FanPost

By the numbers: Splitting up the Offensive Point Shares

Yesterday using the Points Allocation system developed for team scoring by Ian Fyffe, we were able to come up with some conclusions regarding the offensive (or lack thereof) and defensive prowess of the Iowa Wild when it comes to divvying just how the Wild scored their 72 regular season points. Given the model provided in another article, Derek Lalonde’s offensive unit account for just 27 (37.5%) of the 72 possible points. This is on par with Iowa being the second worst offensive team in the AHL in terms of goals scored. This time, I’m going to break it down even further by dividing points among offensive contributors, once again using Fyffe’s model. Using a bell curve, I’ll give each player an offensive grade using the Offensive Contribution (OC) metric.

In order to provide a grading scale based on a bell curve, the team mean for contributors at the end of the season (24) is 0.88 points earned out of 27 possible offensively distributed points. The standard deviation for this team is 0.40 shares per point. Therefore, the top 2% of players will have a score greater than 1.68 using standard deviation. Those are the A range (just think back to when you were graded in college...except for hockey players).

The A students in the offensive class (1.68 <):

Alex Tuch: 2.30

Pat Cannone: 2.06

Sam Anas: 1.68

The B students (1.28-1.67):

Mike Reilly: 1.53

Christoph Bertschy: 1.45

Zack Mitchell: 1.38

Gustav Olofsson: 1.30

Mario Lucia: 1.30

The C students (1.27-0.48):

Colton Beck: 1.15

Zach Palmquist: 1.07

Tyler Graovac: 1.00

Maxime Fortunus: 0.84

Jeff Hoggan: 0.77

Kurtis Gabriel: 0.69

Luke Kunin: 0.55

Gerald Mayhew: 0.54

The "get em next time" students (0.48>):

Mike Weber: 0.38

Hunter Warner: 0.38

Nick Saracino: 0.31

Nolan Zajac: 0.23

Nick Seeler: 0.23

Ryan Carter: 0.15

Dmitry Sokolov: 0.08

Justin Kloos: 0.08

Looking at the grading scale/conclusions:

You might notice a number of defensive players on the wrong side of the 0.88, while the top defensive pairing at the end of the year broke into the B-students. This is not necessarily a bad thing, given that the team was so defensively focused (45 points of 72 are allocated to defense). We can hypothesize that what players like Fortunus, Weber, Seeler, and Warner will most likely be looked more favorably upon when breaking down the defensive side (something I’ll do later). Fortunus as a veteran player is used mostly to prop up the young guys, but it is good to see him floating around that mean average in terms of his point production. Weber was dogged with injury this season, so to make a glaring note of the captain would be inappropriate. Warner is that undrafted prospect who Minnesota is taking a chance on, but is also a rookie. Carson Soucy doesn’t make the list simply because he only played in two contests.

In breaking down the offense, it’s great to see so many offensive rookies at the top of the grading scale, specifically looking at the A-players Alex Tuch (2.30) and Sam Anas (1.68) who sits right on that second standard deviation. Tuch has shown that he can play at the professional level, at least in the AHL, so this should not be a shock. He’s an All-Star. He’s huge. He can handle the puck well. Sam Anas is another prospect who zero ice time at the NHL level. Look at him to make some noise in training camp next season. Otherwise, he looks to be another contributor for the Iowa Wild for (parts, hopefully) next season in Des Moines. As for Pat Cannone, a tip of the cap and a tap of the stick to him. I think he’s worth keeping around for development of wingers down in the AHL (*cough* Kunin *cough* Greenway *cough* Sokolov, sips tea).

In the B-range, it’s nice to see some defensive prospects in here compared to the ones above. Both Mike Reilly and Gustav Olofsson have had a taste of NHL action. Both also were better offensively in the show compared to their counterparts playing forward. At the end of the year, Lalonde put these two on as a solid top four defensive pairing (I saw them play as the top line intermittently). For those who love offensive-defensemen, these are the guys for you. Notably, top scorer from last season, Zack Mitchell, breaks the top 15% of offensive point sharers. Mario Lucia also stayed in the B-group, though he missed a lot of the second half due to injury, only playing in 44 games; at the All-Star break he had similar numbers to Tuch at that time.

In the middle of the pack, there are no real surprises. Tyler Graovac showed some effectiveness as a 4th-line, NHL center, and only played 26 games with the club, ending above the team average with a point share of 1.00. Another player to look out for in camp on a team where center depth is tantamount to success.

Where Iowa has some glaring deficiencies, but where Minnesota fans should be happy, is that three of the top ten offensive point sharers are defensemen, who are already in a better position for the greater point shares for defensive play (something that I will cover later). With the NHL clearly shifting towards D-men who can both patrol the blue line and produce at the top of the offensive zone, this could be something worth looking into when developing a plan for bringing up prospective NHL defensemen for the Wild.

There are some deficiencies when doing a statistic like this, however. Although it works with team based numbers, there are players with inflated grades due to their production in a limited sample. What do I mean? Think about any time the season starts and a player has a hat trick. Auston Matthews comes to mind for the Maple Leafs. By the end of his first ten games in the NHL, he had nine goals. He ended the year with 40, far from his expected weighted projection of 74 goals he was expected to score. This can be looked at on this breakdown via the young players. Newcomers Sokolov, Mayhew, Kloos, and Kunin at the end of the year were all on the back half of statistics, but Kunin stands out in the average range due to his five goals and three assists in so few games played. The others had less than 20 games, so their grades will (hopefully) change as they get more ice time.

Another portion that is left out is the players that were traded and don’t count at the end of the year, but still contributed to a team’s total points. Technically speaking, of the offensive contributors that were still on the roster by the end of the season, roughly 21 of the 27 points allocated to the offense were earned by players on the Iowa Wild at the end of the season. The other six points are comprised of both call-ups and trades, ie. Downing, Dalpe, Pulkkinen (who was the team’s leading scorer at the time of his trade), etc.

The Math (aka "the stuff that dreams are made of"):

OC is one that determines an individual player’s offensive value when it comes to the team’s overall offensive value first in terms of point output and then later compared to shares of offensively earned points. What does this mean exactly? We have a pizza party for an entire roster of an AHL team (sans goaltenders, because they aren’t goal scorers) and we need to divide up 27 personal sized pizzas between 42 offensive contributors (-ish) this season (cut into squares, of course). Based upon lack of offensive play, this turns out to be the worst pizza party ever; the kind where everything is burnt cheese and maybe six or seven pieces of pepperoni on the entire pie. Nevertheless, we still have hungry people and need to split it all up. How do we do this?

The first component is assist based, so you’d think that you’d see balanced forwards, ie. centermen, and defensive players but this goes back to the previous method of finding a team’s assist to goal ratio. Clearly assists are going to be higher in this situation. I’ll break down the math using just one player on the roster, who was first on the team in scoring, but second on the team in OC: center Pat Cannone.

Ian Fyffe explains that the math goes as such:

Players Assists + Player’s goals = Offensive Contribution

Team avg. Assists per goal

(Player's individual assists divided by the team's average assists per goal; added to player's individual goals)

Iowa Wild’s Assists per Goal:

293 team assists = 1.64 goals per assist

179 goals

Pat Cannone:

29 individual assists + 9 individual goals = 26.68 or a round 27

1.64 Assists/goal

With Cannone’s OC of 27, now it is used to determine what share of the Wild’s 27 offensively earned points out of 72 belong to Cannone’s contribution as an offensive player. For the sake of Cannone, and those of you who have already fallen asleep with the glare of a math lecture on your screen, I did this for the rest of the offensive contributors for the Wild this season. The team’s OC is rounded to be about 353 (including contributors who were no longer on the team by the end of the season). Now the math, according to Fyffe, looks like this:

Individual player’s OC x Team’s Offensive Shares of Points = Individual point shares

Team’s OC

(A player's individual offensive contribution divided by the teams total offensive contribution; multiplied by the team's share of points allocated to the offense)

Cannone:

27 x 27 = 2.07

353

The conclusion being that of Cannone’s offensive output this season in terms of points earned by the team offensively, Cannone can walk away with 2.07 slices of those 27 personal pizzas brought to the party (once again, pizza that is conveniently cut into squares). Cannone will head back to New York hungry this offseason (insert joke about the AHL being the "Always Hungry League"). When doing the math for all of the players, only Alex Tuch earned more pieces, walking away with 2.30 win shares.


Next I’ll be taking a look at the defensive side of point sharing as well as how the goalies performed this season.

The opinions posted here are not those of Hockey Wilderness.