selective about which part of the career to include in their candidate’s admission dossier into the Great Captain Hall of Fame.
But I’ve prattled on long enough. Your random thoughts on my random thoughts?