Trying to make MVP Shares a thing in the NHL
It's a thing in the NBA and MLB so.... let's see if it's useful in hockey
In the last 5 years, (2018-2022), the NHL has seen 5 different winners of the Hart Trophy (The NHL’s version of an MVP) - while this creates the illusion of parity among the superstars in the NHL, to any fan who has watched the league develop, it’s clearly been McDavid’s league for the entire portion of this time-span.
Introducing MVP Shares (or Hart shares) - It’s a concept that’s well-known in the MLB and NBA. The primary idea of it is to give value to receiving MVP votes, avoiding the 2nd and 3rd place finishes being forgotten to history.
As this doesn’t seem like readily available data, here’s a quick breakdown of the voting process on awards in the NHL. Each esteemed writer is given 5 votes, ranked by weight, on who they favour for the Hart, Norris, Calder, Lady Byng, Selke, and All-Star Teams at the end of the season. To give preference to favourites, they weigh the votes — 10 points for 1st place, 7 points for 2nd place, 5 points for 3rd place, 3 points for 4th place, and 1 point for 5th place. You’ll notice that tallying those points will add up to 26 points - how much weight a writer has on the award vote.
By dividing the total votes a player receives by the total possible votes, you’ll get the vote %.
That Vote % number is the “Award Share.” By taking into consideration the fact that Matthews had an exceptional season (60 goals in 73 games) - it doesn’t ignore the excellence of McDavid’s 123 points in 80 games. Instead of having a zero-sum winner and loser, there’s consistent tracking of where the player finished year to year. Matthews gets .8359 Hart Shares, McDavid gets .5697 Hart Shares, etc.
So. Where does everyone lie? Well, I broke it up by era.
Since the weighted voting system wasn’t implemented until 1982 (read more on that in the notes section below), I used that year as a marker of pre/post-modern award shares.
While each “Hart Share” is functionally an expected Hart Trophy, it’s important to note that Hart shares ultimately end up being more than the amount of Harts handed out (92.6 in the total list over a 40-year span). Because of this, it’s better to use an accumulation of votes to sum up a Hart instead of predicting when or where their Hart trophy should have come from. It’s not smart to outright claim Brodeur deserved 2 Hart trophies - instead one could point to how he had nearly double Price’s Hart votes despite never winning one thus leading to a conclusion that Brodeur provided more value during his career than Price did.
When looking at pre-1982, when the modern weighted-voting method was introduced, Gordie Howe’s dominance is unmatched. According to this metric, he dominated pre-1982 hockey more than Gretzky dominated post-1982 hockey. His 4.454 Hart Shares are 1.93x larger than 2nd place Maurice Richard - Gretzky’s total only has a 1.68x lead over 2nd place Sidney Crosby.
HARTS ABOVE EXPECTED
Since we have data on how many Harts have been won as well as Hart shares, we can functionally make a “Harts above Expected” model through basic subtraction → Harts won - Hart Shares = Harts above Expected.
This also means the lower the number, the more the player has been undervalued; the higher the number, the more the player got their trophy in their best seasons.
Again, this isn’t a metric to determine how many Harts a player was “robbed” of, but more of a stat to determine how undercelebrated they were within the award. Some notable players on this list are Martin Brodeur, the player with the highest “xHarts” without a Hart, Sidney Crosby and Jaromir Jagr who are both at the top of the list at ~ -3 each, and Nathan Mackinnon who looks like one of the most robbed players in the NHL’s history despite being only 27 years old.
NOTES
→ In 1995-96, they changed to the modern format. Before that, they had 1st place be worth 10 points, 2nd place be worth 3 points, and 3rd place be worth 1 point. In order to maintain consistency throughout the eras, I adjusted the 2nd and 3rd place votes to be worth the modern values of 7 and 5 points respectively.
→ In 1982, they introduced the modern system we use, before then the voting percentage was based on how many votes a player received divided by the total votes given out. Because of the difference in voting styles, vote % was far lower before 1982 - we created a separate list for pre-1982 MVP Shares as it’s hard to really even compare them. The biggest thing to note is just how dominant Gordie Howe was - in this metric, his 1.93x more votes than the 2nd highest on the list is better than Gretzky’s all-time dominance.
→ In 1931-32, there were 2 nominees and no online mark of their vote%, vote total, or placements. We gave both nominees a .40 MVP share so if you feel like the 1932 Hart winner, Howie Morenz, deserves an extra .20-.60 MVP shares, feel free to add it to his tally. Instead of skipping them or adding too much extra, we decided to just give them a 4/10 of a Hart instead.
→ In June, the NHL will host the 2023 awards. That will likely see McDavid jump ahead of Jagr on the post-1982 list at just 26 years old. One more phenomenal season could put him 2nd behind Gretzky.
→ This is the 100th year of Hart trophies!!!
→ The top-10 all-time, regardless of voting patterns are
Wayne Gretzky — 8.3824
Sidney Crosby — 4.9965
Mario Lemieux — 4.8626
Gordie Howe — 4.454
Alex Ovechkin — 4.1342
Jaromír Jágr — 4.0342
Connor McDavid — 3.1485
Dominik Hašek — 2.8876
Maurice Richard — 2.3077
Mark Messier — 2.2324