Experience and Hiring from Within: Where Good GMs and Coaches Come From

Ed. note: Here’s F.O.T.B. Ben Lutz, aka the DC Sports Dork, with an RMNB guest spot all about head coach and GM hires. You can see more of Ben’s work on www.dcsportsdork.com. If you enjoy his stuff, feel free to join his facebook page.

While hockey fans were paying attention to the Stanley Cup Finals, a handful of NHL teams were focused on hiring a new head coach and/or general manager. The Capitals found their guys by hiring Brian McClellan and Barry Trotz as GM and coach, respectively. Pittsburgh pulled off the unthinkable by hiring ex-Hurricanes GM Jim Rutherford as their new general manager.

Hires like these inevitably lead to questions about the candidates’ credentials and organizational philosophies. Is it a good idea to hire a coach that has experience or one with a fresh set of ideas? Does the same apply to hiring a general manager?

In order to find the answers, let’s first observe the list of NHL head coaches and general managers as of the final day of the previous regular season. Then, we’ll use advanced stats to see which type of head coach and general manager hires deliver the most success to an NHL team.

First, let’s start with the list of head coaches as of the last game of the 2013-2014 regular season.

Of all the head coaches to have sat behind an NHL bench last April, 17 previously worked as head coach for another NHL team, seven of them previously worked as AHL head coaches and five as NHL assistant coaches. Patrick Roy was the only coach to come from the junior ranks. With the vast amount of retread hires, it’s no surprise to see the average amount of games coached hitting more than 341 (or a little more than four NHL seasons worth of experience).

Twenty-two of the 30 head coaches were former employees from other organizations. Eighteen of the 30 head coaches were hired when they were between 45 and 55 years old. resulting in an average age of 48.2.

Now let’s look at the crop of NHL general managers.

As you can see, the age range for a current NHL general manager is a little younger than that of an NHL head coach with the average hire being 46.8 years old. Of the 30 general manager positions, 20 came from outside, with nine of them previously holding a GM position at another NHL franchise. Of the 21 previously inexperienced general managers, 17 came from working as an assistant general manager or in other jobs within the front office.

With this information, it’s obvious that NHL franchises tend to hire someone with experience for a head coach and someone who hasn’t been a GM– but does have front office experience– as general manager.

So is this the correct route to take when choosing the best candidate?

In looking for an answer, it is important to note that teams have varying degrees of goals and standards. If the Pittsburgh Penguins get knocked out of round one of the playoffs, it would be considered a disappointing season. However, being eliminated in round one of the playoffs may be considered a successful season for the Florida Panthers. As a result, simply using standings points percentage isn’t enough to come up with a conclusion for head coach’s or general manager’s success rate.

In Rob Vollman’s Hockey Abstract (pages 53-61), Vollman discusses a new stat, points above expectation (or xPts). The figure is created based on behindthenet.ca writer Gabriel Desjardin’s theory that NHL teams always regress/progress to the 50% points percentage mean by 35% every year. For example, if a team records 98 standings points– 60% points percentage– they will be expected to regress about a third of the way to the mean next season, earning just 92 standings points for a 56.5% points percentage. If a team goes above or below that 92-point mark, that difference is their points above expectation (or xPts). Here is a calculation for the 2014 Washington Capitals to help clarify things.

expected points% = last year's points% - ((last year's points% - 50%) * 35%))

expected points% = 59.4% - ((59.4% - 50%) * 35%)) = 56.1%

xPts = (Actual points% - expected points%) * 2 standings points * number of games

xPts = (54.9% - 56.1%) * 2 standings points * 82 games = minus-2 xPts

You followed the most recent edition of the Caps, so this number probably makes sense to you– just as the eye test told us that they underachieved. To put things into perspective, this year’s Buffalo Sabres had minus-30 xPts while the Colorado Avalanche had plus-40 xPts. Only one team in NHL history had a higher xPts number than Roy’s Avs: the 1992-93 Quebec Nordiques.

With that explanation, let’s take a look at the head coaches from the 2000-01 season (the first season the NHL had 30 teams) to see whether one with or without previous NHL head coaching experience had the more successful tenure.

As you can see, experienced coaches are better by quite a sizable margin. In this year’s Stanley Cup Finals, however, both Darryl Sutter and Alain Vigneault had extended breaks before becoming head coaches for another team. Sutter stopped coaching the Calgary Flames in 2006 to concentrate on his general managing duties there, but it took him another six years (at age 53) to become an NHL head coach again. For Vigneault, he was just 36 years old when he became the head coach of the Montreal Canadiens. When he was fired 20 games into the 2000-01 season, Vigneault had to wait six years to become head coach in Vancouver. In the end, these guys persevered and became successful coaches at other venues, but are these types of experienced head coaches good hires in general, or are they exceptions to the rule?

The best type of retread hires are those that were fired in the previous year.

This table shows that there is a valid reason why people like Barry Trotz and Peter Laviolette get snatched up by other teams immediately. The same will apply to Dan Bylsma. As for people like Ron Wilson and Tom Renney, who were rumored to be in the hunt for the Panthers head coaching gig, this might be a sign that an NHL franchise is desperate to find anyone interested.

So how is it that people like Alain Vigneault and Darryl Sutter are more successful coaches when the table above said they shouldn’t have been?

The answer could be in their ages. A head coach with six or more years off between coaching tenures had a 24.0% and 31.6% correlation to the number of games coached in their latter tenure to their age when they fired from their previous tenure and to their amount of time between coaching stops, respectively. However, that correlation becomes a whopping 81.6% when games coached in the latter tenure was compared to the age that coach was hired during in that same tenure.

In short, hiring people like Vigneault and Jacques Martin fits the pattern of a good idea, while hiring people like Craig Ramsay and Barry Melrose did not. Time will tell if the Ted Nolan Experience will pan out for Buffalo.

Now let’s look at general managers since the 2000-01. First, which types of hires are most successful?

Remember that list of current general managers above? Lou Lamoriello was formerly the Athletic Director at Providence College before he became general manager of the New Jersey Devils. (Boy, has their street cred changed since then). However, Lamoriello is the only hire from outside of NHL circles in the last fourteen years, so NHL teams may not want to go down this path or they may be eaten by the small-sample-size monster.

That said, it is evident that former assistants are the way to go– not former head coaches and general managers. The only ex-head coach to currently hold a general manager position is Bryan Murray of the Ottawa Senators, but considering he is 71 years old, one may think that general managers with his credentials will soon be a thing of the past.

One more note: maybe don’t hire an agent or a former player to be your general manager. Pittsburgh, Vancouver and the Islanders: I’m looking at you.

So now that we can conclude that assistant GMs are the right type of hire, was hiring someone from within the best move for the Capitals?

The answer may be yes.

Is the sample size too small to make a strong conclusion? Possibly, but the results are still fascinating. Even though GMs hired from within did substantially better, their tenure is shorter than those hired from the outside. Even if you exclude the interim hires of Jeff Gorton of Boston, Tom Kurvers of Tampa Bay, and Jim Clark of Columbus, the average tenure for a GM hired from within is still shorter than five years (4.72 to be exact). There may be a mentality of giving a shorter leash to the GMs the owners already from other positions in the organization.

All in all, this is some interesting stuff– and helpful when analyzing all those teams who have recently made a new hire.