clock menu more-arrow no yes

Filed under:

The Second Assist: Statistic or Gift?

New, comments

The second assist is a given part of hockey scoring, fully ingrained into decades of statistics. The question is -- what does it really show?

Aaron Doster-USA TODAY Sports

As part of my periodic musings on matters arcane and trivial, my attention wandered to the concept of the second assist in hockey.  It has, of course, become embedded in the fabric of NHL scoring, counting just as much as a goal or a "first assist" in terms of points.  The Art Ross Trophy is just as available to a player who piles on second assists as it is to the most prolific goal scorer in the NHL.   Somehow, that just doesn't seem right.  So what is this bizarre creature?  What does it show?  Why does it rank as the functional equivalent of a goal?

Although there are lots of paraphrased versions of what (and what does not) constitute an assist,  it's always better to go directly to the source.  The NHL standard for awarding assists is found in Rule 78.3, which provides:

Crediting Assists - When a player scores a goal, an "assist" shall be credited to the player or players (maximum two) who touch the puck prior to the goal scorer provided no defender plays or has control of the puck subsequently. Each "assist" shall count one point in the player’s record. Only one point can be credited to any one player on a goal.

After reviewing the definition, a few things strike me as odd.  First, there is actually no distinction between a "first" and "second" assist.  There are simply a maximum of two players who can receive assists.  Next, there is absolutely no element of causation attached to the definition.  In other words, an assist does not involve any requirement that the player's ation have any direct or indirect contribution to the goal.  Finally, note that the player(s) to be credited with the assist need only "touch" the puck  -- not "play" the puck, not "possess" the puck -- just touch it.  Intentionally, unintentionally, or otherwise.  In contrast, however, in order to break the assist chain, a defensive player must "play or control" the puck.  Interesting.

Let's deal with the whole causation issue first, as this seems to me to be the single biggest logical fallacy in the whole construct.  As a lawyer, I deal with causation virtually every day, and there are some interesting parallels.  Of course, as noted above, there is no causation element in Rule 78.3, which is in itself absurd.  So, statistically, a gorgeous tic-tac-toe play is just as impressive as a puck that bounces off the rear-ends of two offensive players, plops down on the stick of a third guy, who knocks it in.  Seriously?

Now I'm willing to concede that a player who directly passes the puck to the player scoring the goal merits an assist. No issue there, but note that I reference "passing" the puck -- not just touching it.  I realize it's the NHL, and the small rink makes for all kinds of crazy bounces, and I also understand that the rule is intended to be absolute, so that no discretion comes into play.  But in so doing, the NHL has effectively gutted any meaning from the number.  While you can certainly argue that this flaw exists with respect to "first" assists as well, the logical canyon really drops off when you get to the second assist.

Let's consider an automobile accident occurring at a four-way stop intersection (known in central Ohio as "Intersection Roulette")  Harry stops at the stop sign first, and then proceeds into the intersection.  Joe runs the stop sign in his direction, and collides with Harry.   Pretty cut and dried, right? Joe is the bad guy, and his actions caused the accident.  Now, assume the same facts, but say that Harry was texting as he proceeded into the intersection, and could have avoided it were he paying attention.  A little murkier, eh?  Joe is still probably at fault here, but Harry might have a piece of the action. Finally, assume the same facts, with the twist that Joe's lawyer argues that if Harry had not dawdled at breakfast, he would not have been at the intersection at that time, and the accident would never have occurred.  While perhaps a factually true statement, the problem is that there is nothing wrong or culpable about Harry taking more time at breakfast.  So, while it is factually a true statement, it is also entirely irrelevant.

The second assist is akin to the latter situation.  While the truth of the second player touching the puck may be undeniable -- of what relevance is it to the ultimate goal?  If a second assist can be awarded to a goaltender who merely drops the puck to a defenseman behind his own net, a defenseman who merely moves the puck down low from the point, or a forward who has a misdirected puck carom off his hindquarters, of what use is it.?  Keep in mind that we are talking about a statistic here -- not just an idle topic of conversation.  Not only is it a statistic, but the "assist" construct as a whole represents almost two-thirds of the scoring calculus.  Consider a football example. When a touchdown pass is thrown, both the receiver and the quarterback are credited with a TD. Fair enough -- they were functionally equal partners in the result.  However, if the NHL construct were in play in the NFL, the center would also get credited with a touchdown.  He touched the ball, and if he doesn't get the ball to the quarterback, the TD doesn't happen, right?  Again, totally true . . . and totally ludicrous.

To my mind, statistics must meet the acid test of bearing a direct correlation to a quality of performance that inherently belongs to the individual or team being evaluated.  Shooting percentage, save percentage, goals against average . . . these are all objective measures that directly pertain to an individual's performance.  As with any team sport, they are impacted somewhat by the actions of others, but over time are reliable and repeatable indicia of performance.  With this in mind, I scoured the past five seasons worth of NHL statistics of first and second assists to determine whether any pattern or correlation to performance could be found.  (For those interested, NHL.com now tracks 1st & 2nd assists, and has the data back to the 2010-11 season.  Go to Statistics, Enhanced, and filter on Skater Scoring)

What did I find? Virtually nothing.  No matter how you slice or dice the data, there was precious little correlation to anything.  However, there were a few trends that were observed:

  • The second assist is a much friendlier statistic to defensemen than the first assist.  Over the five years, just about half of the top 15 in the category were blue liners.  In contrast, on one defenseman appeared in the top 15 in first assists.  Again, not surprising given the defensive role.  It does nothing, however, to resolve the issue of causation.
  • The second serves very much as a derivative statistic.  While there are fewer repetitive names in the top 15 over the five years than you see in first assists or goals, some names keep re-appearing.  Nicklas Backstrom, Derick Brassard, Henrik & Daniel Sedin, Kris Letang.  Of course, the Sedins have that Evil Sedin Twin Mojo that is indescribable, but very real.  Letang spends a good portion of his ice time with guys named Sidney Crosby and Evgeni Malkin.  Backstrom does the same with a guy named Alex Ovechkin and Brassard has shared a line with Rick Nash for the better part of the past five years.  Otherwise, there was no discernible rhyme or reason to the numbers.
Clearly, the second assist is not a gauge of direct offensive production.  Rich Nash has a total of ten second assists over the past two years. Ovechkin has 18 over that same time frame, but led the Capitals in that category in 2010-11.  Most of the elite scorers do not appear on the second assist leaderboard, and even those who dominate the first assist category are nowhere to be seen.  While you can certainly argue that this is entirely consistent with the scoring structure, as the rule provides that nobody can receive more than a single point in connection with a given goal, this only serves to provide the premise.  At best, the second assist is a derivative statistic,  dependent primarily upon the efforts of others further up the food chain.  In this regard, it is not dissimilar from the plus/minus stat -- my least favorite NHL number -- which is inevitably dominated by players from the teams with the most success, or fourth line, low minutes players.  Neither is a valid measurement of individual performance.

As presently constructed, the second assist is little more than a gift presented to random individuals who happen to be on the ice and touch the puck in proximity to a goal being scored.  Sure, the second assist can be meaningful in a given play, but there is nothing inherent in the number that guarantees that it meaningful across the board.  That's how it fails as a statistic.

If you want to preserve the second assist and give it some meaning, change the "assist" definition to require more than mere touching -- make it necessary to either play or control the puck.  That way, the definition aligns with what the defense must do to break the assist chain.  Additionally, for purposes of second assists, I would require that the play come from the neutral or offensive zone.  This enhances the likelihood that the play actually had some causal link with the ultimate goal.

Am I naive enough to think these changes are going to be made?  Hardly.  This is the NHL, and change comes at a glacial pace.  Even today, a decade after the lost season and the sweeping rule changes, the officials do everything in their power to return the game to the Neanderthal days of clutching and grabbing.  So, I understand this is not happening anytime soon, if ever.  But, as I said at the top, this is the off-season, and my mind wanders to remote areas where it is not safe to tread.  Stay tuned.