Kukla's Korner

The Puck Stops Here

Pittsburgh Penguins Fire GM Ray Shero

If a team had a 60% chance of winning every playoff round they played in, they would be a very good team.  That team should be quite satisfied with their achievements and not be tearing itself apart because they are disappointed with their results.

What does it mean to have a 60% chance of winning every playoff series that you enter?  It means that you have a 0.6^4 chance of winning the Stanley Cup in any given year as you have a 60% chance of winning in four different rounds.  That leaves this team with a slightly under 13% chance of winning the Stanley Cup in any given year.  A team with a 60% chance of winning in any playoff series they enter will probably not win the Stanley Cup in any given year.  In fact they should go several years between cup victories.  They should win the cup once every 7-8 years.  Each other year they should lose to a team that they were favored to defeat.

Ray Shero has been Pittsburgh GM for 8 playoff years.  During his reign the Pittsburgh penguins won the Stanley Cup once.  These results are consistent with our hypothetical team that has a 60% chance of winning any given playoff series they enter.  That result got him fired.

Obviously this model has been simplified from reality.  No team could have the exact same chance of beating any opponent in a playoff series.  In some cases the probability was above 60% and in others it was below that figure, but the fact the results fit that model and none of its inputs are too outlandish does support that it isn't too far off from the Pittsburgh Penguins current situation.

Ray Shero has been the Pittsburgh GM since 2006.  He was quite lucky to take over a team with a young Sidney Crosby and Evgeni Malkin.  Most general managers do not have that opportunity.  Nevertheless he built a good team.  A team is not two individuals.  A team is a group that can win the Stanley Cup.  A team is a group that can win their division the last two years and never in the Shero years finish below second place in the division.

Firing Ray Shero is a panic reaction.  It is a lack of understanding of probability.  Pittsburgh has a good team.  They have a solid chance of winning the Stanley Cup in any given season and yet they haven't won it since 2009.  They lose to weaker teams in the playoffs.  This is exactly what our simple mathematical model predicts.  Shero lost his job because the Penguins ownership did not understand how hard it is to win a Stanley Cup.  They underestimated the odds against it and punished the GM for that mistake.

Ray Shero is a very good general manager.  He will have no trouble finding a job this summer.  A new team will be happy to hire him.  Pittsburgh will have a hard time finding anyone better to run their team.  Odds are their next GM will not have the skills of a Ray Shero.  It is always a bad sign when the man you fired is the top candidate to be hired for a vacancy league wide.  That is exactly what Pittsburgh did.

Filed in: | The Puck Stops Here | Permalink


EDJ's avatar

That is a rather poor probabilistic model of the situation. Once you take into account factors like correlation between playoff series or the fact that the probability that you win a playoff series usually decreases as you advance to a higher round because it is usually the better teams that make it to the next round, the model that you have put forth will fail dramatically in many situations. It also doesn’t make sense to model each playoff series as a single random variable, since as you said the probabilities change with each different opponent.

Furthermore, it is not unreasonable for the ownership to expect a GM to not just maintain a “certain probability” of winning a playoff series but to improve it. This is not to say I agree with Shero’s firing, but even if Shero did put together a team that would win a playoff series with a 60% chance, if he did not increase it in several years that could be a legitimate reason to fire him.

Posted by EDJ on 05/17/14 at 10:02 PM ET

PuckStopsHere's avatar

Think about what you wrote

but even if Shero did put together a team that would win a playoff series with a 60% chance, if he did not increase it in several years that could be a legitimate reason to fire him.

If Shero did build a team that has a 60% chance of beating any other team in the playoffs he built the best team in the NHL.  He beat a team that should be favored over other team in the league.  It makes no sense if you get fired for building the best team in the NHL because in your logic you are not making your team better than the best.

Posted by PuckStopsHere on 05/17/14 at 10:17 PM ET

EDJ's avatar

First of all, you can make a team that is better than the best. It just means we’re comparing future teams to present teams.

Second, I never said that he needed to make a team better than the best. What I said is that he needs to increase the probability that his team wins a playoff series, unless you think that maintaining the team’s ability is extremely difficult. That would be a fair contention.

Third, the model is so oversimplified that you can make silly assertions using it. If a GM built a team that has a 50.0001% of beating any other team in the playoffs, that would also be the best team. That team would then be favored over any other team in the league.

Posted by EDJ on 05/17/14 at 10:46 PM ET

PuckStopsHere's avatar

So to follow your logic ... Detroit is not as good a team as they are right now.  Not only has Ken Holland not made Detroit , they have got worse.  Naturally you would suggest Detroit should fire Holland…. right?

Posted by PuckStopsHere on 05/17/14 at 10:55 PM ET

EDJ's avatar

You haven’t followed my logic. First of all, I never said I agreed with the firing.

if he did not increase it in several years that could be a legitimate reason to fire him.

There are many other possible reasons to keep or fire a GM, such as retirements to key players, injuries, drafting, stability, fan expectations, etc. Raw performance is just one measure of a GM’s ability.

What I am saying is that even if somehow one could prove that the Penguins maintained a certain ability to win, that alone does not justify keeping their GM, especially if you realistically thought that the GM could have made better decisions.

Posted by EDJ on 05/17/14 at 11:02 PM ET

awould's avatar

Percentages and probability. Simplistic. Salaries, the cap, actual on ice performance. These things matter too. The Pens are top heavy with no depth and their cap situation is a mess. Thanks to Shero.

Posted by awould on 05/18/14 at 01:59 AM ET

awould's avatar

Shero is why I like Holland so much.

Posted by awould on 05/18/14 at 02:01 AM ET

w2j2's avatar

I agree with PSH.  Firing Shero was a reactionary, stupid mistake. 
One point about firing people that is underrated, is what it says about the ownership. 

If you were a sought-after GM or coach, would you go to work for an owner who throws a hissy fit after falling out of the playoffs and fires your predecessor?
You are going to try to find a Mike Ilitch or a Jeff Vinick, owners with deep pockets and an even keel.

Posted by w2j2 on 05/18/14 at 07:47 AM ET

shazam88's avatar

Firing Shero may or may not have been a mistake but basing your thesis primarily on a “60% chance of winning a playoff series” analysis is simplistic (which admittedly you acknowledge), doesn’t illustrate internal trends, and further, ignores the human aspects that are very tangible and which need to be used in concert with proper statistical analysis in order to best evaluate performance in any field (calling Nate Silver, calling Nate Silver).

PSH, you really tend to pick arbitrary measures and entrench yourself in an effort to defend them, rather than reinforce them with further statistics or other forms of actual evidence. You’ve done this in the past with your treatises on the paucity of “elite teams” and whether we should essentially throw this season’s statistics out the window when evaluating Patrick Roy’s influence, since you question the “repeatability” of the season due to your feeling that Varlamov had an up year.

Here, one could just as easily explain the fact that the Penguins managed to win a number of series against fairly middling Eastern conference opponents in the past, and that the stars were aligned fairly magically allowing them to beat the Wings for the Cup. The reality is they don’t have a great D, Fleury hasn’t set the league on fire, leadership hasn’t improved (I’m no Crosby hater either, but his on ice demeanor hasn’t matured as much as I figured it would) and they have a ton of UFAs this off-season. 

On that last point alone - the UFAs - it gives one pause and suggests that if the complexion of the team is to change, it might well make sense to put in a new GM at this point.

Posted by shazam88 from SoCal on 05/18/14 at 12:37 PM ET

Add a Comment

Please limit embedded image or media size to 575 pixels wide.

Add your own avatar by joining Kukla's Korner, or logging in and uploading one in your member control panel.

Captchas bug you? Join KK or log in and you won't have to bother.


Notify me of follow-up comments?


Most Recent Blog Posts

About The Puck Stops Here

imageThe Puck Stops Here was founded during the 2004/05 lockout as a place to rant about hockey. The original site contains over 1000 posts, some of which were also published on FoxSports.com.

Who am I? A diehard hockey fan.

Why am I blogging? I want to.

Why are you reading it? ???

Email: y2kfhl@hotmail.com