Either Way

Let’s go back in time a little.  Specifically, let’s go back to around 4 am Eastern Standard time on January 3rd of this year.  With me?  Cool. Let’s watch some TSN. Or, I guess, we could just go back to sleep, but if that’s how you roll then this probably isn’t your post, and anyway what the fuck is wrong with you, that you want to use time travel to sleep through hockey games six days in the past? If you don’t want to watch TSN, just go get your own time machine and kill Hitler, why dontcha?

Anyway, in my time travel experiment, Canada hasn’t lost to the USA in the semifinal yet. Canada hasn’t even started to play the semifinal yet. Rather than being three hours of nearly-constant misery culminating in the most embarrassing sort of loss, the game is still a space of bright possibilities. In this time we’ve gone back to, Canada might still beat the Americans like so many rented mules.

In this cheery past, the TSN commentary is optimistic. While, like smart hockey people always do in single-game-elimination situations, they’re hedging their bets (gotta control the puck, need strong goaltending) they’re also looking for the positives, and in doing so, they come up with this: by virtue of their bye to the semifinals, Canada has had more rest. The US has played six games in eight days. They’ll be worn down. This is an advantage for team Canada. The time off will make them better than their weary opponents.

Hockey discourse never puts out only one narrative. That’s part of the bet-hedging- always go into a game with four or five storylines and you’re sure to have one of ‘em pan out. In broadcasting, and even in certain kinds of writing to a deadline, an analyst doesn’t have the luxury of wait-and-see. He’s gotta emplot this shit in real time, and that means keeping more than one plot in play as long as possible. Before the game, this particular story- the well-rested, relaxed Canadians vs. the tired, harried Americans- is just one thread of explanation being laid down in anticipation of a possible outcome.

However, while most advance storylines get dropped the second they cease to line up with events, this one doesn’t.  As the first period oozes on and Canada oozes across the ice like so many stoned jellyfish, the color commenter doesn’t just abandon the whole issue of how rest and time off are supposedly so invigorating.  Nope, he reverses it.  Rather than talking about how good it is to get a break and how exhausting it is to play a lot of games, he says something to the effect that Welp, you know, sometimes it’s real hard when you have these days off, it throws off your rhythm, you get rusty.  The Americans are really prepared because they’ve been playing so much. Canada’s not ready.

See that?  See how that works?  This isn’t just any old hockey narrative here.  This is a MAGIC REVERSIBLE HOCKEY NARRATIVE.  Like one of those dresses you can wear that’s green on one side but has giraffes on the other, it’s perfect for both a wedding AND a safari!  Time off: either it makes hockey teams better or it makes them worse.  Whichever way they played, the time off is a big factor in why they played like that.  Team Canada could come out onto the ice and play lacrosse, and time off could be cited as an explanation:  well, you know Bob, sometimes with a couple days off you can forget that you’re a hockey player.

By the time the game is over, “time off” will have been dropped almost completely as an explanation. The post-mortems on Team Canada will talk about goaltending and coaching, team composition and team cohesion, but they won’t often cite Canada’s bye and the American’s quarterfinal as an explanation of the outcome. They’ll give it up and move on to more fruitful issues, or at least ones that provide more opportunity for a scathing critique of the nation’s relationship with its hockey and its teenagers. But the magic, either-way power of “time off” as an explanation isn’t gone forever.  In fact, with some players coming back from playing half a season in Europe and others having done nothing but sit in Kelowna eating Cheetos for three months, I’d bet that the first few weeks of the season will be filled with appeals to the Magic Either-Way Power of Time Off.

So here’s a question: what is this whole scrap of narrative about the impact of breaks vs. not-breaks doing here? What’s the point? What possible role does it fill? Do we really need an explanation that does nothing but tell us that sometimes things are one way but other times they’re a different way?  Because I could believe that without an explanation.  You know, Bob, sometimes Southern teams are better than Northern teams, but other times they’re not.  It depends.  Teams with more Ryans are often better than teams with fewer Ryans, except sometimes Ryans have been known to be a liability.  It depends.  What’s that, Bob? Can I tell you in advance when it will be one way and when it’ll be the other?  Why of course I can’t.  Nobody knows that.

The problem with this kind of explanation is not exactly that it’s positing one possible cause of two different outcomes, but that it’s positing that cause in a fundamentally unfalsifiable way. Falsifiability, for those of you who aren’t into theories about scientific method, is one of the major criteria by which scientists distinguish the shit that is part of science, like astronomy, from the shit that is not part of science, like astrology.  The basic idea is this: in order for a theory to be considered a scientific theory, it has to be able to be proven wrong.  Not actually proven wrong, mind you, but able to be so.  The advocates of the theory need to be able to state the conditions under which it would be exposed as bullshit.  Moreover, whenever possible, they should conduct exactly the kind of experiments or make the kind of predictions most likely to destroy their own prediction.  That’s how, according to this method, you know that an idea is better: by putting it to the hazard and seeing if it survives. [N.B.: If you want to know more about this, the dude is named Karl Popper and the paper is here.  One of the best reasons to learn about falsifiability is that you get to say "Popper" over and over again.]

Now, there’s no law that says that narratives have to be falsifiable. People can, and do, believe all kinds of things that are unfalsifiable.  Religion, for example, is explicitly designed to be unfalsifiable- everything is God’s will, no matter what, because if something wasn’t God’s will it’d undermine the very concept of God. In a the lecture linked above, Popper cites Marxism and Freudianism as other examples of belief systems that are unfalsifiable by design.  They are able to subsume anything and everything within their worldview, and see that elasticity as a strength rather than a weakness. Their goal is never to be proven wrong.

Hockey isn’t science, and there’s no obligation to be scientific about it. But at the same time, like the sciences, hockey analysis has certain concrete, definable aims against which we can evaluate its usefulness. Yes, we spend a lot of time bullshitting, but the ostensible purpose of sports commentary is not just to spin out a bunch of crap because it’d be real awkward to have four guys in suits sit in front of a camera in contemplative silence. It’s to provide explanations- specifically, explanations of why and how some teams win hockey games and other teams do not.

An explanation that can fit any outcome equally well is an unproductive one.  Saying “sometimes more rest helps and sometimes it hurts” doesn’t get us any closer to understanding why a team wins or loses, and- as TSN has shown us- as it stands, it doesn’t seem to be helping us make better predictions.  It’s useless knowledge, if indeed it’s knowledge at all.

Does this mean that time off means nothing? Maybe. We know that playing back-to-back games have a negative effect on team performance, but to my knowledge the ideal length of spacing between games- and whether there is a point where more time off leads to worse outcomes- hasn’t been well-researched. It’s possible that it doesn’t make no nevermind at all. But, then again, as humans, we know that rest and rust are two sides of the same coin, and that both can and do happen to people. It’s perfectly possible that there is some validity to the explanation.

But that validity, if it exists, isn’t in the explanation as currently presented. “Well, sometimes rest makes teams better, but sometimes it doesn’t” may be true, but it gives no meaningful information. It’s like saying, “Well, sometimes I’m happy, but sometimes I’m not.”  Okay, well, good for you, but unless you’ve got some idea about, you know, the circumstances that make the difference between those two situations, it doesn’t exactly qualify as an insight.  It’s not going to make you happier, nor help me make you happier, and frankly I don’t even understand why you bothered to tell me such an inane thing.

If we wanted to avoid inanity, we could go so much deeper into this whole issue of how time off affects team performance. We could do studies on the length of breaks between games and look for patterns in outcomes. We could interview coaches and trainers about the conditioning strategies they use to manage down time and look for correlations between those strategies and on-ice performance. We could do kinesthetic or psychological studies of the physical or mental effects on players of playing more often vs. less often, and try to develop a typology of human responses that would lead to the building of a rustproof hockey team. And, if anyone really believed that the reason Canada lost that semifinal game was because they didn’t play a quarterfinal, I’d expect exactly that kind of research to be matter of intense debate and advocacy right now.

That nobody really seems to interested in doing this work- and particularly, that “time off” explanations are offered so often during games and so rarely developed after them, as we’ve seen with the Canada loss- suggests to me that nobody actually believes this explanation. It seems plausible, it’s not controversial, it’s easy to say, but it carries no substantive weight. Nobody is actually satisfied with it as a causal explanation of the outcome of hockey games. It’s just a thing to say.

So maybe this is a case of an annoying but basically harmless unfalsifiable truth statement, the hockey equivalent of “God works in mysterious ways”. Dudes have probably been saying for decades “oh, team X will be better than team Y because rest” and then reversing that shit the second team X goes down a few goals, and as far as I know I’m the first person to get my panties in a twist over it. After all, it don’t hurt no one, right?

Not in this case, no. But there are categories of unfalsifiable explanations in sports that do have painful human consequences. Consider, for example, the way we talk about coaches’ personalities and team success: if you get a new hard-ass coach and the team does better, it’s because he’s motivating them with his stern tough-loving ways. If they do worse, it’s because they’re cracking under his brutal assholery. There is no condition under which a change in the team’s results could not be interpreted as the consequence of the coach’s temperament. Similarly, if your team is failing, whether your coach is a sweet-natured old grandpappy or a scorching demon from the black pit, his personality will be cited as a problem. If he happens to be nice, he’s too nice. If he happens to be mean, he’s too mean. If he changes his strategies, he’s panicky. If he doesn’t change his strategies, he’s stubborn. Every possible type of coach has tried every possible response to every possible kind of team struggle and not one has ever found a predicable way to avoid being blamed and fired, because hockey discourse’s commitment to the “coach’s personality” explanation for failure is stubbornly unfalsifiable. In the time-off case, the worst consequence of an explanation that can never be wrong is an irritating trope and a bit of wasted time, and the only reason I can offer that you should question it is by appealing to your desire for Better Explanations and, I dunno, Truth, maybe? In the coach-has-lost-the-room case, though, this explanatory gap leaves people vilified and unemployed.  So even if there is not always a good reason to attack unfalsifiable narratives in hockey, there are certainly good reasons to remember that such narratives do not provide useful explanations. Believe them if you want, as articles of faith, but be careful what consequences those beliefs might have.

Comments (12)

  1. Thanks, Ellen. You cracked me up and gave a great thumbnail of Popper. I’m your fan.

  2. Some people are starting to try and figure these types of things out with advanced stats, but the problem is that there are a ridiculous amount of factors that play into the outcome of a hockey game, some out of the direct control of the players (ice conditions, for one, or simple assembly of talent minus those who are injured). There is no real mathematical way to isolate any one factor and analyze that over a large spectrum of games.

    Even in baseball, where you can easily isolate each individual’s impact on a play (unlike hockey with its fluid dynamics), there are factors that are difficult to isolate and control for. Weather is a big one, as are injuries and personal-life problems that may take away from a player’s maximum ability. That’s why statistics and trends are representations of the whole in the past but only show what may happen in a given game or series or season (as opposed to definitively predict what will happen).

    To your point that the amazing reversible analysis is flawed, absolutely. And the better color guys (as well as those not completely bought and sold by their teams or national federations) stay away from that or admit when their theories are wrong.

    But to the point about finding those scientific truths and explanations, I think there are just far too many factors to really do that with any degree of certainty. You can find larger trends or something that runs counter to traditional hockey wisdom, but something that is applicable to every game in every circumstance is never going to emerge.

    • Oh, I absolutely agree that any quest to perfectly distil all the factors that contribute to the winning of an individual game, in all the correct proportions, is pretty much doomed. There’s just too much luck… I mean variance… I mean… chaos? Whatever you call that. But I still think it’s worth trying to figure out such things, because even where perfection is unattainable, “better” might be.

      Also, honestly, I’d be happier if commentators were more comfortable using variance as an explanation, with the understanding that there might be all kinds of lower-echelon contributing causes that play into that, rather than attributing large-scale explanatory power to said contributing causes. (Yay unnecessarily complicated sentence!)

  3. Another narrative: leadership and grit.

    Going into a series, whichever team has won more in the past has this advanatage, once the series is over, you assign more leadership and grit to players on the winning team. Once you have it, you can never lose it. See Mark Messier, the ultimate leader, even though he consistently “led” the highest-priced team in the NHL to miss the playoffs the last years of his career.

  4. Didn’t the Kings have more time off with each passing series last year? And they opened every one of them 3-0? Small sample and all, but it does suggest that Better is probably a more important consideration.

  5. I’m thinking two big, related parts of why the “rust vs. rest” theme is so often brought into game commentary are 1) that it’s those guys job to have something to talk about and 2) that the ambiguity of rust and rest’s effects on teams, rather than making it a useless topic to address, is actually the primary thing that makes it interesting and causes it to continue to pop up.
    But really Ellen, nothing’s perfect and exposing such obvious flaws as the scientific meaninglessness of certain influences on performance and performance tracking statistics that people throw around all the time, in hockey and everywhere else, is a pretty low level of analysis. I must say that your historical articles are far more rewarding reads than your philosophical articles (like this one). It’s like you tried to make a conversation stopper into an actual conversation, and ended up filling space with even more meaningless commentary than that of many commentators’ “Rust vs. Rest” debate.

    • I’m sorry you didn’t enjoy this one, and of course I appreciate that you’re trying to make suggestions that will help me improve as a writer, and I’ll bear that in mind in future.


      I do this to have conversations, ideally conversations that end with me (and, if I do a good job, maybe some other people) learning things about the game that they didn’t know before, or thinking about things in a way they hadn’t previously. And despite your feeling that this was a “conversation stopper”, I had several good discussions both in the course of writing this and after putting it up. So from my perspective, it did more or less what I want a post to do.

      Also, FYI, the history posts are usually tagged as such, so if you want to read selectively in the future, it shouldn’t be too difficult.

  6. I see what you’re getting at, but this kinda misses, Ellen. There is no field I know of in which the use of completely sensible theories and hypotheses is not completely and utterly trashed by media-based TALKING HEADS. Seriously, you cannot listen to the Weatherman and seriously hope to have a solid discussion about global warming. On the whole they are ill-equipped, and they are certainly driven by a whole set of cross-motivations.

    So why would one imagine that sports commentators would usefully apply a theory about the value of rest to an athlete? You wouldn’t. Do such theories have value? Absolutely, unquestionably, they do. You cannot, for instance, run a boxer out into the ring, night after night. Nor a marathoner. Nor a professional baseball starting pitcher. Rest is absolutely required by those athletes to perform anywhere near their capabilities.

    For NHL teams, it gets if’ier. Players with injuries often require additional rest, and in many cases, can go from “unable to play” to “game ready” with just an additional day or two. During the Stanley Cup playoffs, we quite often know – or hear about after the season is done – that a dozen players were carrying on, with real performance-impacting injuries, that rest would help.

    But does the fact that talking headed big mouths then apply this theory to any and all situations, regardless of questions of injury or the nature of the sport and its demands, turn the EXPLANATION into garbage?

    No, it doesn’t. It’s no “unfalsifiable,” at all. In fact, it’d be really nice to see it studied more, to see when and where and how it applies, sport by sport, position by position. As it is, we know it applies – as discussed above – at least in some sports, and in some cases.

    The key is, don’t ever look to media personnel- who get paid for blabbing and throwing and and every sort and type of of shit available at the listening and watching wall – to help you determine which theories work and which don’t.

    • Well, the weatherman/global warming example is maybe a bad analogy, as meteorology and climatology are simply very different sciences. You wouldn’t expect a physicist to be able to give someone an accurate psychiatric evaluation either …

      I think Ellen’s point, though, is not that there’s no impact of time off on performance, but just that the WAY colour commentators (and also many analysts) employ the “importance of/problem with time off” narrative is problematic because of it ends up being entirely unfalsifiable. Used in this way, it is unlikely to help us ever understand the causes behind a win or loss.

      But as she says, “we could go so much deeper into this whole issue of how time off affects team performance,” so I think you’re actually agreeing with her when you say it would, in fact, be really nice to see it studied more. Just studying it more, though, might not be enough – it also needs to be studied and analyzed properly, e.g. through the formulation and rigorous testing of falsifiable hypotheses (perhaps?). But if it’s just going to be some unfalsifiable narrative we can lay over any win or loss, there’s good reason to think that that’s not going to be good enough at all, at least not for those who want genuine understanding rather than just the illusion of such, or those who are just happy to have some noise to fill the dead air while they’re watching a game.

      • I get what you’re saying, but thing is, she’s arguing that because MEDIA TALKING HEADS can’t handle a hypothesis correctly, it has no explanatory power. But media people can’t handle ANY hypothesis correctly, because they’re not trained in anything relevant. They’re there to look pretty, to say shit that lots of people will nod at, to pump up the volume, turn up the heat, create horse-races. And they like having multiple voices on their shows, to help create conflict and also to lay out different narratives. But they care so little about any one theory, and know so little, that they’re entirely happy to just smudge it in both directions. That way, nobody’s ever permanently wrong, everybody gets invited back, no theory dies, no team is ever out, the show must go on.

        Thing is, it’s actually interesting to wonder, MSM to the side, whether a theory actually works or not. I think that’s where we all agree. ;-)

        • I have (perhaps misplaced) faith that it is totally possible that media talking heads can handle all sorts of hypotheses correctly- IF they’re pressured to do so. I just wish I was able to exert more pressure…

  7. Amen, amen: The talking heads drive me crazy with the kind of analysis that gets recycled without any reflection or investigation.

    If some guy can come with his own factor (Corsi number), then perhaps some multi-variable analysis can be done on the whole question of rest vs. no-rest. Perhaps…

Leave a Reply

Your email address will not be published. Required fields are marked *