Randy Carlyle’s Impact on Hockey Analytics Unmatched

Whatever way you slice it, Randy Carlyle’s
tenure in Toronto wasn’t spectacular by any means. Though Carlyle was behind
the bench for parts of four seasons, he spent just 188 regular season games as
the Leafs head coach, winning 91 of those contests. Carlyle was in Toronto for just one
full season as the Leafs’ coach, to go along with a lockout shortened
year, the end of one season and the first half of another. Three playoff wins
was ultimately all he had to show for it. Even if you don’t think his time was horrible, it really wasn’t great.

But in that short period of time, Carlyle’s impact on the advancement of hockey
analytics was massive, perhaps more than any other single individual’s
contributions. No, seriously.

Randy Carlyle was the perfect villain that the analytics movement needed to
gain legitimacy.

Without the Joker and other criminals, Batman has nobody to protect Gotham
from. Without Voldemort, Harry Potter’s just a regular boy with wizard powers.
And without Randy Carlyle, professional and amateur statisticians alike don’t
have a whipping boy to try to bring down.

It’s important to note that the term “villain” is used only to portray Carlyle as
the antagonist in the strange saga that has unfolded during his tenure as the
Maple Leafs head coach, but not an attack at his character or personal intentions. While his
methods, systems, player usage, and thoughts may have been critiqued time and
time again, he believed -however flawed they were- that his actions were best
for the hockey team he was paid hansomely to run. And ultimately, it was clear, to both
analytically-inclined as well as others- eventually, it was time for Carlyle to
go. Whatever his plan was, it wasn’t working. That much was clear.

Among a bevy of new ways of evaluating the game, possession stats was and still is the most frequently referenced and arguably most important aspect of hockey analytics. Despite the traction the field was gaining and the negativity it showed towards his team, Carlyle and others in the Leafs brass ignored the red flags many had brought up.

But imagine the following scenario for a second: What if the Leafs had kept Ron
Wilson and Brian Burke, and never hired Carlyle in the first place? Ron Wilson managed a team who, despite other issues,
had years of solid possession numbers. In 2008-09, Wilson’s team ranked 13th
in 5 on 5 shot attempts (Corsi)- the next year, that number ran up to 4th.
Wilson did have poor possession teams as well, dropping to 25th in
2010-11, but was never the absolute worst team in the league. Wilson never made
the playoffs as Leafs coach, but was never exactly behind the bench of a
historically awful team. Wilson’s teams suffered their own sets of problems, for sure, but typically had bigger issues than puck possession. 

Meanwhile, Carlyle’s teams in Toronto were, analytically horrible, particular
in his two and a half seasons under Dave Nonis as GM. In terms of 5 on 5 Corsi,
Carlyle’s team’s in Toronto ranked 30th, 30th, and 27th league-wide from the start of 2012-13
until the day he was let go. In his only full season, the Leafs’ inability to
possess the puck led to a defensive nightmare that saw the most shots against ever
allowed by a team over an 82 game season.

Imagine a second scenario: Wilson was still let go, Carlyle was still hired, but the Leafs were an average, middle-of the pack or better possession team, and Carlyle still failed to find value in possession statistics.

Drew Doughty thinks corsi is “crap“, while the following two player tweets show certain players opinions of their views on analytics: 

When’s the last time you saw Lupul criticized for his view? Doughty? Ryan? The
issue goes away after a few days or hours. They’re obviously not the only players to speak up, but really, after a bit, no one seems to care.

With Carlyle, it didn’t go away. It became a part of
him and the way he was perceived. Because Carlyle both spoke out against the value of analytics and coached a style that led to an analytically ugly game, his reputation shifted greatly to the aforementioned whipping boy of the analytics contingent. 

This season, Carlyle, to his credit, did show signs of change. A partial overhaul of the Leafs management, including the hiring of president Brendan Shanahan, assistant GM Kyle Dubas, and the introduction of an analytics department might have been enough to transform Carlyle’s team- but it was too late. Add in the fact that Carlyle’s team only went from “historically bad possession team” to “very bad possession team”, and it’s easy to see why a chnage needed to be made.  

But besides giving food for the blogosphere
to criticize, Carlyle’s tenure trickles down to a level you probably haven’t
thought of. Like it or hate it- the Leafs hold the biggest stake of
them all in terms of the 30 NHL teams in terms of influencing the game. The
market, the money, the media presence, the number of fans, the pressure to
perform- it’s widely known it’s the hegemon of the NHL system. And the head coach of that team holds a lot of power. With either a different coach or a better process or results, analytics wouldn’t be quite the same as they are today.

I started off as a reader at TLN and other sites, and was fortunate enough
to transition into a writer here at the start of the reason. Without Carlyle
criticisms, I doubt I would’ve stayed long as a reader. My entire
knowledge of analytics came solely due to my curiosity as to why people were
thinking the Leafs were bad, even when they were winning games. When last season’s collapse happened, the legitimacy hit home. I know that there are hundreds, or more likely even thousands more who were introduced to the field as a direct or indirect result of the Leafs under Carlyle.

But personally, I don’t see myself as holding a very big influence over too
many people’s opinions, and neither does the average reader of a website who criticizes
the head coach of a hockey team, which is all I was until recently. However,
the lead columnist of a national newspaper that covers the team- most certainly
has a little sway.

Without Carlyle’s Leafs, James Mirtle and Bruce Arthur’s stories aren’t as exciting or
controversial. “Leafs continue to perform average due to average underlying
numbers” isn’t as much of a story as, “Leafs continue to win despite poor
underlying numbers, expect that to change soon”. Again, this isn’t a knock on
the two journalists quality or the media’s role itself- it’s just an
observation. There isn’t a unique story there without Carlyle coaching a team so
historically poor. Going against the grain as these two writers earned a
reputation for doing, criticizing teams who were winning, wouldn’t have been as
frequent had these teams been winning.

Both started covering the Leafs in the late 2000’s, and despite still being a
key focus for the two-both have expanded his reach far beyond the Leafs’
readership demographic. Without analytics to set them apart, they’re just two regular
media members for the most part. Both are extremely talented, nonetheless, but
it’s been their usage of analytics in the mainstream that have, from my
perspective, helped to separate them from the pack. Without a knowledgable audience, their work becomes less valuable.

On the flip side, without Carlyle, does the negative energy (sometimes warranted, sometimes not) towards other members of the Toronto media continue? To some degree, sure, but it’s likely not quite the same level.

Without Carlyle, hockey broadcasts wouldn’t be using analytics right now, most
likely. Over the past few seasons, Sportsnet and TSN- not to mention CBC- have
all developed, to some degree, a form of analytics in their broadcast. The
trend has extended to American broadcasts as well. Many of the networks have
unveiled the usage of analytics on their website, hiring writers from various
blogs, while other online media outlets such as the TheScore frequently post
analytically-themed articles. If you’ve been watching the games, you really don’t
need fancy stats to see how much more common analytics have been entrenched in
the mainstream. The aforementioned outlets, with national outreach, all have their head offices in, well, Toronto.

Without Carlyle, it’s easy to rationalize that Leafs assistant GM Kyle Dubas probably
doesn’t get hired, and who knows about the rest of the analytics department
would end up. Perhaps Brendan Shanahan never gets brought aboard, who knows?

Without Carlyle, the “summer of analytics”, where several teams hired probably
doesn’t happen, either, or at least not as publicly. It’s no secret the Leafs get
a disproportionate amount of coverage across the country. Is there any surprise
the most public few months of analytical changes come after a Leafs collapse
predicted by poor analytics? And if those hires were to still take place, you
can bet they wouldn’t be as big a deal. In the past, most analytic departments
were either well-kept secrets of teams or virtually non-existent.

Of course, it’s not like the analytics movement was nothing prior to Carlyle’s
hiring. There have been many, many, talented people over the past ten years or so, developing the
stats that are slowly becoming the norm. They didn’t need Carlyle to create their work. But he provided a mouthpiece- the easy target to use as a glowing example of their validity. Eventually, we would have gotten to
this point, where analytics are becoming more and more the mainstream. But once you take a step back, it’s easy to see just how massive an
impact the targeting of Randy Carlyle has had on the field.