A Scathing Indictment of Randy Carlyle

When the Leafs traded Carl Gunnarsson for Roman Polak in the summer, I was one of the only people who thought it was a good trade. My thinking was that the Leafs had Gunnarsson playing too high in the lineup, and had too many offensive defensemen and that getting Polak would nicely balance the lines and make the Leafs defense better overall.

I was wrong. 

I don’t want to be too critical of Polak, as it is only 6 games into the season, but so far he looks slow and soft. I had heard he was a big hitter, but there just aren’t that many opportunities with the speed of today’s NHL to lay big hits. What you have to do is be hard on your man and take him out by simply finishing your checks – basically the kind of hits anyone who wants to can do.

To my eye, Polak has looked weak positionaly and slow for an NHL player. He is next to useless on offense and does not seem to be very good at moving the puck. I don’t think Carl Gunnarsson was worth playing on the top line – and maybe there’s some bias inherent in my take – but I don’t remember Gunnarsson ever being as bad as Polak has been.

YouTube player


After six games and roughly six hundred articles on Advanced Stats, I have changed my mind about the necessity of acquiring Polak. Without getting too stat heavy here, the upshot of Advanced Stats – Corsi is the most well-known and is basically plus/minus with shots being used as the metric instead of goals – is that over time, if you have more shots than the other team you will have had the puck more and subsequently won more games.  This is a mathematically sound proposition and most of the people who slag advanced stats just don’t understand that its a aggregate stat meant to be used over a large sample size and thus can be extremely screwy in the short-term.

The most interesting thing about these stats – to my  mind – is that they show that players who skate fast and create offense are much more effective than defensive specialists.  The math shows that there is a distinct possibility that the NHL analysts, coaches and players have been overrating the utility of the stay-at-home defenseman.  As the game has become faster and more skilled over the last decade, it is hard to argue with this assertion. Luke Schenn probably would have been a superstar in the 90s, just like Adam Foote was. In today’s game, he’s next to useless. Same with Roman Polak.

What these statistics show is that Jake Gardiner was the Leafs player last year who most drove possession. The stats reveal that guys who conventional thinking suggests are “bad at defense,” guys like Keith Yandle, Eric Karlsson and Jake Gardiner, are actually far more valuable than defense-only players, despite their sometimes maddening defensive lapses.


While I originally thought that the Leafs were better off with a more traditionally balanced defense, new information has forced me to change my mind. I now think teams might want to try icing six offensive defenseman, or at the very least, six defenseman who can all skate fast and move the puck. Carl Gunnarsson isn’t really an offensive defensman, but he isn’t a slow one-dimensional hitter either. There is room in the NHL for Gunnarssons, but I think the Gleasons and the Polaks and the Schenns are going the way of the enforcer. I now  think the Leafs made a mistake in trading the solid but unspectacular Gunnarson for Polak. I also think, after watching him skate, that signing Robidas to a three-year deal was insane – he must be the slowest player in the NHL.

Compounding the Mistake

(Greg M. Cooper-USA TODAY Sports)
(Greg M. Cooper-USA TODAY Sports)

OK, so maybe the team erred in bringing in Polak, but with the hands-down worst coach in the NHL, they just seem to be doubling down on stupidity. Like the Polak trade, Randy Carlyle is another guy I have grown to hate (not personally, just as a coach, you understand) as I learn more about analytics.  The Leafs have hired a slew of analytics guys and I can’t for the life of me understand how they continue to allow such a regressive coach as Randy Carlyle to mange their roster.

Sitting Jake Gardiner for even one game is unforgivable. This season the team should be about growing into a contender. They do not have any chance of winning the Cup this year, so they should be laying a foundation which, you would think, would include living with any mistakes Jake Gardiner might make. Over the weekend, upon his return to the lineup against Detroit, I thought he among the Leafs best players both nights.

Insanely, with Gardiner back, who does Carlyle sit? Stuart Percy! The guy who has been the Leafs’ biggest surprise of the season so far. Ironically, he looks like a faster slightly more offensive Carl Gunnarsson. At worst, Percy is the Leafs 4th best defenseman after Phaneuf, Rielly and Gardiner and sitting him is the easy thing to do because he’s a rookie, but it’s a horrible roster management decision.

The Leafs have 7 healthy NHL defensemen. Robidas and Polak should be alternating who watches the game from the press-box. Sitting one of their young defenseman just denies them the opportunity to learn and grow, in a season where the results shouldn’t be as important as the process.

Bringing Carlyle back to coach this team was almost as dumb as having Percy or Gardiner watch the game. With the team having been demonstrably worse in just about every measurable way under RC as it was under Ron Wilson, and with the team seemingly intent on moving into the future by creating an analytics department and hiring Kyle Dubas as Assistant GM, it  made little sense to bring back Carlyle who is as old-school a coach as there is in the NHL.

Stuart Percy Marlies

And what have they got for it so far? Terrible decisions on who to sit, a loss to Montreal with 2 rookies on the ice in the final minute, Daniel Winnik on the second line, a refusal to try Kadri with Kessel, and a fourth line still playing almost no minutes. An insane experiment to move their best defenseman to the left side even though he’s played his entire career on the right. Phaneuf doesn’t look comfortable and it creates even more problems because Gardiner, Percy and Rielly all play left side too. So the solution, apparently,  to having the worst defense group in the NHL is to sit your best possession player, acquire two players who wouldn’t look out-of-place skating in quicksand and then moving your best all-round defenseman to the wrong side in order to make it so at least one of your three prized young defenseman have to play their wrong side, which compounds their inevitable growing pains. OK then.

The conspiracy minded among us seem to think a final season of RC was worth it to secure Wings coach Mike Babcock. I don’t buy that for a second because it’s just not likely he’ll actually be be available. What I suspect is just regular old human kindness: They like and respect Carlyle and want to give him a chance. Unfortunately, he is demonstrably bad at his job and needs to be fired immediately.

Does anyone know what Dan Bylsma is up to right now?

3 thoughts on “A Scathing Indictment of Randy Carlyle”

  1. I had thought the.new analytics the Leafs were usig would work. I have my doubts now. An easy way of understanding all analytics is to see the immediate impact. If the analytics being used were properly used you would see higher scores with fewer goals. The analytical data available in the NHL allow for that. Look how the Wild turned the tables on Chicago in game three and four in the Stanley Cup. Then look at what they did to the Avalanch in their opening two games.

    You may not be able to win every game but you will have high scoring games if you are on the right track.

  2. A few points. First of all you could not possibly have watched that game Saturday night and seruously thought Gardiner was one of the best players on the ice for the Leafs. He was terrible. He looked completely lost all night. He could very well have been their worst player.

    You write about the benefits about using Corsi but then question Winnick playing on the 2nd line. It is because of his corsi numbers that he is on the line.

    As far as defencemen. The #1 job of a defenceman is to prevent the other team from scoring. Points or no points, I would take 6 Zdeno Chara’s any day over 6 Eric Karlsson’s.

    Advance stats or not, the Kessel, Bozak, JVR line is one of the top scoring lines in the NHL. Unless they struggle for an extended period of time why would you break them up? Not sure if you noticed who is leading the Leafs in scoring right now.

    Finally, Randy Carlyle, regardless of what “school” he is from he still has close to a .600 winning percentage over his career, which is pretty successful for the worse coach in the NHL.

    • His point with a defenseman like karalsson is not about points but about puck moving which makes a lot of sense, besides chara is one of a kind. Also the JVR,-bozak-kessel line may be good but bozak if clearly the weak link there and not at least giving kadri some minutes with kessel and JVR is ridicules, Bozak has some strengths but he will never be the offensive force Kadri can be.

Comments are closed.