Yesterday, I took a look at some of the possible indicators of momentum in sports and found that there was scant statistical evidence of momentum existing in any meaningful way. Today, I'm going to finish up by talking about some of the logic holes in applying the concept of momentum to sports, and how it is used to distort story lines and tie together strings of independent events with false narratives.
A Mea Culpa
I screwed up. The numbers I provided yesterday with regard to the points produced on drives inside the five-yard line were inaccurate, and I apologize. The IgglesBlog Twitter account pointed out that it was unable to reproduce my findings, and after running the numbers again myself, I found that I had been including drives from the final five minutes of the second and fourth quarters and excluding drives that began during the first 10 minutes of those quarters; my intention was to do the opposite. My mistake.
After running the new numbers, there are some noticeable changes in the outcomes for each type of drive, but I still think that the general takeaway regarding momentum is the same. Remember that yesterday, I found that teams were scoring more points on drives that began deep in their own territory after a punt than they were from a takeaway or a stoppage on downs. With the correct data set, that's no longer the case:
Very interesting. Now, the (possibly) momentum-inducing takeaways are well ahead of punts in terms of expected offensive return. That in itself suggests that momentum might mean something in terms of changing possessions, but if that were the case, why are teams still not seeing any sort of boost when they stop the other team on fourth down inside their own 5-yard line, which should be an enormous momentum swing? It could be that 105 drives isn't a large enough sample, but given how frequently coaches and announcers refer to such a stop as a momentum shifter, it's disheartening to not see any evidence of a fourth-down stonewall meaning anything for the ensuing offensive series.
If both takeaways and fourth-down stoppages were producing more points per drive than ones beginning with punts, there would seem to be something to the argument, but given what the data suggests, I'm inclined to wonder whether correlation has anything to do with causation in terms of the momentum shift of the takeaway and the subsequent success on the drives in question. More data would help, of course, but at the moment, I still don't think history tells us anything about momentum shifts producing meaningful swings on these possessions. But I did want to provide the correct numbers.
The Logical Argument Against Momentum
With that said, I think there are a few arguments that really put the limitations of momentum into perspective. As with the statistical arguments, because momentum is a limitless abstract concept, it's impossible to provide a case against every single definition of what momentum is or could be, but there are a few loopholes that come up in most cases worth driving a truck through. My list of the simple logic arguments against momentum:
• When momentum is discussed after the fact, it's only ever discussed in situations where you can make a case for momentum existing: In other words, nobody ever makes a case that the incidents in a game or a season ever occurred without the concept of momentum being involved. If there's a seemingly clear turning point, momentum is invoked. If there's no such turning point, some other aspect of performance becomes the topic of discussion.
Take the Dodgers and Cardinals. Because the Dodgers were a bad team to begin the season and started playing way better almost immediately after calling up Yasiel Puig, you can make the case that the arrival of Puigmania in Los Angeles was the team's turning point and began the process of turning things around and developing momentum. Not an unreasonable point at all. On the other hand, the Cardinals have simply been good all season; there's been no notable point when they went from one level to the next. Nobody ever says that the Cardinals are good despite the fact that they don't have some obvious momentum sustaining them. When we talk about the Cardinals being good, we talk about how good their starting pitching is or how fantastic their player development is.
• Likewise, when momentum is discussed in the moment and it proves to be irrelevant, there's never a critical analysis of whether that momentous incident actually meant anything: Go back to football and imagine a key turnover; for fun, let's say it's the Giants forcing a Tom Brady interception at midfield early in the fourth quarter of a tie game. After that takeaway, you'll often hear the announcers talk about how the Giants have taken the momentum away and how they can capitalize on their opportunity to keep the momentum. (This conversation doesn't always happen, of course, but I think you would be lying if you said this didn't happen a decent amount of the time.)
Now, if the Giants promptly score on the ensuing drive and win the game with that touchdown, every story after the game will talk about how the takeaway was the turning point that gave them the opportunity to win. On the other hand, if the Giants go three-and-out and punt, will there ever be any discussion of the fact that the "turning point" actually didn't make the momentum shift whatsoever? Of course not. During the game itself, the announcers will focus on whatever happened on that third-down play that forced the Giants to punt, say something nice about a defensive player, cover the ensuing punt, and then go to commercial as Tom Brady runs back onto the field. In the game wrap-ups the following day, you'll read about some other momentum-shifting moment later on in the game or about some other measure of performance that has nothing to do with the (in fact nonexistent) momentum shift of that Giants interception. That's an unfair advantage for momentum: You only hear about it when there's evidence it actually meant something.
• Momentum is appropriated to explain acts we know are random, independent events and/or unsustainable acts in a small sample size: Step away from sports for a second. This one's as simple as flipping a coin. Flip a coin 10 times and you've got better than a 10 percent chance of getting either eight or more heads or eight or more tails to appear within those 10 flips. You're smart, though, and know that the coin flips are independent events and that one has nothing to do with the other, right? Stand around a roulette table for an hour and watch the people pass by who look at the video board and shout that a red or black number is due.
Now, apply that to football for a second. The Arizona Cardinals were supposed to have momentum when they went on that arbitrary-endpoint-friendly 11-2 stretch between the end of 2011 and the beginning of 2012, but a slightly closer look at the numbers saw a team that had gone 10-1 in one-touchdown games over that stretch. Because we know that teams over the long haul will win about 50 percent of their close games, the momentum argument was just covering up for random chance and a good amount of luck in those tight games. Sure enough, when they finished the year losing 11 of their next 12 games, nobody was wondering about what happened to the momentum from their 4-0 start.
• Momentum arguments stack on top of one another without any regard to whether one might mean more than another: Let me go back to last November, when the Chiefs and Broncos played out a classic case of momentum vs. momentum as spelled out by the two announcers for that game, Greg Gumbel and Dan Dierdorf. (I wrote about this at the time here.)
At the moment in question, the Chiefs were facing a fourth-and-2 from the 4-yard line, up 3-0 with 4:35 left in the first quarter. They chose to kick a field goal, to the boos from their home crowd, over going for it. While Dierdorf pointed out the value of coming away with something at the end of a good drive (something that certainly seems pretty unimportant), Gumbel noted that the Broncos would get "tremendous momentum" by stopping the Chiefs on a failed fourth-down attempt (something that doesn't appear to be true).
Now, the implication there is clear: By going for it, the Chiefs would retain the momentum of their early start, during which they had kicked a field goal, forced a three-and-out, and then returned a punt into Broncos territory before producing a 33-yard drive to get the ball to the 4-yard line. The field goal would put the Chiefs up 6-0 against Peyton Manning with 49 minutes to go. Does it seem likely that a 6-0 lead would hold up against one of the league's best quarterbacks with that much time left on the clock? Does that thought really require hindsight?
Furthermore, is it clear that the Chiefs really have the momentum? The Chiefs were 1-9 heading into the game, in the middle of a season that had been rendered totally irrelevant and suiting up for a lame-duck coach, Romeo Crennel. The Broncos were 7-3 and riding a five-game winning streak into Kansas City. The Chiefs had lost their previous seven games. Up until kickoff, if you believed in momentum, you would have sworn that the Broncos had all the momentum and the Chiefs had absolutely none going for them. Are the events of what happened during those first 10 minutes enough to wipe out the massive momentum gulf between the two teams from before the game?
And, on top of all that, isn't the field goal really a momentum boost for the Broncos? They're the ones whose defense took over on a short field and forced the Chiefs to kick a field goal. The Chiefs offense has to go to the sidelines and know that they couldn't punch the ball in from 37 yards out and that their coaching staff thinks they need all the field goals they can get to try to win the game. If you were Jamaal Charles, would you feel like you had a ton of momentum if your coach didn't think you could get two yards? In this case, the supposed momentum shift requires your players to be impossibly naive to the task at hand.
• Momentum scales to fit any argument you possibly want to make over an indeterminate amount of time: Let's say that you wanted to make the case that the Puig arrival helped the Dodgers drastically improve in the National League. You can follow along with that argument by looking at his game log here. Depending on when you were writing your article or attempting your analysis, you would have found it either incredibly easy or impossibly difficult to make the case that Puig's arrival turned around the Dodgers.
Consider that Puig showed up when the Dodgers were 23-32. After his arrival, Los Angeles promptly won four of their first five games, with Puig starring in many of those games. You could have argued then that Puig's arrival was a momentum shifter and gotten few complaints. Puig continued to play incredible baseball, but after that early stretch, the Dodgers went 3-9 in their next 12 games, leaving them 7-10 with Puig in the lineup. The opposite would be true now: Instead of suggesting that Puig's promotion had stoked a fire in the Los Angeles clubhouse, you'd have to say that Puig's brilliance simply wasn't enough to get a dormant team going.
Since that 3-9 stretch, the Dodgers have promptly gone on an incredible 41-10 stretch, leaving their record with Puig on the major league roster at 48-20. Now, if you were to argue that Puig's promotion wasn't a turning point for the team, you would be laughed at, and probably with good reason.
Think about how the selective endpoints and perspective change the story that's being told. Puig's arrival goes from being the cause of a turnaround to a success existing in spite of a turnaround and then becomes a cause again. It's the emotional spark that drove the Dodgers to turn things around, even though it took them three weeks to actually start winning far more frequently (despite what Puig was doing to try to carry them to wins during that 7-10 stretch). If it was a momentum shifter, and then it wasn't, and then it became one again, was it ever really a momentum shifter at all?
OK. You're probably sick of reading the word "momentum," and I have to admit that I'm reasonably sick of typing it. So let's move on for now. But during this upcoming NFL season, do me this one favor: When you hear an announcer talk about a momentum swing in the middle of a game, or hear a coach talk about how important is to keep the momentum in his team's favor, or hear a player talk about how his team is gaining momentum by doing this or that, try to follow that argument through to its logical conclusion. See if the game-changing momentum swing actually changes things after all. I suspect that you'll find it to be a case of "nomentum" more often than you might think.