Imagine how frustrating it would be for a knowledgeable general medical practitioner to pick up a copy of his or her favorite medical journal and read an article extolling the virtues of the humoral theory. Put yourself in a lab coat and pretend there’s a stethoscope resting around your neck as you peruse text proposing that when the four humors – blood, phlegm, black and yellow bile – are in balance, good health is guaranteed. It would be absolutely maddening to read claims that an imbalance in the humors is at the root of sickness.
The existence of more effective measures doesn’t always guarantee the use of more effective measures. While the specific example of a doctor practicing humoral medicine in the 21st Century would likely result in several malpractice lawsuits, our predilection for doing things not by the best means but rather the most familiar continues on a smaller scale that seldom ends up a matter for a court of law, or of life and death.
Consider the relatively unimportant endeavor of sports. When we talk about athletes, our conversations often lead toward opinions on their ranking. Player A is better than Player B because Player A does this. No, you’re an idiot for thinking Player A is better than Player B because Player B has this number of whatevers, and Player A only has that many. And so the world of sports “discussion” turns.
Statistics are the weapons of these duels. Even those dismissive of analytics rely on them. And this is the funniest bit of business to do with the futile arguments that erupt over sports. Once one refers to an advanced metric in support of an opinion, one is most likely to hear a counter argument that suggests numbers aren’t the end all and be all, followed by a less advanced number to prove the opposite point. Your fancy PER doesn’t tell the whole story. He’s averaging 25 points a game.
It seems that traditional numbers are not statistics measuring performance, but rather a generally accepted misunderstanding of tallies that are now mistaken for representations of true talent. As soon as these figures are accepted as numbers, it’s difficult to reasonably refute that they’re not the right numbers. After all, the more advanced metrics simply take more factors into account than the simple and traditional. Doesn’t that mean that they tell more of a story, and isn’t that what we look to glean by quoting numbers?
This is perhaps best seen in the case made for former starting pitcher Jack Morris’s Hall of Fame candidacy. Most metrics don’t support the idea that he’s an elite pitcher deserving of the post-career honoring. However, over his 18-year career, he was credited with winning 254 baseball games. Only 42 pitchers since 1871 have collected more victories in their career, and so, in a lot of people’s minds – the same who rely on pitcher wins to be telling – he was a dominant pitcher.
Judging the abilities of individual athletes who participate in team sports based on the overall outcomes of the entire roster of players is unreasonable because even the very best of performers playing the most important of positions are incapable of controlling every single aspect of a game, and so, as long as they compete in a team sport, the tallying of the group’s results remains misleading to the evaluation of the individual.
But raging against the counting of wins for individual players is nothing new. Its cons have become as much of a crutch to the new school as its pros have been to the old guard. And it’s not as though it’s completely inaccurate. It just doesn’t tell as complete of a story as other measurements. Wins are typically assigned to the player most responsible for a single game’s outcome: a pitcher in baseball, a goaltender in hockey and a quarterback in football.
Ridiculous? Yes, but the degree of ridiculousness varies for each of the three. I understand that attempting to decipher the least stupid of the bunch is a bit like trying to determine the best cancer to have, but tilting the typical Player A vs. Player B sports argument toward a discussion of a different sort is something that at least has the appearance of fun.
And so, here’s why I believe baseball’s starting pitcher to be the most deserving of player wins, and hockey’s goaltender to be the least.
The Starting Pitcher
Without a doubt, the starting pitcher is responsible for a larger percentage of a baseball game than any other player who steps into the batter’s box or plays the field. Unlike any of the other winner/losers, his action is necessary for the game to be played out. However, he isn’t responsible for everything that happens in a game. He’s not even responsible for half of what happens.
Baseball is split up first between offense and defense. The offense is responsible for half an inning and then the defense is responsible for the other half. Already, there’s 50% of a game that a pitcher has nothing to do with. We take defense, and under this category falls pitching and fielding. Fielders are obviously responsible for fielding, further lessening a starting pitcher’s responsibility, leaving us with pitching, which also must account for relievers which brings down the amount for which a starting pitcher is responsible even further. And even within that percentile, there’s an entire element of luck (pitchers can’t control where balls are hit and no hitter is good enough to put balls precisely into play) for which we’re not accounting.
It’s his necessity to baseball’s action, and the corresponding responsibility for each pitch he throws that makes assigning wins to a pitcher the least ridiculous of the three options. Every single time he releases the ball toward home plate, in the direction of the batter, the opportunity exists for the opposition to find a measure of success against the pitcher’s team. This tempting of fate, no matter what the odds say, is a constant flirtation with disaster. His capacity for failure earns him a heavy responsibility, and it’s not all as horribly silly as we often make it out to judge how he handles that weight by assigning wins and losses to him as an individual.
The quarterback is similar to the pitcher in that he’s only responsible for half of his team’s play. Football, like baseball is broken up between offensive and defensive turns. However, a pitcher performs during his team’s defensive stand while a quarterback performs during his team’s opportunity on offense. It’s his duty, not to stop runs, but to lead his team in the scoring of points.
Even within his control of the offense, scoring isn’t the complete responsibility of the quarterback. Last year, NFL teams averaged 34.7 passing attempts to 27.2 rushing attempts per game, meaning that a running back is almost as likely to run with the ball as a quarterback is to throw it. Then, even when he’s throwing the ball, he’s relying directly on another player on his team to catch it. As we’ve established, a pitcher in baseball relies on his defense to deal with batted balls, but its not in as direct of a manner as a quarterback throwing to his receivers as they battle the opposing secondary, and it’s not built into every single play.
All of this also assumes that the play calling and the opposition is of equal stature to both pitchers and quarterbacks. And we’re not even taking into account the roll that kickers and special teams play in offense.
At least with a hockey goalie, he or she – merci Manon Rheaume – actually participates in the entirety of the game (unless they’re pulled for an extra attacker). However, a goaltender, despite constant presence on the ice, isn’t exactly leaned on very often to contribute offensively, which remains an important part of winning a game in the National Hockey League. Still, dismissing their won/loss records isn’t done as easily as it is with pitchers and quarterbacks.
That doesn’t make it any more relevant though.
There are several elements outside of a goalie’s purview that are likely to influence the outcome of the games in which he plays, including the talent, tactics and discipline of the entire team. There’s also the small matter of luck which plays a not-insignificant part in contributing to scoring chances, which ultimately lead to goals.
That’s not to suggest that all goalies perform at the exact same level. Yearly, in terms of “goals versus threshold” those playing the position are listed among the top players by Hockey Prospectus. The only problem is that year-to-year there’s a whole lot of variance among those goalies due to the factors I’ve already mentioned. Even when we do find consistent performances among goalies, there are factors beyond their true talent as a puck stopper that play a large role in explaining their ability to recreate positive performances.
For instance, a lot of what determines our perception of whether a goalie is good or bad is the quality of shots that they face. A team like the Boston Bruins employs a system that limits the quality of the shots their opposition takes, and as such, they’ve consistently had their primary goaltender – whether it was Tuukka Rask or Tim Thomas – among the league leaders in save percentage every year. They appear to be good goalies, but in another system, facing a different quality of shots, those numbers would be expected to fluctuate.
In fact, Tom Awad, a writer for Hockey Prospectus, estimates that a hockey goalie is responsible for about 5% of his team’s winning percentage. Yet he’s attributed with wins and losses on an individual level as though he was 100% responsible.
This is the issue many of us have with player wins. It’s incredibly silly to imagine pitching, quarterbacking or goaltending to be the complete causation of victories and losses in their respective team sports. And yet, that’s what happens when we assign them to players for no other reason than tradition dictating it.
We have improved analytics to help us have a better experience following the sports we love, better methods to help us have better conversations. But if you want those discussions to evolve, it’s important to not come across as arrogant or dismissive about what has been traditionally used. Player wins do tell a story about the players to which they’re attributed. It’s to varying accuracy in each case, but it’s still something, and it’s something on which we’ve built improvements.
The more advanced metrics put more weight on the factors under the control of the individual, rather than accounting for things outside that control. It’s not difficult or revolutionary or perfect. It’s merely better at providing a clearer picture of what we want to determine. It’s moving forward without forgetting the past.