Picking the 2019-20 Hart Trophy Winner Using Analytics

It's not all about the point totals.

Predictions for the 2018-19 Hart Trophy - Last Word on Hockey

Years from now, we probably won’t remember much about the 2020 Hart Trophy race. Sure, we’ll remember that Leon Draisaitl won it for winning the scoring race by 13 points, but will we be able to recall the months of vitriol on all corners of social media, the hurt feelings, the data visualizations, the eye test, etc.? Probably not. So let this stand as a permanent record of one of the most vicious MVP races in recent memory.

You might have already guessed that point totals aren’t what I’m gonna be looking at here. I’ll be using a number of stats including EvolvingWild’s Wins Above Replacement and RAPM stats (which estimate the isolated impact of a player on goal-driving and play-driving stats) as well as other stats from NaturalStatTrick like goals for percentages and scoring rates. Something I will value (although it’s not my be-all-end-all) is the percentage of a team’s total wins above replacement a player provided. This, to me, is a pretty good proxy for how crucial a player was to the success of his team.

While this shouldn’t be necessary, I will also remind readers that I do not have personal grudges against their favourite player/team/city, and I am not part of a conspiracy to deny them trophies: I am just a simple neighbourhood stat man.

Who is Not On My Ballot

Leon Draisaitl - C, Edmonton Oilers

% of Team Wins Contributed: 15%

Alright, alright, let’s get into it. Leon Draisaitl had an impressive season, there can be no doubt about it. He is one of the league’s best players, certainly one of its best offensive players, a dynamic powerplay asset, and the league’s best finisher over the past three years. But he isn’t on my Hart ballot.

When we’re talking about the cream of the crop, the absolute best players in the league, the most valuable players, every part of the way you impact your team counts. Which means we can’t just look at the points and the offence - we have to face the fact that Leon Draisaitl was one of the worst defensive forwards in the NHL this season. He was brutal defensively at even strength, where his run-and-gun style allows opponents to counter-attack and gain excellent scoring chances - finishing in the 3rd percentile in RAPM expected goals against and 2nd percentile in actual goals against. In his role on the penalty kill, often cited by Oilers fans as evidence of his hidden defensive talent, he was 4th worst among forwards in expected goals against per 60 minutes at 4v5 - especially appalling considering how strong the Oilers PK was aside from him. Did his defensive numbers improve after January? Yes, he was merely below-average instead of one of the worst. But unfortunately for Draisaitl, 2019 is a part of the 2019-20 season; I’m judging a body of work, and it is bad overall.

Offensively, look. His talent is absolutely undeniable - the way he plays is clear based on the huge gap between his shot attempt driving (below-average) and his goal driving (exceptional). He has proven wrong every doubter who said that he was purely a product of McDavid and not worth $8.5M per year. At the same time, it is ridiculous to deny that his powerplay production - the edge that gave him the scoring title - owes quite a bit to getting to play with the best offensive player in the world. Draisaitl ranked third on the Oilers (and in the NHL) in PP Wins Above Replacement, the only non-points powerplay metric we have. McDavid was just as - if not more - central to that powerplay compared to him.

I understand why Oilers fans might be upset by the anti-Drai-for-Hart thing, because they feel like the goalposts keep changing - first it was the arbitrary playoff cutoff in 2018, now suddenly points don’t get you the Hart? This isn’t an anti-Edmonton conspiracy, it’s a reflection of new ways we have of looking at the game that aren’t restricted to what you’ll find on the back of a hockey card.

Nathan MacKinnon - C, Colorado Avalanche

% of Team Wins Contributed: 12%

This was a tough one. As a Haligonian who was in the Metro Centre when MacKinnon and the Mooseheads won the QMJHL championship (and who actually took the SAT with him), I would like nothing more than for him to win an MVP trophy. But he’s not quite there for me.

A lot of MacKinnon’s case is similar to Draisaitl’s, with differences reflected by the way they play the game. Unlike the Oiler’s run-and-gun rush game, which leads to lots of goals on relatively few shot attempts and dangerous counter-attack chances, MacKinnon drives offence at an elite rate across the board at even strength, possessing the puck, generating excellent scoring chances for him and his teammates, and using his talent to turn them into goals at an even higher rate than Drai. When he’s on the ice he makes his teammates better in a way that Draisaitl doesn’t. He’s also stronger defensively - not strong defensively, but about average at least. The Oiler’s advantage over him comes from finishing talent and especially on the powerplay, where WAR is pretty ungenerous towards him despite strong production.

MacKinnon was certainly one of the league’s best players this year, and it wouldn’t surprise me to see him end up on the ballot. Where he misses the cut me is that the top three were not only better, but also more critical to their teams’ success. The Avalanche were the league’s best WAR team this season, thanks in large part to excellent depth. The same is not the case for the other teams represented on this list.

The Nominees

#3: Elias Pettersson - C, Vancouver Canucks

% of Team Wins Contributed: 27%

This one might be surprising to those who only care about point totals - Pettersson finished at less than a point per game, and second on his team in scoring behind J.T. Miller. How can he be a candidate for most valuable player if he wasn’t even the best on his own team?! Once you get deeper than the HockeyDB page, though, things get clearer.

Pettersson plays fewer minutes than a lot of superstars, due to Travis Green’s overreliance on Bo Horvat. What he does in those minutes (which are still clearly first-line level) both in terms of rates and cumulatively is special. He finished 2nd among skaters in total WAR and 1st in WAR/60, and above the 81st percentile in every statistical category except shot-attempt driving. His defensive game is underrated, finishing a very respectable 84th percentile in preventing quality chances against using RAPM metrics adjusted for quality of teammates and competition.

Also in Pettersson’s favour is how important he is to the Canucks. His 4.2 WAR represented over a quarter of the team’s total WAR; on a team that struggles to generate offence overall, Pettersson’s contributions to goal driving, quality chance generation, and the powerplay are absolutely invaluable. On top of that, he doesn’t give any of it back in his own end. No sophomore slump here.

#2: Artemi Panarin - LW, New York Rangers

% of Team Wins Contributed: 36%

Words can’t really do justice to how dominant Artemi Panarin was at even strength this year, and how much the New York Rangers relied on him offensively. So I’ll let the numbers do the talking. At a full season pace, Panarin’s 23.5 even strength offence goals above replacement would have surpassed Connor McDavid’s 2017-18 mark to become the highest since data exists (2007-08). He led the league in even strength points, even strength points per 60 minutes, and even strength assists per 60 minutes while playing primarily with Ryan Strome and Jesper Fast. He was borderline elite on the powerplay without playing with the league’s best player. And he did all of it while being an above-average shot attempt suppressor and a very strong scoring chance suppressor as well.

When he was on the ice, the Rangers outscored their opponents an insane 66% of the time - 75 to 38. Without him, they got outscored by 0.6 goals per sixty minutes. The team accrued 12.3 WAR this season in total, and Panarin was individually responsible for 4.4 of them, the highest percentage among skaters playing for a team that made the play-in. Considering how some people feel about giving the Hart to goalies, Panarin would be a great choice. But he’s not my guy.

#1: Connor Hellebuyck - G, Winnipeg Jets

% of Team Wins Contributed: 38%

Without Connor Hellebuyck, not only are the Jets not in the play-in round, they’re not even close to it. They were a horror show defensively this season, 3rd last in the league in terms of expected goals allowed per 60 minutes and last (behind even the Red Wings) in expected goal percentage with a meager 43%. They got outplayed a ridiculous 75% of the time. And yet they have a chance to win a Cup, and it’s all because of their goalie.

Hellebuyck had a harder workload than any goalie in the league this season - he played the most games, faced the most shots, and withstood the most expected goals. And yet he improbably stood his ground, finishing with almost 20 goals saved above expected, the highest or second-highest (depending on your preferred goalie model) save percentage above expected, and the 3rd highest quality start percentage. He didn’t get fatigued as the season wore on; in fact, he was even better in February and the first half of March than he was in October and November. All told he contributed 4.8 wins to the Winnipeg Jets, of their meager 12.8, leading the league in that category. Rarely do we see a goaltender singlehandedly prop up such a dysfunctional team and get them into the playoff race to this extent, and for that I believe he deserves to be the first goalie to win the Hart trophy since Carey Price.