Thursday, January 12, 2017

The Game: One Stat Tells Story of Wild's Success

By Nick

No, I'm not counting my chickens before they hatch.

Yes, I still remember my own blog post from earlier this week talking about how hard it is to just accept that this team is good.

No, I'm not burning the candle at both ends here.

I got to thinking about Duby vs. Price tonight.  And how that feels like the story - the best goalie in the league, versus the best goalie in the league this season.  And you know what, maybe that is the story of tonight.  But it's not the story of the Wild's success this season-to-date.  Not by a long shot.

So, I looked into the stats that backed up my premise.  Because there's another sub-narrative that comes into play if those stats actually do support the premise.  Which they do.  Let me explain.

My theory was that it's not defensive play that is the difference for the Wild relative to its historical performance.  It's offense, and more specifically finishing.  The first part is obvious to anyone who has paid attention to the Wild throughout its history.  The Wild was born as a team that had to rely on defense because it didn't have enough offense to rely on that.  The historical data back that up.  I went to and looked at the Wild's goals against/game ranking for each season since it was born.

Those rankings, in order from oldest to most recently completed season: 12th, 24th, 4th, 5th, 5th, 1st, 8th, 2nd, 21st, 16th, 13th, 15th, 7th, 6th, 9th.  This year Duby and Kuemper (and the defense) have them ranked 2nd in the league in goals against/game.  In fact, the average league ranking for GA/GM before this season was 9.87th.  The Wild has had more completed seasons where its GA/GM ranking was in the single digits (9) than in the double digits (6).  Taking nothing away from Duby, Kuemper, and the defense, the kind of defensive effort the Wild is producing this season is not special within the full context of the history of the Wild.

On the other hand, looking at the rankings for goals FOR/game, the Wild has traditionally struggled relative to the rest of the league.  Those rankings, again GF/GM, in order from oldest to most recently completed season: 30th, 25th, 24th, 25th, 25th, 19th, 17th, 22nd, 20th, 26th, 30th, 22nd, 24th, 12th, 18th.  This season to date, Staal, Coyle and the fellas have been scoring at a rate that ranks....4th-best in the league, as of this morning.  You heard me right: 4th-best. As you can see, the Wild has never had a complete season where they finished in the top 10, let alone top 5.

Put another way, the average difference between the Wild's GF/GM and 15th place in the league, from inception through the 2015-2016 season, was -0.24 goals for/game.  The Wild scored a quarter of a goal per game less than the median team over the first 15 seasons of its existence.  The average difference between the Wild's GF/GM and the 1st place team in the league, for the first 15 seasons, was -0.88 goals per game.

This season, the Wild is scoring 0.41 more goals per game than the 15th-ranked team in the league, and  -0.35 goals per game than the 1st place team in the league, as far as GF/GM.  -0.35, if you hadn't already guessed is the lowest spread to the 1st place team that the Wild has ever had.

So, given that the Wild's defensive performance this season (as measured by GA/GM) is in line with their historical data, clearly the difference this season is that their offensive performance (as measured by GF/GM) is significantly better than their historical data.

But therein lies the rub.

A) We have been conditioned to expect the other skate to drop any time.  Until this season is over, I'll be worried about regression.

2) The Wild's GF/GM performance is so much better than historical that it is practically screaming for regression before the season is over.

Well, to that end, I looked at the top 15 scorers on the roster this season, and compared their season-to-date points/game to their career pts/game.  I was pleased with the result.  The list includes, in order from most points to least (among the top 15): Staal, Coyle, Granlund, Koivu, Zucker, Niederreiter, Suter, Parise, Sprugeon, Dumba, Pominville, Brodin, Haula, Stewart, and Scandella (JEEK has one more point than Scandella, but he's no longer with the team, so...).

Those 15 players have played in a combined 551 man games this season, amassing 305 points, or 0.554 points/game.  Those same 15 players, over their careers through the end of last season, had played in a combined 6,900 man games, amassing 4,057 points - OR - 0.588 points/game.  That means that those 15 players have been producing at 0.034 fewer points/game than they had over their collective careers.  Now, that's largely driven by Pominville, who is producing at 0.33 fewer points/game than his career rate, compounded by the fact that he's played in the 3rd-most games of the players on the list.  But Parise (#4 in games played on the list) is also producing 0.147 fewer points/game  than his career number.  And, while Pominville's performance is now a couple seasons in the making, and thus appears to be trend line-able, it is harder to make the same case when it comes to Parise.

Believe it or not, Koivu is producing slightly lower than his career number (0.718 pts/gm this season, 0.729 pts/gm for his career through 2015-2016).  Stewart and Scandella are also under their career numbers.  Coyle is producing at the greatest spread to his career number (+0.374 pts/gm), so you hope he has turned a corner, but prepare for the possibilty that he regresses.  But, countering that, Staal is only producing at 0.057 higher than his career number - which is more likely to be sustainable.

Among those 15 players, the top 7 players ranked by biggest spread to their career pts/gm are: Coyle, Zucker, Granlund, Niederreiter, Brodin, Spurgeon, and Dumba.  That's very interesting, and one hopes at leat a couple of them HAVE turned a corner and will be able to sustain this improved scoring - assuming a couple others regress.

So there you have it.  You can have all your advanced stats.  They all have their uses.  But you really only need one stat (and that of the non-advanced variety, to boot) to show you what is making a difference for the Wild this season.

The Minnesota Wild is scoring.  Whodathunkit?

Please join us in the message boards to discuss this article, and everything else.