Chris Phillips, our defensive weak link: A statistical analysis

Disclaimer

Before I say anything, I will make a confession. I started this research because I wanted to prove my instinct that Chris Phillips is one of the main causes of our dismal season. This was not a general research where I just happened to stumble on to that fact, I specifically wanted to prove it. Having said that, the data is out there and paints a clear picture with or without my bias. So here it goes. The data will actually show that Filip Kuba is performing even worse than Phillips, however the focus of this article is on Phillips.

Introduction

As of February 1, 2011, the Ottawa Senators are 17-25-8. They have scored 108 goals and allowed 157 (the standings say 160, but three of those are from shootout losses). Out of those, the Senators have scored four empty net goals, and allowed two empty net goals, which I will not include in my analysis. I am however including the two penalty shot goals that the Senators have scored. So the basis for this analysis is 108 - 4 = 104 goals for and 157-2 = 155 goals against.

As another disclaimer, Phillips was on the ice for all four empty net goals for (three were even strength, one was a power play empty netter). He was on the ice for neither of the two empty net goals against. As empty net goals are not considered in this analysis, this hurts Phillips' statistics slightly.

A more detailed analysis follows after the jump.

Even Strength Goals for and Goals Against while on the ice
Chris Phillips has been on the ice for 28 of the 104 goals for, and 72 of the goals against. That of course is a bit biased against Phillips as he plays the most shorthanded time with virtually no power play time. So considering only even strength play, Phillips has been on the ice for 25 goals for, 53 against. Chris Phillips has been on the ice for more than twice as many even strength goals against as even strength goals for. The Senators have allowed a total of 118 even strength goals against, meaning Phillips has been on the ice for 45% of even strength goals against.

How does this compare with the other defencemen?

Note: Powerplay and short-handed numbers are from the perspective of the Senators. So when Erik Karlsson has a 2 in the GA column under powerplay, it means the Senators have allowed 2 goals while they are on the power play with Karlsson on the ice, i.e. two shorthanded goals.

Filip Kuba since his return, has been on the ice for 12 even strength goals for and 34 against. That's almost three times as many goals against as goals for! David Hale has the best goals for to goals against ratio at 1.75, but Hale along with Brian Lee have played relatively few games (13 and 19 games respectively). What is interesting here however is that Erik Karlsson and Sergei Gonchar, both offensive defencemen criticised for their defensive play, have actually been on the ice for fewer even-strength goals allowed than Chris Phillips, our supposed defensive stalwart.

Of course, one might think that Phillips has simply been on the ice for more even-strength situations than Karlsson or Gonchar, so to gain a more nuanced understanding, here is the data normalised for ice time. The table below shows the goals for and against in even strength, power play and shorthanded situations. The last set also includes the overall goals for and against in all situations.

Again, it can be seen that Karlsson and Gonchar allow significantly fewer goals per 60 minutes than Phillips. In defence of Phillips, he plays against better opposition than any of the defenders. However, he also plays with the best teammates around him. Check BehindTheNet here for Quality of Competition and Quality of Teammates listings, as well as Corsi.

Here is also a similar listing for GF/60 and GA/60 to the data I provided, however note that BehindTheNet considers only 5 on 5 for EvenStrength (ignoring 4 on 4), only 4 on 5 for penalty kill (ignoring 3 on 4 and 4 on 5) and only 5 on 4 for the power play (ignoring 5 on 3 and 5 on 4). The numbers are very close though as can be seen.

Gonchar even does a slightly better job on the penalty kill than Phillips (6.48 goals allowed per 60 shorthanded minutes versus 6.70 for Phillips). Erik Karlsson, meanwhile, has an incredibly low 2.91 goals against per 60 minutes shorthanded. While Karlsson rarely plays shorthanded, the Senators have allowed only two goals in the 41:18 that Karlsson has played on the penalty kill.

One unexpected statistic, however, is that Phillips being on the ice does not account for significantly fewer even strength goals scored for than Karlsson or Gonchar. However, consider that Phillips has played with Gonchar and Karlsson for most of the season. However, it will be shown below that Phillips actually hurts the performance of Gonchar and Karlsson.

Goals for and against by pairing

While watching the games, I had a feeling that Karlsson's and Gonchar's defensive play has suffered while being paired on the ice with Phillips. So I calculated the number of goals for and against by Gonchar and Karlsson when they are paired with or without Phillips. These numbers consider even strength play only.

As can be seen, Sergei Gonchar plays 50% better when paired with someone other than Chris Phillips, while Erik Karlsson performs a whopping 77% better. Note the sample sizes for with and without Phillips are both significant. Gonchar has been on the ice for 9 goals for and 18 against with Phillips and 18 for and 24 against without Phillips. Erik Karlsson has been on the ice for 10 goals for and 21 against with Phillips and 17 goals for and 20 goals against without Phillips.

Conclusion

Chris Phillips has served the Ottawa Senators tremendously over the years. He was a great defensive presence and, while on shut-down tandems with Zdeno Chara and later Anton Volchenkov, helped neutralise some of hockey's best forwards. He's never been a contributor offensively, but his defensive play made up for that. Now, his defensive play is not only not above average, but he is no better than a second year defenceman known for making defensive mistakes. He is no better defensively than Gonchar and Karlsson without any of the offensive upside. He is no better defensively than Chris Campoli or Matt Carkner either. In fact, the only player Phillips has performed better than, at least based on these statistics, has been Filip Kuba--and that is not saying a whole lot.

Luckily for the Senators, Phillips' contract is up. The Senators should either trade him for a draft pick, or if there is no trade available without taking bad salary back in return, the Senators should let him walk.