Within the past week, the Florida Panthers and Buffalo Sabres became the latest NHL clubs to fire their head coach, joining the Philadelphia Flyers in moving in a new direction behind the bench. Mid-season coaching changes can go either way, and each team that makes a move has different expectations both before and after the event. As a result, every change is judged against different factors in hindsight.
In order to get a deeper understanding of why the moves may or may not deliver the desired outcomes, we tasked resident stats guru Neil Greenberg and former NHL assistant GM Frank Provenzano with exploring the impact of certain underlying statistical metrics on how a team performs before and after the coaching change. In addition, we had them examine these particular teams in order to assess what happens next for Florida and Buffalo, given their new bench bosses.
The role of 'puck luck'
It takes both skill and luck to succeed at the NHL level, and Florida was certainly not getting any puck bounces its way. By using the advanced metric PDO, which is simply a sum of a team's shooting percentage and save percentage during even strength, we can see how just how much "puck luck" a team was seeing, or not seeing as the case may be.
Over the course of a season, these type of bounces tend to even out, so a team that is below the league average of 1000 will tend to gravitate toward that mean over time. That doesn't mean that every team will be at 1000 by season's end, just that long stretches of good or bad fortune typically don't last an entire season.
Florida had an even-strength shooting percentage of 6.6 percent and save percentage of .895 at the time of Kevin Dineen's firing, which is a PDO of 961. This is well below the league average; however, many teams go through this type of slump over the course of a season.