Momentum vs. Confidence
Hey there, readers. We apologize for the lack of new content lately. We're busy finishing cleaning up our 2013 game charting and dealing with some other things, and the Adjusted Games Lost numbers are coming soon. In the meantime, I was cleaning out my old e-mail and came across this e-mail discussion between Football Outsiders staff members that took place right after the Colts' legendary comeback against Kansas City in the first round of this year's playoffs. It brings up some interesting points about the concept of momentum, and how it relates to player confidence, and what that might mean in a sport where there are 22 men on the field at the same time. I'm putting it out there for you to all discuss in the comment thread.
I hope that the reader whose e-mail to me is quoted below does not mind. -- Aaron Schatz
Aaron Schatz: Some of you may have followed me on Twitter Saturday arguing with people about momentum. I challenged some readers to come up with actual proof of momentum, involving a definition and then statistical analysis. I had a volunteer, but he didn't quite do that. Which is too bad, because like would be a lot easier if we could actually prove that momentum exists. It feels like it does, right? Here's his e-mail:
Paul Petrone: I’m writing this article after watching the playoff game between the Chiefs and the Colts. The game was a great one; Andrew Luck has led his team back to a win after falling behind by 28 at the beginning of the third quarter.
The game had plenty of plays and plenty of arguments for momentum and against momentum. It seemed like Luck had all the “momentum” in the second half, but he also threw a devastating third quarter pick. The Colts played well, but they also got some luck of a fumble coming right to Luck that he ran in for a touchdown. There are arguments on both sides.
That said, while I certainly believe momentum is used all-to-often in sports as a tired metaphor by announcers, I do believe it exists to some degree. The key is to clearly define it.
So What Is Momentum?
Having momentum is the same as “being in the zone,” or “feeling it,” or whatever other phrase that describes playing really well for a sustained period of time. That’s it. We’ve seen it in just about every sport: a hitter who can’t help but hit a line drive, a shooter who can’t miss a three, a quarterback who throws for 300 yards in a half.
Okay, so what really is the reason for those hot steaks? Is it the talent of the player? Is it just the randomness of doing anything, that after awhile it is fair to assume that people will both have hot and cold streaks? Yes, that’s part of the reason. But also, your emotional state plays a role.
So How Does Momentum Play into Sports?
When things start going your way, you begin to feel more confident. For example, when you have a good day at work, you generally have a good day at home. And conversely, when you have a bad day at work, you have a bad day at home.
Over a long period, our emotional state can have huge impacts on our health. Study after study shows that people who are consistently negative are more likely to get sick and have shorter lives than generally happy people. Why? Our emotions matter and have a significant affect on our bodies.
Let’s bring it back to sports. Let’s say a quarterback has a bad game, or a string of bad games. What do you begin to do? Question yourself and your confidence level will begin to fall. That will affect your play, both physically and emotionally.
On the physical standpoint, when things don’t go your way, you tend to change what you normally do. If you throw a few picks going deep, your going to have an aversion to throw deep (unless your particularly mentally tough). If the defense reads that, they can take advantage and sit on short routes. You might try to do different things, change what has worked for you in the past and play worse.
However, if you have success throwing deep on a team, your confidence will rise. You won’t question the game plan, you’ll do what you’ve been taught and continue to do it well. Physically, your game won’t change, which increases your chance for success.
Also, if you feel good emotionally, you’re going to play better. For the same reason people who are negative throughout their lives get sick more often, people who feel good about themselves will play better. The better you play, your better you’ll feel emotionally, which will only help you play better.
Is Momentum Overstated?
Yes. I don’t believe the emotions of professional players change on a play-to-play basis. But do I believe it can change if a player has a terrible half? I do. Do I think it can improve with some coaching or a great play? I do.
I also believe players are stronger mentally than others and are harder to rattle as well. And I believe that can separate players of equal talent.
So, in conclusion, do I believe your emotional state matters in doing anything in life, particularly something as emotional as sports. Is it overstated? Yes. But is it still a factor? Yes.
Here was my response to the reader:
This is not the kind of article we were looking for. It's just a subjective explanation of what you think momentum is, but not in any way a test of its existence.
What we want is the following:
- a) A definition of what momentum is: in other words, after Event X, team should play better than otherwise expected.
- b) Statistical analysis showing that this is actually true.
Cian Fahey: Whether momentum exists or not is irrelevant to me. It still affects the belief of those who do believe in it, so it's important regardless. Belief/Confidence is one of the most important aspects of playing sports. It's less an issue at the top level because most of those playing at that level have huge confidence/egos. When I played sports, at a much lower level obviously, the effects were always obvious.
Matt Hinton: "Momentum" was everywhere during the BCS title game, especially after FSU rallied from 21-3 down following the fake punt. Even Nick Saban cited it – hey, coaches have to make bold calls like that sometimes to spark some momentum. Mack Brown sent out at least a dozen tweets about momentum. Players and coaches definitely believe in it.
That's a little surprising to me because it seems like they'd be more in tune to the mechanics of what's going on, which they can understand and control, than a vague emotional state subject to lord knows what. But the emotional/effort aspect of it is a very real part of it for them, and thinking of performance in terms of probabilities and degrees of randomness isn't very productive if you're the performer, so they attempt to define it and control it to whatever extent they can.
[ad placeholder 3]
Peter Koski: Momentum is a psychological crutch used to validate the result of a series of random events because people are uncomfortable that we live in such a universe where events unfold the way they do "just because."
What constitutes momentum in an NFL game? When exactly does it begin? Momentum's origin is always so easily marked in hindsight, yet remains very tricky to locate in real-time. It seems that momentum's origin can never be pinpointed and instead exists as a Schrodinger's Cat that either exists or does not exist until you go back to measure it from some future point in time. In Green Bay on Sunday, the 49ers defense forced three straight 3-and-outs and held a 6-0 lead after the first quarter and were driving into Green Bay territory to start the second quarter. They had "THE MOMENTUM!" However, after riding that momentum 30 yards into GB territory, Kaepernick threw an interception and GB drove to score a touchdown and took the lead 7-6. So, was Momentum tricking us into thinking SF had the momentum, when really GB had the momentum all along because they only allowed two field goals despite being massively outgained offensively? Well, the 49ers scored five plays after GB's touchdown to grab the lead back, but did they also grab the momentum? The "Fumble Luck" touchdown was not random, it was the Momentum's righteous providence! These are but two of the enumerable examples that we see each week exposing the folly of "momentum" in the NFL. We know what momentum means within the NFL lexicon: One team's victory was possible because they achieved a series of successful outcomes that overwhelmed the opponent's ability to counter. The true alternative is frightening to many fans and leads to them to cling to "momentum;" events unfolded whose results were heavily influenced by randomness and ultimately outside direct control of either team to a certain degree.
There's a definite confusion between confidence and momentum. When negative events unfold and confidence is lost, it's possible that focus is also lost and momentum is observed as a self-fulfilling prophecy. All it takes is one play to "swing momentum," which I think in itself invalidates the definition of momentum.
Danny Tuccitto: I chimed in on Twitter at the time, but I will expand here. If we define momentum as "A player having success on Play A results in increased confidence on Play B, which leads to increased likelihood of that player having success on Play B (or vice versa)," then yes, momentum exists. The sport psychology literature is clear on it. The problem is that, in football, you have 11 vs. 11, and so these positive (or negative) feedback loops are diluted across 22 players, not to mention backups and coaches. That's why, when someone like Bill Barnwell (read here and here) or Brian Burke (read here) does an admirable job of trying to relate presumable momentum-producing situations to subsequent performance, they find no evidence of momentum in the data. With so many interactions, the overall team-wide effect seems to me to be practically imperceptible. There are situations where you have, say, most of the players on the field for Team A experiencing the positive feedback loop, while most of the the players on the field for Team B are experiencing the negative feedback loop (i.e., the optimal situation in which to find evidence of momentum), but those situations are so few and far between that they get drowned out by the "it's a wash on the field" situations.
[ad placeholder 4]
Scott Kacsmar: I think momentum works much better in basketball or hockey where it's a fast, back-and-forth game with fewer players on the field. When you watch a scoring run, especially from the home team, in a NBA game, momentum swings are undeniable. But measuring this in football becomes almost impossible because of the mini-breaks that take place between plays, and as Danny said, the 22 players really complicate things. It only takes one small mistake from one guy to throw off an entire play. Someone like LeBron James can practically go on a 9-0 run by himself by dominating both sides of the ball.
Danny Tuccitto: If I could offer a practical example for my point, consider something like "wide receiver A is destroying cornerback B on every play in the first quarter" or "pass rusher A is destroying offensive lineman B on every play in the first quarter." OK, so the WR/OL are in super-confidence mode and the DB/DE are in super-woe-is-me mode, and the former are probably screaming to the heavens on the sideline that they're owning their guy 1-on-1. But then what? How's the QB's confidence doing? How's the opposite WR/OL's confidence doing? How likely is the offensive coordinator to damn the torpedoes and keep feeding the ball to that WR or running behind that OL on every play until it stops working? There are just too many things that need to come together to make one guy's "momentum" translate to perceptible (data-wise, I mean) team-wide, momentum-based success.
Aaron Schatz: Actually, I think Peter's point about confidence and momentum being confused for each other combines well with Danny's points. First, that there's so much noise with 22 (actually, 90 including backups on both units) emotional feedback loops going on at once. Second -- and this ties into Cian's comments -- I would suggest it takes a lot more to shake the confidence of a professional athlete than it would to shake the confidence of a high school athlete or something. (By professional, I would also include the top college athletes.) And thus, players' lack of confidence is unlikely to heavily affect their performance, even when they are losing by three touchdowns. And, it follows, this helps explain why we have never found evidence in NFL play-by-play that big plays naturally lead to more big plays, and failure naturally leads to more failure.