FN Mailbag: March 13, 2015

article_98dba575-5e09-4042-9218-5f988d760dde

It’s about that time again! The Flames continue to defy gravity and the very real chance they make the playoffs for the first time since 2010 looms large on the horizon. Hazzuh! Calgary’s improbable success continue to cause tension between fans and “advanced stats” analysts however, so I got a lot of questions related to that in this week’s mailbag. 

We also touch on Monahan comparables and some hypotheticals regarding the Flames blueline. 

Speaking as part of the advanced stats crowd, the point is often to gird against unreasonable expectations and bad conclusions. As someone who has written about the team for nearly a decade, I can attest that sudden, seemingly inexplicable spikes of success can lead to faulty conclusions or decision making when things cool off. If you don’t properly account for your strengths and weaknesses while things are good, it can lead you to asking the wrong questions when they aren’t. Of course, to many fans this feels like misanthropes raining on the parade more than anything.

That said, this clash is also more fundamental than what I’ve described. Different fans have different reasons for following the game and different ways of supporting a team. There’s definitely an element of “wanting to be right” (confirmation bias) when stats and non-stats fans argue, but sometimes it’s less about the intellectual disagreement and more about how each relates emotionally to the club. 

For example, I personally don’t follow the Flames strictly for escape. I am far more engaged by the intellectual challenge of determining causality, finding new innovations and predicting outcomes than just hoping the club wins night to night. 

One of the reasons I started writing about hockey independently, rather than hanging out a message boards, was a need to investigate and express my own fanship without having to engage in useless No True Scotsman discussions with other people who also seem more concerned with inspecting other fans’ credentials than actually talking about the game. My interest is in talking about the game in a serious, skeptical and critical manner – though I’ve come to accept that it’s not at all others fans expectation or experience. 

That’s a tough one. Unlikely? Uncanny? Gift from the gods? The Flames success this season is a crazy confluence of improbable (high shooting percentage, great OT/one-goal game record) and ability (low PIM’s, quality goaltending, Excellent top defense pairing, growth from Gaudreau and Monahan). 

Like everyone else, I’ve become relatively weary of the dug-in factions on either side of the analytics battle. I am working on an article that will seek to re-frame the debate and potentially reconcile the two sides, as much as that is possible. I can’t get into specifics yet, but I want to move beyond calling low or high possession teams “bad” or “good” and talk about fragile and robust clubs instead.  

The next stage of analytics will be effectively teasing apart individual contributions from external circumstances like teammates, coaching strategies and opposition. This goes for both possession (shot volume) and percentages (goal frequency). There are some early attempts to do this in development currently, but we’ve only taken the first few steps. Getting more (and cleaner) data from SportVu in the near future may help.

Interesting question. One I hear all the time currently is Jonathan Toews, owing to Monahan’s high personal SH% and penchant for two-way play (which has improved greatly this season after a shaky rookie season). I’m somewhat hesitant to make that comparison because it’s extremely optimistic and Toews is a superior puck distributor in my eyes. 

Jordan Staal has been the most natural comparison for me, particularly early career Staal – big centre with a  good shot and very good hockey IQ. 

I ran a Hockey Reference query to help suss out a list of comparables for Monahan. It’s a very encouraging list of names, so be sure to check it out.

At this point, probably not. Sieloff is just 21 so he has time to recover, but he’s lost copious amounts of key development time to injury during his formative years and seems to have tumbled down the organization’s prospect rankings as a result. He has frequently been a healthy scratch on the farm this year. 

Reinhart, on the other hand, is running out of runway. He’s nearing the end of his entry level contract and isn’t any closer to making the team full time than when he arrived. The influx of quality forward prospects fighting for ice time and his huge step back offensively this year (10 goals, 19 points in 50 games) has likely sealed his fate in the Flames org. 

The Wiz was acquired by the Ducks in exchange for Rene Bourque (unwanted roster player), a 2nd round pick and prospect William Karlsson (a former 2nd round pick). Wiseniewski is signed until 2017 at $5M/season, so it wasn’t a rental deal. 

I would have probably supported a move like this by Calgary. Wisniewski has been decent for years and would slide comfortably into the Flames top-4 rotation. His deal also isn’t onerous for a team with so much cap space and what the Ducks gave up hardly qualifies as essential. 

Green is known as defensively suspect offensive defender, but the truth is he’s been a positive possession player against tough competition for most of his career. He’s a very good player (when healthy), though it seems like putting up 70-point seasons from the blueline is week behind him.

That said, he’s also 29-years old and already showing some signs of breaking down. I would be willing to sign up for 3-4 years at $5M or so, but probably not longer than that.

Haha. Hahahaha. HAHAHAHAHA…

  • DoubleDIon

    Kent did you see the comment on the Pike article about xdiff?

    Are you familiar with it at all?

    For me it reconciles data better with what my eyes tell me. Ie. The Flames are worse than the Kings, but not worse than the Oilers or Coyotes.

    Corsi is useful if it doesn’t stand alone. My main issues with the stat is it tracks attempts, not shots and also that is weighs a breakaway and floater from the side boards exactly the same.

    Xdiff appears to track similarly to corsi, but accounting for shooting percentages from specific areas of the ice.

    The Flames rank 23rd in xdiff which seems more accurate to me of their actual level of play. 28th is off base in my view even accounting for the good bounces or PDO we’ve had.

    I also feel corsi is more useful for individual players when taken in context with WOWY, QoC, QoT and zone starts. 5-5 corsi is less relevant to a team because your penalty differential should come into play when measuring a team, but not as much with an individual player.

  • mattyc

    When you guys talk about Possession stats, don’t really just mean shot volumes?
    There is no actually possession tracking correct?
    It is just assumed that the possession follows the shot count?
    Is this correct?, and if so do you think the 2 really do correlate so closely?

    • mattyc

      Not assumed, it has been shown. Was a very high correlation (r2 >0.90). It’s also debatable whether possession time or shot attempts actually correlate better with winning too IIRC. Either way, both variables are strong predictors of future success.

  • mattyc

    I thought the whole point of cheering for a team is about getting away from every day concerns and having fun.

    I never understood this type of thinking. I genuinely enjoy analyzing the games, both from a strategy POV and quantifiable POV (although I think the quantifiable part can/should ideally inform strategy as well). If anything, the added insight increases my enjoyment of individual games and also the macro-scale team building.

    I also can’t understand why there are ‘analytics wars’ and such a divisive binary. I can see why lots of sportswriters don’t like it: it often contradicts or disproves some of the trite bs they spew during games; but for the average fan, all it does is add extra information, with the bonus of it being quantifiable.

    • piscera.infada

      You’ve never understood cheering for a team just because they’re your team?

      Personally, I’m with you on analysing the game, but I don’t think it’s unreasonable to just enjoy the sport for no other reason than enjoying the sport. Some people just don’t want the “negativity” forced down their throats every day. Maybe be they shouldn’t come to an analytics-centred blog–that’s a fair criticism. But to dismiss the above thinking–that someone just wants to cheer for a team–comes off as a little bit arrogant (with all due respect).

      I completely agree with the false binary though. The stats are here, they explain a good deal, I (again) personally enjoy them. Perhaps a little less arrogance from both “sides” though, might just create a more reasonable dialogue.

      • mattyc

        completely agree – must have mis-portrayed my idea.

        My point is that i don’t understand why some people think there’s only one way of enjoying the game. I like analyzing it with numbers, some don’t, no problem either way. But why should I have to enjoy the game the way you do?

  • mattyc

    I am looking forward to the article you mention in Q2. I am somewhat interested in stats (like having you folks fill me in, but not really getting into the numbers too far myself), but I am getting tired of articles (yer ol’ buddy Lambert)that really really flog how bad the flames are.

    I think the stats have a place, but I think there are things that are not quantifiable too. Insisting ad nauseam the flames are terrible when they are having a surprising season is annoying.

    I prefer the sort of caution and optimism that you espouse rather than simply saying “the flames are bad. Very bad. Worst in the league.” etc

  • piscera.infada

    Why do most analytic gurus become so personally offended when the Flames win?
    You have to admit the vitriol that gets sent towards the Flames is a little overwhelming.
    Kent is one of the few who seems to avoid this and tries to find alternate solutions as to why the Flames are having success despite poor shot volumes.

  • Bean-counting cowboy

    Some thoughts:

    Bad possession teams are not necessarily bad and just plain lucky..i.e. the Flames.

    Good possession teams are not necessarily good and just unlucky..i.e. the Oilers!

    A little knowledge is dangerous! Coming to conclusions from a set of data and just the data is placing far too much reliance on that set of data. Making conclusions from a set of emotions of how we see the team is also faulty.

    “Other things being equal” has a nice ring to it, as is often stated in analysis, but it is not a true statement applied to games humans play.

    I appreciate reviewing data after a game is over, but I don’t get wrapped up in it!

  • Derzie

    Kent says: “the advanced stats crowd, the point is often to gird against unreasonable expectations and bad conclusions.”

    This applies to NHL GMs and Fantasy League players but what does this have to do with the fans? Win = happy. Loss = sad. Try Hard = Happy. Slack off = sad. Questioning why is best left to those that get paid to make actual decisions.

  • ssamze

    I feel Kent is one of the few analytic writers who strongly favours analytics without being condescending. I completely agree that it is the terms such as “bad team” and “good team” that gets people off. Volatile is more accurate.