: Player Discussion:
View Single Post
02-03-2012, 07:31 PM
Join Date: Jul 2006
Originally Posted by
Maybe it's because I have a natural aversion to mathematics but it's this kind of argument or point of discussion that does nothing for me. Trying to rationalize with numbers what we all see with our own eyes may help some people come to grips with things on a holistic level, I get that. But something as "advanced" as team shooting % when a certain player is on the ice is still a very crude measurement. Does the team volume shoot when lesser skilled players are on the ice because they lack the creativity to generate any semblance of offense otherwise? Are less shots taken when Benn is on the ice versus any other moment for whatever reason at all? Or do they take more and happen to score at a better rate? Regardless of any of those answers, his mates scoring at a better rate per shot when he's on the ice brings things back to the old eye-ball test - the team is most dangerous when Benn is on the ice.
You can slice things up any way you want mathematically but at the end of the day people are left to make conclusions and subsequently make predictions based on those conclusions. As scientifically based as you want to make the thing it still takes a human mind to give the numbers any kind of meaning and human minds operate differently and can come to (similarly reasonable but) different conclusions based on the same data.
Jamie Benn isn't "picking up" secondary assists that inflate his point totals he's picking up an earned statistic based on his play in all areas of the ice.
I wasn’t making any comment on Benn’s ability as a playmaker or his overall talent level in that previous post. When considering a subject through the lens of stats (I won’t use “advanced” because that just drawing an arbitrary distinction between one set of statistics that have been counted for years and another set that have only come into the mainstream recently) it’s often not very useful to just look at results. The fact is that just giving credit for what happens at the end of the play only takes into consideration a very small amount of what actually contributes to success. So Benn’s ability at both ends of the ice that you referenced is well born out in the possession numbers that we can generate just off of what is provided on the score sheet – numbers like relative CORSI ratio. It’s also obvious from just watching the game. I don’t think you and I disagree on this topic as much as you might think.
When I reference Benn’s ratio of secondary assists to ice time and suggest that it may not be sustainable, I think the best way to consider that is to break the situation down into its individual components. It’s logical that when the puck leaves Benn’s stick, he can exert no further direct influence over the play until the puck returns to his stick. In between it’s up to the other players on the team to make the correct decisions and execute properly. This is probably most obvious when we’re talking about a buffer of another player in between Benn and the eventual goal scorer. When the puck does eventually get to the guy taking the shot, there’s any number of things that could happen – hopefully he puts it in the bloody net. Unfortunately much more frequently he will miss the net or the goaltender will make a save or the defenseman will block the shot. Sometimes the player’s stick will break, sometimes he won’t even take a shot and instead choose to carry the puck into the corner or pass it to someone else. The point I was trying to make is that is it really so difficult to imagine that in some years a higher ratio of these chances will go in the net than in others? I don’t think that’s a very controversial statement. And yet that very ratio will have a huge effect on a player’s counting stats at the end of the year due to factors that are completely out of their control.
If we could imagine NHL players just constantly playing games to the point that there were 1,000 games in a NHL regular season, we would see these ratios even out over time and we could make some real statements about how a player’s true talent is reflected in one stat or the other. That condition doesn’t exist though so the best we can do is compare players to each other and pick out clear outliers for examination. We can ask ourselves some questions to determine if there’s actual talent driving these outliers or if it’s just a result of random chance. So far in the studies that have been done there’s been no definitive proof for the concept that players can improve their linemate’s chances to score over and above their career performance. That doesn’t mean it doesn’t exist, but to date we can not measure it with the data that we have (which is what I was trying to get at in the 2nd part of my previous post.
Another (and much clearer) example of this is Mark Fistric’s 2009-10 season where he posted that ridiculous +/- number. Well as it turns out one of the reasons he was able to finish so highly in that metric is that the Dallas goaltenders posted a .963 save percentage while he was on the ice at even strength. Is that because 2009-10 Mark Fistric was just so ridiculously good in his own end that he could prevent goals just by his very presence on the ice? Maybe Marty Turco was imbued with the powers of Dominik Hasek and Patrick Roy during those minutes? I think there’s a much better explanation for that success. It also explains why we haven’t seen the same level of performance in this category from Mark Fistric since then.
In closing, I actually work with statistical modeling as a property assessor, so I guess it’s not surprising that I’d latch on to this stuff and talk about it with my friends and on the boards here. But one thing to remember is that while we have to ensure that the results of statistical testing can be verified by observation of the real word, and I would never suggest that these methods are anything more than one tool in a large arsenal of ways to look at the hockey players, at some point we also have to be open to changing our perspective on how things work and be willing to consider ideas that may seem puzzling, or counter-intuitive, or different than deeply held convictions that we desperately want to be true.
View Public Profile
Find More Posts by Alistar