Looking Forward for Cornell Hockey
Photo by Duncan Williams
Another year, another great regular season with tournament disappointment for Cornell. For the second straight season, Cornell has followed up a loss in the ECAC tournament with a first-round exit from the NCAA tournament. What makes this year significantly worse than last year, however, was that this year’s team was so much better than last year’s version, winning the Cleary Cup as the best regular season team in the ECAC en route to a one-seed in both the ECAC and NCAA tournaments. But with a semifinals loss to Princeton and a first-round loss to Boston, Cornell’s season has once again come to an end without any gold.
Undoubtedly, this has been one of the most frustrating ends to a season in recent memory for Cornell hockey fans. Now, normally, when something like this happens, it is common for fans to try to be positive, pointing at the success of the past year and the bright future. After all, the reason this year was so disappointing is because Cornell played so well it raised our standard of expectation. Besides, Matthew Galajda will be even better next year, and there are some talented players committed to Big Red hockey for the coming year. Next year could be even better than this past one!
Unfortunately, I cannot force myself to look at this “bright side” because I do not see anything changing for Cornell. This article will explain in great detail the fundamental issues with Cornell hockey, where those issues arise from and why I cannot reasonably expect them to be fixed. This will be a somber article for sure, but I would rather be a realist than an optimist; after all, being an optimist led to me buying into Cornell when all of the evidence warned me not to.
This is there I will start, by looking at the story of the 2017-18 Cornell men’s hockey team. Before I go any further, however, here is a list of the previous articles I have written this season, which you can return to for future reference.
(3/2/18) ECAC Tournament Preview
(3/16/18) ECAC Tournament Lake Placid Preview
(3/23/18) NCAA Tournament Preview
And, as always, a reminder of the advanced statistics I use to analyze teams’ performances:
Corsi: the percentage of total shot attempts in a game taken by one team; considered to be the best predictor of future offensive output
PDO: the sum of a team’s shooting percentage and save percentage; considered to be the best quantitative measurement of “luck”
All stats come from the website, collegehockeynews.com
If you have, infant, gone back and read those articles or have been following these articles as they were posted, first of all thank you, and second there is one key theme I have been drilling in since day one: this team was not as good as we thought it was. Yes, compared to last year, Cornell was significantly improved in pretty much all aspects of the game, but to a large extent it was riding on Galajda’s success.
If you happen to believe this is just an attempt to rationalize what has happened in hindsight, then let us look at the advanced stats: with the season now complete, Cornell’s Corsi sits at 51.5%, just 22nd in the country. That is equivalent to Michigan, the team that made the Frozen Four out of Cornell’s bracket, and higher than Notre Dame, which I have maintained is an extremely weak team. News flash: Notre Dame is not winning the tournament this year. It is significantly lower than both Ohio State and Minnesota-Duluth, the two other Frozen Four teams, and perhaps the two best teams left in this tournament.
Even outside the context of the NCAA tournament, that Corsi of 51.5% indicates Cornell was, at best, a slightly above-average team in terms of its level of play. Ranked 22nd means it should have been, at-best, a borderline tournament team. So why was it so dominant in the regular season? The answer has to do with the other statistical value mentioned above: PDO, better explained as numerical luck. Cornell’s PDO was 104.5 (shooting percentage of 10.8, save percentage of 93.7). Consider that every shot on goal either has to be a goal or a save, meaning the median PDO will always have to be 100. That means Cornell was getting absolutely insane amounts of luck during the season. This was the highest PDO mark in the nation, with second-place Northeastern (who also lost in the first round) checking in at 104.1. This indicates Cornell was not one of the best teams in the country, but instead one of the luckiest.
Now, there are obvious drawbacks with using PDO, which I will comment on now. The most obvious is that a PDO measurement ignores the quality of a team’s goaltender, since save percentage could ultimately come down to how good that individual is. With Cornell, this is not doubt magnified, given Galajda was and is one of the best goaltenders in college hockey. Because of that, it is not out of the realm of possibility that a PDO over 100 is the baseline level for the Big Red. Still, 104.5 is insanely high, even for a team like Cornell.
There is also the obvious argument that a team needs luck to be able to win a hockey tournament, given the very nature of the game. But the purpose of PDO is to show whether a team is in line for a regression from its level of play. There is a fine line between the amount of luck needed to win, and being too lucky to carry a team’s performance. If you are still skeptical, consider the following table of former champions and their advanced statistical data (dats is only available for the NCAA since the 2014-15 season):YearChampionCorsi (%)PDO2016-17Denver56.0103.22015-16North Dakota56.6103.92014-15Providence54.3101.6
Small sample size aside, there is a clear lesson to be learned: in order to be able to win the NCAA tournament a team needs to be lucky, but it is more important to be good. None of these teams had a PDO over 104, and both of the teams over 103 needed Corsi values over 56, which is the mark of an overwhelmingly dominant team.
But, the strength of these advanced stats is that they are not just supported by the numbers: they are supported by logic as well. The beauty of hockey is in its randomness, and in the fact the very nature of the game is fluky. It is because of this that PDO was designed as it is, to prioritize shooting data, since shot percentage and save percentage are some of the least reliable indicators of performance, and instead are more indicative of streakiness or, in other terms, luck. Whenever a team has as high of a PDO as Cornell did, it can expect regression regardless of how talented it is.
This is what we saw with Cornell in the postseason, and this is what I have been warning against all season. Over the course of a 40-game regular season, a team can overcome bad offense or poor goaltending, but once tournament play comes around and teams are in one-and-done situations, what matters most is the team’s method. Obviously, a bad bounce here or there could do in a good team, and any team could win on any given night, but if a team is playing the right way, it is more likely to be able to come back from adversity such as bad bounces or a poor goaltending performance.
This was, in my view, the curse of Cornell this season. The team rode hot goaltending to a drastically better record than it deserved, and when the regression hit at the worst time possible, it was unable to adapt since it was forced to go up against a much higher level of competition than it faced in the ECAC. It did not even matter that Cornell’s Corsi was actually not bad, and was in fact significantly better than last year. In the NCAA tournament, above average is not good enough, and that is all Cornell was this past year: above average.
If none of what I have said sways you, all I ask is you re-watch the third period of the Cornell-Boston University game from last Saturday in the NCAA tournament. When Cornell fell down 2-1, and it was more important then ever it tilt the ice and hold onto the puck, Cornell could not even touch it. Boston played such good possession hockey that Cornell had only one shot on goal in the last six-plus minutes of the game. That, folks, is known as the eye test, and is the go-to response for many hockey purists when presented with statistical data. The numbers don’t support a team is playing well? They eye test proves the numbers. Well, the eye test says Cornell did not know how to play possession hockey, that the Big Red did not know how to turn its focus to offensive pressure when it mattered most.
Why was a team with so much talent up front, including Hobey Baker nominees Alex Rauter and Trevor Yates, and top-scorer Anthony Angello, incapable of even mounting pressure when it needed a goal? The answer has to do with something I talked about in my very first article , excerpted below:
“Men’s hockey head coach Mike Schafer has long been considered to be one of the best defensive coaches in college hockey, yet seems to be skeptical with the growing sentiment in the hockey world that speed and skill are the key ingredients to a championship-winning team. ‘We use some statistics in preparation and evaluation,’ Schafer told us in an email. ‘I have seen some of the most complicated reporting of some NHL teams with their statistics and I don’t believe that the coaches truly know how to use them and I don’t know if the people preparing them really understand the intangibles when producing the statistics.'”
Schafer, one of the most respected coach in the country, used one of the biggest clichés in hockey when discussing statistics: the concept of “intangibles.” Ideas such as “passion,” “hard work” and “grit” which coaches believe are true measures of a good player and a good team.
Look, obviously these qualities are important, and statistics in now way tell the whole story. Cornell could have had a Corsi of 60% and nothing would have been guaranteed. But, in hindsight, these sentiments, which are still very common amongst the old school hockey community, are incredibly irritating. These are the kind of statements that justify objectively poor performance in the name of subjective observances from the coaching staff. These are the kinds of beliefs that prioritize blocking shots over taking them, prioritize positioning over possession or, to put it more simply, prioritize defense over offense.
Again, defense is incredibly important in the game of hockey. I am not disputing that, and there is truth to the common trope that “defense wins championships.” But, guess what? Offense wins games, and without winning games, a team is not going to win championships. On top of this ,t here is major harm in prioritizing blocking shots, positioning and other such important concepts. What do blocking shots, hitting and similar ideas have in common? They all involved giving up possession of the puck. This is what Cornell hockey has learned to prioritize, has learned to expect, so it is not surprising that when possession is what matters, Cornell is caught and unable to adapt.
Returning momentarily to the idea about intangibles, while hard work is obviously essential to good performance, it is just the beginning. Take studying for an exam as an example. A student can be very passionate about a subject, can put in hours of preparation to get ready. But if the student is studying the wrong way, all of the time in the world will not be enough. It is the same with hockey: a team needs to be able to work hard and play with passion, but if the method is wrong, or the team is prioritizing the wrong concepts, performance will not improve.
All of this is to say that, in my view, the main reason Cornell hockey did not win the NCAA tournament this year is because the mindset leading the team is flawed. Head Coach Schafer is no doubt one of the best in the country, and one of the smartest hockey minds in the sport, but his method is flawed. There are merits to an old school philosophy, as I have detailed, but the issue is that is is still old school, thus not applicable to today’s playing style.
It is for all of the above reasons that Cornell fell so early in the tournament, and why I do not believe next year will bring any major improvements. Cornell hockey has been so successful for so long that disappointments such as this one will be written off by fans as being “unlucky,” and no heat will come on the coaching. For its part, the coaching staff has earned trust through years of success. Unfortunately, it is because of this trust that the necessary changes are unlikely to come, the focus on offense that is key to winning a national championship will not come to fruition. Cornell will continue to be a very good regular season team, and may even progress further in the tournament, but the gold will not come with it, unless…
Unless the team decides to emphasize possession with defensive effort. Cornell’s offensive talent was better this year than it was last year, and will probably be better next year. Even an acknowledgement of the importance of shooting and puck possession will likely be enough to propel Cornell from an average to a great team. This is a talented team with a great coaching staff that remains subscribed to an antiquated coaching system. That last part is the only thing holding the Big Red back. Fix it, and the championships will come back to Ithaca.
Unfortunately, I am not convinced the change in perspective will come, thus I remain skeptical of Cornell’s chances next year. Still, there is always the chance it will right the ship, and that chance is all the fans should need to maintain some hope.
Overall, it was a disappointing season for the Big Red, but the focus must now turn to next year. Expectations are going to be higher than ever, and we can only hope the team will meet them.