When narrative goes too far, and inequalities fail

Regular readers will know that I'm pretty interested in how narratives shape our perception of sports. One of the most significant ones lately has been in the Vancouver-Chicago series in the NHL playoffs, where the Canucks led 3-0 before the Blackhawks won three straight games to force tonight's seventh game. Of course, that's led to plenty of pieces on how this is a defining moment for Vancouver, a historic occasion and all the rest. Those stories aren't necessarily wrong, as there certainly is a significant mental aspect to sports, and that mental element will be involved tonight; I give it more credence than Joe Posnanski does, even if I share some of his other opinions on storylines. What's happened in the series so far does have a bearing on tonight's game in my mind, so it's perfectly relevant to talk about the pressure, the situation and the rest.

What I don't like is when that analysis takes the next step, though, and ascribes narrative reasons to why one team lost and another won and narrative solutions as to how to remedy this in the future. We've seen this plenty of times before, with certain teams or players being labeled simply as "chokers" for poor performances in small playoff sample sizes, or authoritative declarations that there was some clear flaw in the team that lost; they didn't have enough depth, enough grit, good-enough goaltending or anything else. None of those claims are necessarily wrong or problematic on their own, as it's certainly worthwhile to try and analyze what went wrong and think about how it could potentially be solved. What bothers me is more along the lines of the shades of grey discussion; in essence, any particular claim about size, scoring, goaltending or the rest isn't necessarily wrong and could in fact be right, but pointing to one of those things as the definitive cause of a team's downfall and something that has to be remedied if they're going to win in the future is generally inaccurate.


Many people tend to look at sports as a system of mathematical inequalities. If Chicago beats Vancouver in this series, some claim that's proof Chicago's team > Vancouver's team. From the particular perspective of results in that series, that is correct, but when it's expanded throughout the playoffs, it's more problematic. For one thing, we don't know how teams would have fared against different opponents. Just because team B beats team C, and team A then beats team B, that doesn't mean A > B > C.

Even more importantly, though, it's worth recognizing that every single sports game carries a notable element of randomness. For example, the Canucks finished the regular season with a league-best 117 points, while the Edmonton Oilers had a league-low 62, but if you watched just the Oilers' 4-1 thumping of Vancouver earlier this month, you'd conclude that Edmonton was the better team. It's not just a small sample-size issue, either; in an excellent piece for Baseball Prospectus, Tommy Bennett calculated that even baseball's mammoth 162-game regular season doesn't necessarily tell us which team is the best. Here's part of what Bennett wrote:

hat do we mean when we talk about the best team? The team that had the best regular season record, most likely. But it turns out that performance over 162 games isn’t even enough to say for sure which team is the best. If an entire league with a “true talent” level of .500—that is, one in which God told you all the teams were .500 ballclubs—played a million 162-game seasons, two or three teams would end up with more than 90 wins each time. You’d probably look at those teams and assume they were the best, but we’ve specified that all of the teams were of exactly the same quality. So there’s a real way in which regular season record does a pretty lousy job of telling us which team is the best.

That's applicable to the playoffs too. Playoff outcomes matter, and good for the teams that win in the playoffs, but that's not necessarily proof that they were "better". It's proof that things broke their way on at least four occasions in each series. Part of that's down to their physical skill, part of it's coaching, part of it's mental and part of it's matchups, but a significant part is random, too; it's been found that even the best teams only have about a 55 per cent chance of winning any given playoff matchup. Thus, there's at least a 45 per cent chance that the team we wouldn't describe as "better" (in terms of true talent level) will win.

In essence, the most important thing to take away may be not overreacting to the result of any given playoff game or series. Winning a game or even a round doesn't magically make you an amazing team, while losing doesn't make you an awful one. It goes against the sports talk radio narrative that demands heads of coaches and executives and trades of players when things go wrong, but it's worth keeping in mind that a lot of the outcome of any sports event is pretty random. Sports and math have a lot in common, but I think playoffs are more about chaos theory than straight mathematical inequalities.