clock menu more-arrow no yes mobile

Filed under:

A Few Words about the Recent Posts at ILWT & an Open Thread for Suggestions and Hockey Talk

I’m going to be out for the evening, yelling at people representing Columbus, Ohio in professional soccer. In lieu of, say, Paul Martin’s on-ice impact in 2007-08, the first season he was leading defenseman on New Jersey or a look at Ilya Kovalchuk’s on-ice impact from this past season; I’m going to indulge in some navel-gazing. Don't worry, it's just for tonight.

If that doesn’t interest you, take a look at SB Nation’s coverage of the USHL Futures Draft. Why? Given the growth of the USHL in terms of quantity and quality, more and more USHL graduates are appearing in the NHL. It is the United States’ Tier 1 Junior league, after all. Plus, if you go to Western College Hockey Blog, you can get even more about the USHL entry draft including profiles on the newest organizations in the league. For all you know, the New Jersey Devils of a few years from now could very well include some of these players. As they used to say, keep on scrolling.

Also, please feel free to use this post for an open-ish thread. Talk about tonight's Eastern Conference Final Game 3; hopefully, Montreal can stop suffering from all the bounces finally going against them and beat the Flyers. Please also use the comments to make suggestions on what you want to see on In Lou We Trust in coming weeks, in addition to responding to the following dollop of navel-gazing writing. Maybe there's something you're interested in. Maybe you'd like to see something different (content? a new writer?) Maybe there's something you asked and I stupidly forgot it or didn't go all the way through it. I can't read your mind, so please tell me what's on it with respect to this site.

Over the past few weeks, I’ve written at length about the team’s performance and focused on a few players utilizing a few advanced statistics. Stats that aren’t in the mainstream but have been developed and compiled by people far more insightful and knowledgeable about the game and its trends. Thanks to the work of people like Gabe Desjardens (Behind the Net, and of course, his blog), Vic Ferrari of Time on Ice, everyone at Puck Prospectus, and so many others, I’ve been able to take closer, deeper analysis of the team.

While I’ve made a point of it to state what it was I was looking for and how I accomplished what I found, I never really explained why. Part of it has to do with my educational background; I graduated from Rutgers University with a B.S. and a M.S. in Industrial & Systems Engineering. Much of my course work focused on probability, optimization, quality functions, and most of all, statistical analysis. So I’m already inclined to look at stats, especially after being submerged with simple and complex techniques alike for 6 years. That said, I forgot a lot of those techniques and/or no longer have the capability to do them – so I’m not going to be able to bust out a correlation analysis with ANOVA and variable analysis and the whole 9 yards. So please don’t ask.

The main reason of it has to do trying to understand why the Devils or a player succeeded or failed. Something every Devil fan on a train platform, in their car, at home, on the phone, on the Internet, etc. discusses after a game. It’s not so much an issue of what we see is automatically wrong or misguided; but a stat is powerful in describing performance to someone who hasn’t seen the team or player perform.

For example, you can tell me that Zach Parise is a hard worker, and I’d agree with that – but there isn’t any evidence to show Hypothetical Person C who hasn’t seen Parise or doesn’t like Parise or doesn’t know much about hockey that Parise is a hard worker. A lot of the work and study into advanced statistics, in my opinion, is really to figure out the answer to why and how a team/player truly performed without the confirmation bias from what we see. Many, if not all, of the stats are flawed in some way or form. There's no perfect stat for goaltending, defending, offensive contribution, and so forth. It’s therefore very important to understand what each stat is and what it is not.

However, with multiple advanced stats, we can get an more complete picture of how good a player really was in a season, or how good a team was over a set period of time. Just to continue the example, if we find that in 2009-10 Parise (just to continue the example) had a positive effect on most of his teammate’s Corsi percentage, he had a high total GVT among forwards, and his on-ice impact was largely beneficial (coming soon?) – then it would suggest that Parise is an exceptional player and he accomplishes all of this by hard work and his talent.

Yes, it’s a lot of work to follow at times, much less put together; and I fully understand that micro-stats are rarely going to come up in conversation. That’s why the blogging medium is so great. We can take the time to do the research and present the numbers with charts, graphs, links that reference methods or other evidence, and then discuss it immediately.

Just going through the numbers at Behind the Net, organizing them to see where a player ranks is illuminating to me first. Most of the time, my intent was never to go through a stat just to say, "This method shows X about a player and therefore what you know about the player is wrong, wrong, wrong." I’m definitely not trying to come across as some kind of know-it-all. Often, I find my own predispositions were wrong when I do this sort of analysis. Take the Kovalchuk GVT post from yesterday for example. By defensive GVT, there is proof that he’s not a defensive liability – I never thought he was a total waste but not all that good either. Yet, it blew my mind that not only Kovalchuk actually contributed on defense, but did so better than players who may not have that rep (and led to an easy, attention-grabbing title).

If I come across that way, I sincerely apologize; that’s definitely not my goal - especially when I'm looking to see if I'm right or wrong to begin with. Similarly, when I ask for evidence of a claim in a comment or in a FanPost or whatever, I'm not trying to be confrontational or mean. I want to know what basis backs up what people say about a player or a team. If you think Kovalchuk is selfish, then I want to know why that is - especially in light of his assists, his very good on-ice shooting percentage from season to season, etc. That's all. No beef. No animosity. No insults. I just want to know more.

I do realize that not everyone may embrace the numbers or how I look at them. I do realize that they can often be sobering, sometimes I’m forced accept the harsh reality around a team’s performance – regardless of how I feel about it. For example, you and I can grouse about how the Devils lost to the Flyers all we want, but there’s not a player or a coach or a philosophy or a shake up that’s going to turn a 2.6% shooting percentage at even strength into the expected percentage of 6%. If there was, the team that figures it out first will undergo a dynasty like we’d never seen before; and would never in a million years share the secret if they were smart. That sometimes the reality is that it was bad luck, or rather, the team just didn’t get the bounces. (Aside: That’s the other mind-blowing strand about this, some of the clichés you hear among hockey players, coaches, management, and talking heads are actually provable.)

So this is where I’m coming from with all of the statistical analysis as of late. Admittedly, I’m still learning (catching up?) to how some of the stats work, how to utilize them, and what they largely say. I like to post up the results of something I found Devils-wise with them because A) it can give some insight into performance, B) it can advise us on what we should hope the Devils do or not do, and C) I just find it quite interesting and perhaps interesting for you. I’m also still planning to do some smaller projects – e.g. reviewing every goal Martin Brodeur gave up in this past season like I did here and here, maybe next week I’ll start that? – that may not be based on numbers but could easily provide insight into performance. Again, the main goal is to make stronger conclusions on what we think about how the Devils should go about being a successful hockey team.

It’s definitely not all perfect, I’m certainly not perfect, I’m not going to throw away all emotion, and I’m willing to learn whether or not something is clear, done correctly, or the method can be improved upon. It can be time consuming, as putting all of those charts and conclusions together takes as much if not more time than putting numbers together. I think it’s worth it if only because of what knowledge could be gained.

Hopefully this explains at least to some degree how I’ve operated as of late. I think some of you already know this (and I’m very heartened by the fact that others do see where I've been coming from), making this navel-gazing sort of pointless; but it may be useful for others. As always, thank you for reading.