The League of Extraordinary Statisticians: Size

I'm not ashamed to say that I'm about 5'8" soaking wet (haha), and that size did me no favours with scouts while playing defenseman in high school.  A good buddy of mine was scouted despite being 5'5", but he was an exceptional forward (comparable, if not better than, Joe Pavelski in our conference at the time).  Whatever the circumstance, it was no secret to us that our size would likely step in the way of our future in hockey.  I'm not bitter, but it does lead me to this topic which still lingers years after Eric Lindros came along and convinced every GM in hockey that he needed one of his own.

Size certainly is the question today, when we acknowledge that the game is no longer the war of attrition that it was in the mid- to late-1990s and guys like Martin St. Louis and Brian Gionta challenge what we think is possible for smaller players in a physical league.  James Mirtle had an interesting article earlier this year on the larger size of NHL goaltenders compared to previous seasons, and this seems as good of a time as any to see what the LOES had to say about the advantages or disadvantages of larger players.

The question for this week: Does size matter in today's NHL?

Yes, size does matter, but not in the way you might think.  Check out some of Iain Fyffe's research, which demonstrated the slight bias in the NHL Entry Draft based on size - with a good scouting staff you can get some real pint-sized gems with later picks.  Fundamentally the only way to win in the salary cap era is to sign players at discount rates, because the market has undervalued them, and that appears to be the case with smaller players.  
 
Another undervalued skill is the ability to draw more penalties than you take, and though I haven't completed a study proving this, check out those who consistently lead the league in this department - they're generally smaller guys like Martin St. Louis.  NHL officials appears to look at size differentials between players when decided whether to put the whistle to their lips.  So size does matter when building a team - but you want small!

- Rob Vollman, Hockey Prospectus

All you need to know to answer yes to this question is that the average NHL aged male in Canada and the US is 5'9" or 5'10" and the average player in the NHL is 6'1" to 6'2".  That is about four inches above the average in the population.In fact, the only significant player I can find on a quick search through the NHL Guide and Record Book who is shorter than 5'9" (which is essentially average in the North American male population) is Steve Sullivan (5'8").  I could attempt to present the same kind of numbers regarding weight except they don't work out as well because an NHL player is muscular and the average male population is fat.
 
That isn't to say that a smaller player cannot make it in the NHL, but it is rare.  The last significant player who was significantly shorter than average was Theo Fleury who was listed at 5'6".  His prime in the NHL was over ten years ago.  I imagine another successful player of Fleury's size might be out there in the future, but they are rare.  However there are several current players in the NHL who qualify as successful NHLers who are bigger than average in the general population.

- Greg Ballentine, The Puck Stops Here at Kukla's Korner

The idealistic answer is "of course not."  If you have the talent, that talent will bear itself out over the development of the player, and the cream will rise to the top.  But I still think that "you can't teach size" is a common crutch/excuse for coaches and scouts, and we all know players whose official height is only true with their skates on.  I'm not sure the bias is just in the NHL, but throughout the developmental pipeline.

I know an owner of a USHL team that insists that undersized defensemen (< 6-0) and forwards (< 5-10) are less likely to make it to the NHL via the CHL than through the NCAA pipeline, and this is something that's on my short list of things to study.  One could argue, however, that if true this is more a function of the respective player pools of both countries.  In Canada, the better/larger athletes propagate to hockey, whereas in the USA they usually go to other sports (Rams QB and Heisman Trophy winner Sam Bradford played both hockey and football until high school).  I'd love to see the football players at my high school here in Texas playing hockey, but I doubt we'll be seeing any 6-5/250 Tongans performing a pregame Haka in the NHL anytime soon.

That said, the statistics bear out the obvious trend of players getting larger.  NHL defensemen today average 6-1.75 and 208.6 pounds whereas twenty years ago they were about an inch shorter and over ten pounds lighter, and two inches and twenty pounds smaller in the last season of the Original Six era.  Forwards today average 6-0.8/201.  They were an inch and a quarter and thirteen pounds lighter twenty years ago, and were over an inch and a half and about twenty pounds lighter in 1966-67.  Goalies today are huge compared to twenty years ago, and I'm not just talking equipment.   Goalies have gone from 5-10.5/177.5 to 6-1.7/197, and there are only a couple of players around the average size of twenty years ago.

- Marc Foster, Hockey Prospectus

All other things being equal...a taller goalie is advantaged over a smaller one as he covers more of the net.  A bigger defender protects more turf.  A hulky forward is stronger on the puck.  Size is a dominating factor.

 

If you draft a big guy who disappoints with his skill game, he can still play a role for you.  But draft a skill guy who fails to develop and he has no second dimension on which to base a career.  So there is a natural selection bias towards size.

 

But other things are rarely equal.  Smaller usually means more nimble.  Smaller usually means faster.  Smaller players, to succeed, have been forced to become skill players.  So there is usually a size/skill trade-off.  The NHL has been working to give the skill player more room and relative value.  But size will always matter - big time.

 

- Alan Ryder, Hockey Analytics     

Yes.

 

Size in a player is an asset, just like any tool is be it speed, hand-eye coordination, lower body strength etc.  The thing with the size debate is I hate when it becomes too extreme and people draw a line on each side. Those sides being things like a player has to be X tall with Y weight as a minimum standard and the other side being, "Well look at Martin St. Louis and all the other small players." 

 

Physical assets do help players, and David Staples has shown there are plenty of things in the physical game that contribute to a team’s success that don’t get recorded by the NHL. However the flip side of that is not all big players are good physical players, while some small players like say Brian Gionta can be an asset in the physical game.

 

There is no straight-forward answer to how valuable size is, but it definitely is valuable, and players can definitely get away with poor size if their other tools are of high quality. It’s all part of the package of many things that makes up a player and has to be evaluated as such.

 

- Corey Pronman, Hockey Prospectus

Though I couldn't track him down for this LOES, Iain Fyffe has done at least one good article that I've seen on the topic (and definitely worth reading).

I appreciate Rob's care for the little people.  In all honesty, though, I think he raises an interesting point about markets in today's salary cap...bigger players likely fetch more money and minutes from NHL teams, making the signing of more physically-gifted small players economical.  Their success would still be contingent (as per the other responses) on being talented enough to play physical or produce while avoiding the physical element.  It certainly says something when guys like P.A. Parenteau and Andrew Ebbett, capable NHL players, have struggled to get time at the highest level.

In general, a league with greater parity puts a premium on smaller advantages, and size can often be that advantage.  Bigger players have the edge in blocking shots and disrupting passes and stickhandling; if they use their size effectively, they can also gain or maintain puck possession and get scoring chances in front of the net.  Zdeno Chara can use his size to hit the puck 563 miles per hour.   And so on.

It's worth discussing how size might manifest itself in the data; what other advantages to being larger are we missing?  What about potential advantages to being smaller...maybe in terms of drawing penalties, as Rob suggests?  Any hypotheses on how a team completely stocked with current NHL players under 6'0" would fare?  Can we slap sensors on Theo Fleury and figure out how he did it?