A Statistical Review of the Sens Season

Fortune, which has a great deal of power in other matters but especially in war, can bring about great changes in a situation through very slight forces. – Julius Caesar

Let’s start big picture. Here’s where the Senators wins and losses came from this season, broken down by strength:


I’m going to use Win Probability +/- a lot in this article. Win probability +/- is the amount of win probability added and subtracted from goals. It’s like plus minus, except that the goals are weighted by their importance. What this stat tells us is how the games were won and lost, which I think makes it a great tool for investigating a team’s season.

In any individual game, you need 50 points of win probability to win the game, since both teams start at 50% (50% + 50% WP added = win). The Sens won about 15 and three quarters games with a man advantage, and lost 18 games with a men disadvantage, meaning that in aggregate they lost themselves 2.24 games on special teams, and 3.38 games at even strength.

The PDO here is adjusted to the quality of the shot, giving the stat a little more usefulness when we talk about the actual ‘luck’ of on-ice shooting attempts. I go into how I calculate shot quality here.



The Hemsky-Spezza-Michalek line, in their half a season together, was the most potent line the Sens had all season in terms of win production. The worst line was Zibanejad-Spezza-Michalek. Their 923 on-ice PDO might do something towards justifying their dreadful 32 GF%, and their 53.2 CF% screams a line abused by the hockey gods as much as they were by their opponents.

I can deal with the performance of the Senators bottom six this year. Sure, it would have been nice if they contributed offensively, but their somewhat poor goal and shot differentials mask the fact that they did manage to prevent opposition offense at an above average rate.

Cory Conacher played on one line that contributed to losing more then winning: the first line with Spezza and Michalek.  Whatever happened to that guy?



The 4th most common defensive line, Gryba-Cowen, is the first one we find that had a negative corsi for percentage (47.1%). A fun fact: For all the flak the defensive corps got this year, they actually contributed 0.01 wins as a group minus the Karlsson-Methot line. That duo saw 4.6 losses worth of goals while on the ice together at even strength.

Cowen got grilled by a lot of Sens fans this season got this season, but he wasn’t really what cost this team wins. That doesn’t mean he wasn’t bad, he really was, it just means that his badness wasn’t what cost the Sens a playoff spot.

So who on the blue line cost this team wins at even strength? Phillips, Chris.


Phillips cost the team more wins at even strength then any other player, and there’s little sign things are getting any better from here.


Chris Phillips has consistently seen more and more shot attempts against while on the ice, and there’s little sign it’s going to get any better. Chris Phillips had a goal scored against him  30 seconds or sooner after a defensive zone faceoff 13% of the time. The team allowed one only 3.1% of the time with him off of it. If we look objectively at what Chris Phillips is, right now, we’re looking at a guy who should be fighting for a place on an NHL roster.


sens11The Sens had serious issues on offensive zone faceoffs. Opposition had more shot attempts from the 11th to 20th second after a Senator offensive zone faceoff. The real problems, though, were in the defensive zone. The Senators scored 5.2 wins worth of goals 30 seconds or sooner after offensive zone faceoffs. Their opponents won 14.8 in that same timeframe.

Performance directly following faceoffs has been considered a place where coaches might have some amount of influence on their team. As I mentioned above, putting Chris Phllips out for a defensive zone draw cost the team a tenth of a goal, so that’s certainly a place to start.The Senators inability to relieve pressure from defensive zone draws is something I would bring up in Paul MacLean’s exit interview, if I was Bryan Murray.

Leverage, or, The Sens Sure Picked Some Great Times To Suck


The Senators got worse in the most important minutes. Something to be concerned about going forward? Likely not, but their awfulness in the key situations cost them a ton of wins.

The bulk of their awfulness in important situations comes from key defensive moments, as opposed to situations where the Sens were in need of a goal. The Senators, at all strengths, had a 48.8% goal differential when trailing (defined as situations with an under 30% win probability), and a 43.6% goal differential when leading in big situations (>70% win probability). A poor goal differential when leading can be okay if the leading team suppresses the number of total goals scored, since a goal for is not as important as a goal against. The Sens didn’t do that, seeing the 8th most goals scored while leading on a per minute basis.Not helped by a 44.2 faceoff percentage in situations where a goal against would swing the win probability of the game by more then 50%.

A collective save percentage of .913 while leading sounds less palatable when one considers that the league wide save percentage in leading situations is several points lower then league average. Speaking of the goalies, in an recent post over at Hockey Prospectus I identified Robin Lehner as the least clutch goalie in the NHL this year.

The Future of Jason Spezza

I shall always be consistent and never change my ways so long as I am in my senses; but for the sake of precedent the Senate should beware of binding itself to support the acts of any man, since he might through some mischance suffer a change. –Tiberius

I figured I should tackle this, since Senators ownership is salivating at the idea of dropping some serious payroll by trading Jason Spezza away, and already starting to supply potential justification in his poor season this year. As I covered, Jason Spezza was on the ice for a whole lot of loss this year, but his on-ice shooting luck was very unfavorable and he finished the season with a reasonable 9 GVT.

Using a super secret projection system I built, we can look at who Jason Spezza most resembles through his age 30 season, by GVT.

Rick Martin
Jari Kurri
Lanny McDonald
Paul Kariya
Jeremy Roenick
Keith Tkachuk
Joe Nieuwendyk
Brian Propp
Mark Recchi
Mike Gartner
Luc Robitaille
Steve Shutt
Craig Ramsay
Pavol Demitra

Rick Martin is the worst case scenario and also the best comparison for Jason Spezza thus far. Martin plummeted from stardom after his age 29 season due to a heavy injury history. Here is the projection for Spezza’s age 31 to 35 seasons.



According to history Spezza’s march to mediocrity wont start until 2015. But, you know, not all players in that comparable list flamed out in their 30’s. In the aggregate, though, Jason Spezza is a player that can’t reasonably be projected to be a first line center on a quality team going forward. There are lots of teams that continue to buy players on their past success as opposed to a sensible projection for how they will perform in the future, so.. sure, a) if the Sens can find a team willing to spend assets on an overpaid, slightly above average player, do it and b) if Spezza can find a contract next off season better then 3-4 million dollars annually (which he probably will), he’s probably not a player the cost-effective Sens should be targeting.

Looking Forward

While there’s life, there’s hope -Marcus Tullius Cicero

There’s a lot of great statistical analysis out there, but statistical analysis is rightly focused on the kinds of things that are repeatable traits. I wanted to write something that talked about how this season happened. That’s why I used +/- and Win Probability +/-. These stats don’t have the best predictability for future performance, but they do give us the best idea of how the games were won and lost. The Senators came into the season with high expectations. They didn’t reach them. The answer in large part is bad luck. But there were other problems too. The team had exactly one competent defenseman (okay, Weircoch wasn’t bad), their first line was a complete mess, and the goaltending was disappointing.

Hockey fans are slowly making less and less definitive conclusions based on 82 games, but the wins still count, and the transactions that invested future wins into this season are still on the books. The bad luck that kicked this pretty good team out of the playoffs is the kind that can lift next year’s mediocre version of the Sens into competency, and there will probably be at least 2-3 players still on the board that can contribute by the time the Bryan Murray picks for the first time in this upcoming draft. NHL front offices are still poorly enough run (the Leafs, unfortunately don’t apply here) that the possibility of  clawing their way out of this dire lack of assets through shrewdness can still be a reasonable hope, but the Sens are in a bad place. 

It is not easy for men to rise whose qualities are thwarted by poverty.Juvenal

With a apparent lack of financial support from ownership to buy their way out of this, the Sens ‘natural’ win curve (the hypothetical expected trend in season to season winning, absent of market factors such as trades) is at it’s peak (!), and Senator management is facing a decline phase coming off of a year where they didn’t reach 90 points. A lot of phone calls NHL GM’s answer this summer will be coming from the 613 area code.

About mattpfeffer

Matt Pfeffer is the statistical analyst of the Ottawa 67's. His work appears in the virtual pages of Hockey Prospectus and The Hockey News. His favourite Senator is Oleg Saprykin.