Skip to Main Content
Behind The Numbers

Larkin lurking in Calder Trophy race

College stats forecast impressive start for Red Wings rookie

by Rob Vollman / Correspondent

Larkin's 3rd-period goal

NYI@DET: Larkin scores after stealing puck

2/6/16: Dylan Larkin takes the puck away from the Islanders, then scores a top-shelf goal, the Red Wings' fifth of the game

  • 00:51 •

At the start of the 2015-16 NHL season, the discussion about the Calder Trophy centered on forwards Connor McDavid of the Edmonton Oilers and Jack Eichel of the Buffalo Sabres, the first two picks in the 2015 NHL Draft.

Five months later, a different name is prominently in the conversation: forward Dylan Larkin of the Detroit Red Wings.

His traditional stats alone make a compelling case:

-- His 38 points lead the Red Wings and are second in the rookie scoring race to Artemi Panarin of the Chicago Blackhawks, who has 52.

-- His 35 even-strength points lead the Red Wings, rank ninth in the NHL, and are second to Panarin among rookies.

-- His 18 goals lead the Red Wings and are tied with Panarin for the most among rookies.

-- His 143 shots lead the Red Wings and are second to Jack Eichel of the Buffalo Sabres (160) among rookies.

-- His plus-26 leads the League. On the Red Wings, forward Pavel Datsyuk and defenseman Danny DeKeyser are tied for second at plus-13.

Video: DET@FLA: Larkin beats Luongo on the power play

A deeper look behind the numbers can help reveal if Larkin is the real deal or if his numbers are being boosted by linemates, by advantageous playing conditions, or by puck luck.

One argument in favor of Larkin's continued success is that his high scoring rate was predictable, given his success with the University of Michigan last season. As rare as it is for an 18-year-old to produce at his level in NCAA Division I hockey, it has been shown to virtually guarantee immediate NHL success.

There's a system originally designed for baseball by Bill James in 1985, first applied to hockey by Gabriel Desjardins in 2004, that estimates how many points a player will score in the NHL coming from another league. It is based on past players from the target league and what portion of their scoring remained when they arrived in the NHL.

Larkin, the No. 15 pick in the 2014 NHL Draft, had 47 points in 35 games for Michigan last season, an average of 1.34 points per game. The translation formula for the Big Ten Conference is 0.35, so the projection for Larkin was that he would score 0.47 points per game in his NHL rookie season, a pace that would put him at almost 40 points if he played a full season.

Larkin has 38 points in 53 games, which is 0.73 points per game.

He is defying the projection in part because of the age at which he competed in college. NCAA hockey is dominated by older players, and they are the majority who set the projection number. Scoring more than a point per game at age 18 is a clear indicator of atypical talent and historically yields a translation factor of 0.51 instead of the usual 0.35 for older players. That means Larkin's actual projection is closer to 0.68 points per game, which is three points removed from his actual scoring rate.

As an example, consider his closest statistical peer, Jonathan Toews of the Chicago Blackhawks. When Toews was 18, he scored 46 points in 34 games for the University of North Dakota. In his NHL rookie season (2007-08), Toews scored 24 goals and 54 points in 64 games, which is 0.84 points per game.

Each season, at least one 18-year-old player scores more than a point per game in NCAA Division I hockey, but none typically are as fully developed as Larkin and Toews. Lack of size is what kept Kyle Rau and T.J. Tynan out of the NHL, or resulted in other players being brought along more carefully, including Jason Zucker of the Minnesota Wild and Jaden Schwartz of the St. Louis Blues. Size-based exceptions aside, anyone who scores at Larkin's pace in college will absolutely perform near the projected level in the NHL.

It is too early to compare a 19-year-old Larkin to Toews, who is the captain of the Blackhawks, has won a Conn Smythe Trophy, a Selke Trophy and the Stanley Cup three times. Toews also dominates on the defensive side, in the faceoff circle and in shootouts, so his value extends beyond the skill set shown by Larkin to this point.

Larkin has shown some defensive aptitude and, as mentioned, his plus-26 leads the League. Though plus/minus still has its uses, its shortcomings can sometimes lead to misleading interpretations.

Another way to shed light on the plus/minus of a player is SPSV percentage, which adds the shooting and save percentage of a player's team when he is on the ice, Even though a player's SPSV% is largely outside his control, it's the strongest single determining factor in assessing a player's plus/minus.

In Larkin's case, prior to playing Wednesday against the Ottawa Senators, Detroit scored on 11.3 percent of its shots when he was on the ice at even strength, and the goaltenders had a .954 save percentage, which produced an SPSV% of 1,067. That was the fifth-best SPSV% in the NHL among those who have played at least 10 games, and the highest among those who have played at least 25.

For the most part, SPSV% finds a natural resting point around 1,000, eliminating the effect of short-term surges and slumps. For example, Dale Weise of the Montreal Canadiens led the League last season with an SPSV% of 1,053; he is at 999 this season. Consequently, his plus/minus has fallen from plus-21 to even this season, despite more ice time and a much higher scoring rate.

The component of plus/minus that is within a player's greater control is how many shot attempts are taken and allowed while he is on the ice. In this regard, Detroit is no more likely to control the play when Larkin is on the ice than when he isn't. That means that Larkin can't be compared to Toews beyond his college scoring rate, and that his plus/minus will soon revert to the team average.

Despite the myth busting of his League-high plus/minus, Larkin's offensive talents are exactly as expected, should keep him on the short list of Calder Trophy candidates, and further strengthen the high confidence statistical models have for players who score at a high rate at age 18 in NCAA Division I hockey.

View More