In May of 2013, I wrote an article on Dion Phaneuf and his play when the team is leading vs when they are trailing. It presented a pretty compelling case that Phaneuf is a significantly better player when the Leafs are trailing, a situation in which he is given a more offensive role.
Today I will update that article while looking at some new statistics.
I recently posted an article on how to conduct a WOWY analysis and in it I introduced the concept of TM (short for Team Mate) statistics. To summarize, TM statistics are like a combined WOWY and are a TOIWith weighed average of all of a player’s teammates’ Without You statistics. So, for example, TMCF20 is a TOI With Phaneuf weighted average of all of Dion Phaneuf’s team mates CF20 when they are not playing with Phaneuf. I always find this a little wordy to explain, but essentially it is how Phaneuf’s team mates perform when they aren’t playing with him. In this post, I am going to look at Phaneuf’s RelTM statistics, which is nothing more than Phaneuf’s statistics minus his TM statistics. So, for example, Phaneuf’s CF20RelTM is nothing more than CF20 – TMCF20. In the case of CF20RelTM, higher numbers are better since it would imply Phaneuf’s teammates are more productive with him than apart from him. Lower numbers are better for defensive statistics such as CA20.
I am going to look at three offensive statistics. CF20 to represent shot generation, Shooting % to represent rate of converting shots into goals, and GF20 to represent the overall goal production. Let’s start with CF20.
Since CF20 is an offensive statistic, it is better to have a higher number and in each of the last 5 seasons (Phaneuf joined the Leafs with 26 games left in the 2009-10 season) Phaneuf has had a better impact on his team mates CF20 when trailing vs. when leading. Even last season, when Phaneuf’s teammates had a worse CF20 with him than apart from him when trailing, it was still not as negative of an impact as when protecting a lead.
How about shooting percentage?
Except for 2009-10, when he played two-thirds of the season with the Flames, Phaneuf has had a more positive impact on shooting percentage when trailing vs. when leading.
How do the above two charts translate into GF20?
Mostly as expected, Phaneuf mostly has a more positive impact on scoring goals when trailing vs. when leading. In short, Phaneuf has a positive impact on his teammates’ goal production when they are playing with him.
Let’s now look at the equivalent defensive statistics, CA20, Save Percentage, and GA20.
For CA20, the lower number is better as it is obviously better to reduce the number of shots against. Aside from 2011-12, Phaneuf has had a better impact on his teammates when trailing when leading. Anyone noticing a trend here?
How about save percentage?
For save percentage, it is better to have a higher number as it indicates a boost in his goalies save percentage. Yet again, in every year Phaneuf has had a better impact on his teammates’ statistics when playing catch up vs. when protecting a lead.
As an aside, I always hear about how save percentage isn’t persistent and how is is affected significantly by randomness and luck. Well, out of all these charts, the save percentage one seems to be the most consistent with no real significant fluctuations.
And of course, how does this affect GA20?
Here it is better to have a lower number; it means a lower goals against rate and, not surprisingly, Phaneuf has consistently had a more positive impact on his teammates GA20.
Now that we have looked at offense and defense separately, let’s have a look at Phaneuf’s overall impact by looking at GF%RelTM.
As expected, Phaneuf has a significantly more positive impact on team performance when trailing versus when leading. When he is protecting a lead, he consistently has a negative impact on his teammates statistics; when leading, he consistently has a significant positive impact.
It is difficult to say why this is the case, except that when the Leafs are trailing Phaneuf is asked to play a more offensive role, and arguably he is more suited to that role. Most interesting is that Phaneuf’s defensive statistics are most significantly impacted (in a positive way) when he shifts to a more offensive role (although his individual offensive statistics see a big boost, too). We don’t see huge differences in his quality of competition or his quality of teammates, so those are unlikely significant factors, although they may play a minor role. He does start in the offensive zone more often when the Leafs are playing catch up vs when protecting a lead and this will have some impact, but it would surprise me if that alone is the significant reason. Team-level score effects are are not a factor as the stats presented above compare Phaneuf to his teammates in the same situation (i.e. leading vs trailing). Phaneuf is experiencing more significant or different score effects than the rest of the team.
That said, I recently wrote an article on rush shots and the defensive zone play of Leafs defensemen in which I observed that Phaneuf is really good at limiting the opposing teams clean entries into the Leafs zone. I posited that, knowing his effectiveness in that area, it might be wise to have Phaneuf start more in the offensive and neutral zones. The evidence shown above just seems to reinforce that. Phaneuf is best utilized playing in more offensive roles and starting in the offensive zone, where he can utilize his best skills: A good shot and an ability to impede opposition rushes against.