The inspiration for this article comes from the caption of this past S7S post. Please keep in mind that this is satire, and if Montréal was willing to trade Subban for O'Brien straight-up, I'd do it in a heart beat. There has been a bit of talk lately about advanced stats, and how they can be useful, but they can also be misused. I’m far from being a #fancystats guru, I just want to show that, like any statistics, stats in hockey can be used to demonstrate any point you want when manipulated correctly. As Mark Twain once said, "There are three kinds of lies: lies, damned lies, and statistics."
Note that all stats are for 2012-2013 regular season unless otherwise indicated, and come from behindthenet.ca or NHL.com. Also, please keep in mind that I am making some deliberate fallacies with respect to advanced stats use, such as using Corsi QoC and ignoring Relative Corsi QoC.
Jim O`Brien is a better hockey player than P.K. Subban. I know, I know, Subban won the Norris trophy as the league’s best defenseman this past season (quite deservedly), and Jim O’Brien was a healthy scratch for the second-half of the season and most of the playoffs. But if you look at the underlying statistics, I think you may just end up agreeing with me.
Last season, both played on teams that punched above their weight class. The Canadiens had finished in the running for a lottery pick the season before, and on paper didn’t look much better. The Senators, most pundits believed, were lucky to have made the playoffs in 2011-2012, and were expected to regress back to the mean. However, Subban emerged as a number-one defenseman on a division-winning team, and O’Brien lost his job to rookies Jakob Silfverberg, Mika Zibanejad, and Jean-Gabriel Pageau, as well as enforcer Matt Kassian on an injury-depleted squad that goaltended their way into the playoffs. On the surface, it seems that Subban came out on top in terms of performance in
2012-2013, but many of the available stats disagree with this conclusion.
We can start immediately with Corsi Quality of Competition, or QoC. The higher the QoC, the tougher the opponents your coach puts you out against. O'Brien comes in at an underwhelming -1.253, while Subban comes in ahead at -0.890. However, we should let that sink in for a bit. After all, Subban is a number one defenseman, and his Corsi QoC is negative - how much confidence does his coach really have in him as a defenseman? Meanwhile, O'Brien, as a fourth-line centre, faces only marginally less difficult competition. The small difference shows that Subban can't handle much more difficult competition than JOB.
Next, when looking at PDO, Subban's came in at 1023, while O'Brien's was 985. A PDO greater than 1000 suggests a good deal of luck helping out a player, which would lead one to think that their stats should regress sometime in the future. So clearly, Subban was lucky while O'Brien was not, and this was with Subban facing only marginally more difficult talent than O'Brien. Not looking so hot.
When looking at zone starts, Jim O'Brien had 47.1% offensive zone starts, and 50.0% offensive zone finishes. In other words, he was used slightly more as a defensive centre, and came out on top in this regard. That's what you hope for from your fourth-liners. Alternatively, Subban had an offensive zone start % of 53.6, and a finish of 49.3%. In other words, he was detrimental to getting the puck out of his own zone. His minutes were slightly sheltered, against fairly bad competition no less, plus he was rather lucky, and he faltered anyway. It looks like his coach didn't have much confidence in him, and he still couldn't live up to the expectations. Offensive zone starts against weak competition are given to a shoot-first sniper, or a rookie being eased into the lineup, not a top defenseman. JOB seems to play much better defense.
O'Brien had an on-ice save pct. of 0.940, while Subban's was 0.926. Save percentage is always boosted by stronger defensive play, so O'Brien is clearly more reliable in helping to clear his goalie's crease after rebounds. Subban finished the season with a plus-minus of +12, while O'Brien's was -2, but as any good stats guru can tell you, plus-minus is meaningless. Subban dominates this traditional yet meaningless category, which begins to show us why we probably expect Subban to be the better player. "He's so much better!" we say when we watch the games, yet the numbers show us that he is less defensively reliable and much luckier than JOB.
When looking at their goals per game, O'Brien comes in at 0.172, while Subban comes in at 0.262. However, all is not as it seems. Subban played far more minutes than O'Brien this past season. JOB comes in at 1.15 goals/60 min and 1.39 points/60 min, while Subban comes in at 0.96 goals/60 min and 3.33 points/60 min. JOB appears to be the better goal-scorer, while Subban appears to be the better point-getter. But, consider that Subban got 7 goals and 19 assists on the powerplay, while JOB got a whopping 1 goal and 0 assists. Any follower of advanced stats can tell you that special teams tell you nothing about a player's value - of course you get more shots and points on the powerplay! That's akin to saying you belong in the NHL! So only looking at even strength, O'Brien comes to 0.90 goals and 1.12 points per 60 min, while Subban gets 0.33 goals and 1.00 points per 60 min. O'Brien's goal-scoring dominance becomes obvious, while Subban's point-getting dominance evaporates. And shouldn't a top-pairing, offensive defenseman who is lucky and sheltered be getting much better numbers than an oft-scratched fourth-line centre playing in a more defensive role?
Speaking of which, we should take into account their roles. Subban is a number one defenseman who plays all situations, including the powerplay, penalty kill, and press box when he wants more money. He is expected to generate offense. After Jason Spezza got injured, Jim O’Brien spent the season as the fourth-line centre (behind Kyle Turris, Zibanejad, and Zack Smith), later relegated to the fifth centre (Pageau), and soon-to-be sixth (when Spezza returns). Production of 0.17 goals/game from your sixth-string centre is highly impressive.
Let's forget the advanced stats for a bit, and look at fighting. According to hockeyfights.com, PK Subban has registered 10 NHL fights, 2 AHL fights and 5 OHL fights in his career, while Jim O'Brien has no registered fights. This despite the fact that both are 24, so have played about the same amount of competitive hockey in their lives. Fighting leads to one player sitting in the box for 5 minutes, and for a player of Subban's calibre, this shows selfishness by being willing to put his team down by their top defenseman for a full five minutes. Now I find fighting as entertaining as the next person, and I'm not arguing that fights on your resume are inherently bad, but I would much rather see my team's O'Brien than my team's Subban sit in the box. How happy would you be watching Karlsson drop the gloves, risking injury, and putting his team without his services for 5 minutes? Plus Subban has shown a willingness to risk suspensions by instigating a fight with Turris, then continuing to punch him after the whistle blows, at the end of a playoff game out of reach, when losing the next game would put your team on the brink of elimination. Not to mention that in these past playoffs, Subban put up 3.3 penalties taken per 60 to go along with only 0.7 penalties drawn per 60. Clearly he is not a playoff performer, but rather one who is selfish and gets easily frustrated. Looking at JOB's 2012 playoffs (the last year he played), he put up 0 penalties taken per 60 and 1.4 drawn per 60. He knows how to aggravate the opposition without taking penalties himself, clearly a prime playoff performer. Besides, if we're talking about JOB in the playoffs, you all know what link this is.
In general, JOB’s contributions to his team include being entertaining, having a good attitude, and making people laugh on Twitter. PK Subban gets his teammates riled up by yelling at them and as seen above, risks suspension to get even. These may not be advanced stats, but I’m sure you can all read between the lines and see what these say about who you’d rather have on your team.
All in all, it’s obvious PK Subban leaves a lot to be desired in his position, whereas O’Brien plays far better than a press-boxer (aka nouveau Brian Lee) should. Also, looking at the fact that O'Brien has 70 total NHL games played (regular season and playoffs), it’s quite possible that O’Brien is simply a late bloomer, and his best seasons still lie ahead. Subban has 228 total NHL regular season and playoff games played, and as such is much closer to his ceiling. Subban relies on sheltered minutes (during which he still falters), luck, and easy competition to generate fewer even-strength points (and far fewer goals) than O'Brien, all while being a poor teammate and significantly inferior playoff performer. If there was a Norris trophy for healthy scratch forwards, O'Brien would be a shoo-in if there is any justice in this world.
The bottom line is that PK Subban just can't compete with Jim O'Brien. If you don't like the conclusion, don't blame me; blame the numbers.