From a statistical perspective, the NHL is at least a decade behind Major League Baseball.
Moneyball — the book, not the movie — was written in 2003 and baseball has moved quickly from old school to the new school. Research the pedigree of any MLB front office these days and you’ll see a high-priced mix of Wall Street traders and Ivy League MBA’s crunching complex algorithms to predict player performance.
I get the sense that the hockey is headed in the same direction, based on conversations with NHL executives.
That said, most believe that the inherent complexity of a free-flowing team sport combined with smaller budgets for front office brains will keep progress at a snail’s pace.
Major League Soccer (with their Match Analysis tracking software) and the NBA (with their league agreement to install SportVU cameras in every building) will probably glean insights into player movement and complex statistical analysis first. A few NHL teams have experimented with SportVU as well, but they’re finding that the OCR technology has trouble identifying player names on jerseys during on-the-fly line changes.
From a stats perspective, the teams who seem to be making progress behind the scenes (San Jose, Phoenix, Nashville, Detroit, Pittsburgh, Boston, Tampa to name a few) will never reveal details publicly. They only hint at general ideas privately. In the meantime, hockey fans are left with a hunger for superstats, yet no Moneyball’esque hockey bible to work with.
Credit is due to those who delve into the data and have been willing to publish their work for all to read and critique on the internet. The problem is that for every solid article, there seems to be five more with writers confusing or ignoring statistical concepts like sample size, causation/correlation, etc.
The noise only serves to confuse readers who want to learn, but still cringe at the sight of a calculator.
As a writer/analyst/general fan of the game, I use stats as a starting point or a gut check. This season, for example, much has been made of the Pittsburgh Penguins’ lack of forward depth.
Ryan Wilson had this to say in January:
The bottom 6 has played at a very poor level for most of the year and it has forced Bylsma to use Crosby in more of a shutdown role than he would like. It’s crazy to think but Crosby could be producing even more points if he was given some more advantageous matchups….
…The bottom 6 grouping does not score, they do not possess the puck well, and they are not showing any signs of improving.
The Pensblog took it a step further last month and showed how the Penguins’ forward lines stacked up against other Cup contenders:
…[the Penguins are] successful primarily based on the strength of their top two lines. The good news is that those top two lines are very, very good, so the Pens can win despite having poor possession numbers in their bottom lines. The bad news is that a team like Chicago, that has strength up and down the line-up, would be much more likely to possess the puck more often and therefore to win a seven game series if the two teams met.
Both writers keyed in on the Fenwick Close stat (click here for an explanation of this stat) in their articles, which I think is one of the more telling proxies for puck possession over the course of a season.
The Penguins finished 16th in the NHL this season in Fenwick Close percentage at 50.2% and the number steadily declined throughout the year.
For a little perspective, no team has won the Cup with a FenClose below 50.1% or ranked below 14th since the stat inputs became available in 2007-08:
- 2008 Champ: Detroit — FenClose 59.6% (Rank #1)
- 2009 Champ: Pittsburgh — FenClose 50.1% (Rank #13)
- 2010 Champ: Chicago — FenClose 58.1% (Rank #1)
- 2011 Champ: Boston — FenClose 50.7% (Rank #14)
- 2012 Champ: Los Angeles — FenClose 53.7% (Rank #4)
- 2013 Champ: Chicago — FenClose 56.1% (Rank #2)
Even the ’09 Pens had a FenClose of over 55% after Dan Bylsma took over midseason, which would have placed them easily into the top five that year.
Needless to say, the circa 2014 Penguins are in trouble from a puck possession perspective. The important question I’ve been asking myself all season is: why?
What has changed about the Penguins style of play, roster, or coaching strategies that is influencing their FenClose number — essentially a measure of even strength shot differential? Is this a result of having over 500 man-games lost to injury this season or is it permanent?
After months of watching the Penguins, I’ve come up with five reasons why I think the Penguins are struggling in this statistical category that’s so often connected to Stanley Cup success. I’ll reveal those reasons here on Wednesday morning — and they stretch from goal-line to goal-line.
If you’ve watched the Penguins this season, come up with your own reasons.
Leave a comment below, send me your thoughts on Twitter (@MikeColligan), or send an email to MJColligan@gmail.com.
I’ll include a few of the best in tomorrow’s breakdown.
__________________________________
Simply put, the Penguins need their Dmen healthy. Look at the first graph in this post. http://www.pensburgh.com/2014/4/14/5610794/know-thy-enemy Now, it doesn’t have dates, but I’d bet it fits well with this timeline
Start season with all 6 healthy in October, elite team.
Scuderi goes out, slight drop.
Martin, Letang, Orpik go out, long steady decline.
Martin, Letang, Orpik, Scuderi return before Olympics, gradual increase.
Martin and Letang out post-Olympics, sharp decline
Everybody health again over final month of season, steady climb back to above average
Ed, honestly when’s the last time any line with Adams or Glass – or God forbid – Adams AND Glass on it, got sustained pressure in the opposing zone?
I prefer to use the eye test first, THEN look at advanced stats to see if they confirm what I’ve been watching. What I’ve been watching for the last 2 seasons is Adams and Glass spending an inordinate amount of time 5 on 5 running around in their own zone and me praying they don’t get scored on – which they do. Advanced stats show exactly that.
Whatever worth those 2 guys have in killing penalties is rendered moot when you consider what a train wreck they are 5 on 5. Anyone who thinks Tanner Glass throwing hits 2 seconds AFTER an opposing player has gotten rid of the puck results in anything other than a statistic doesn’t understand what they’re watching. A simple exercise in evaluating Tanner Glass’s “toughness” when it comes to a tangible effect in helping you win hockey games is this: take a pen and paper, then write down the number of times he wins a board battle for the puck during a game. I guarantee you won’t need the pen or the paper. Don’t even get me started on Adams, who is nothing more than a human meat shield.
Harry Z., Megna, Gibbons, Vitale would do nothing but add energy and speed to the lineup. I know, I know, they’re not as “tough” or “big” as Glass or Adams, but there’s a difference between physicality and effective physicality.
When you add that to the Penguins breakout, which consists of defensemen firing 80 foot bomb passes to forwards standing flat footed in the neutral zone who can’t generate enough speed to retrieve chip ins, it’s hardly rocket science trying to figure out why they have poor possession numbers.
Bylsma was finally revealed to be a fraud during the Olympics, and if the Penguins fail and look stupid in doing so yet again, hopefully it’ll be the last time he coaches this team.
I wonder if puck possession numbers lie about our 3-4 lines some. A lot of times, when the the bottom 6 lines have a great controlling shift, they control,a are very physical, and cycle the puck but don’t get shots on the net. The puck possession numbers like corsi and fenwick are all based on shots. While shots are important (because you aren’t scoring without them. SHOOOTTTT!!! Sorry got carried away.) They exclude these shifts which are very important at wearing down an opposing team. All of that being said I think these lines have performed horribly. In addition you want these lines to be your best defensive lines and they get trapped in their own end way too much, especially against the opposing top lines.
No! Reveal them NOW…. Also, put more articles online.