Prior to the 1999 NHL draft, only a few scouts saw a Swedish forward named Henrik Zetterberg -- and many didn't even notice him. But as Detroit Red Wings scout Hakan Andersson said, "this little Zetterberg guy" always had the puck. We know the rest: He ended up being one of the biggest draft steals in recent memory.
But, for our sake, the important part isn't Zetterberg's success. It's the failure of NHL teams to value him correctly.
There are many factors that create the disparity between expected performance (relative to draft slot) and actual performance; the players creating the biggest disparity are labeled draft steals or draft busts. But, of all the contributing factors, three are easily quantifiable: height, weight and home county.
So we set out to find out how teams miss on guys like Zetterberg -- what biases NHL teams have against them, as far as physical stature and scouting exposure. We looked at the top 10 percent of draft steals, and the bottom 10 percent of draft busts, as well. And we compared those players' height, weight and home country to the rest of the field.
A preview of our results: It's not a huge surprise Zetterberg exceeded expectations.
Overlooking short guys
If aliens came to Earth and saw us playing hockey, they would probably guess that height doesn't play a huge role in the game. There are no Dwight Howards trying to block your shot; no passes being lobbed to a giant Randy Moss.
But NHL teams have consistently opted for taller players. First-rounders are, on average, the tallest (6-foot-1.8), followed by second-rounders (6-foot-1.5), third-rounders (6-foot-1.2) and the rest of the pack. This indicates that more talented players tend to be taller, since the history shows that NHL teams are generally pretty good at projecting players.
Gare Joyce wrote for ESPN The Magazine on this happening among goalies.
But undervalued prospects don't fit this rule. They are usually shorter than the field:
The top 10 percent of overachieving players -- both forwards and defensemen -- are about a half-inch shorter than the average drafted player. It's not a huge difference, but we're looking at 15 years of draft picks. So a half-inch disparity suggests height plays a role in players sliding in the draft.
Also, the biggest busts are more than an inch taller than the biggest steals. So, it seems, NHL teams are much more willing to take a gamble on a tall prospect, while letting the shorter ones slide.
Those alien visitors would probably think weight plays a pretty big role in NHL success -- and they'd be right.
There is a high correlation between weight and draft slot, independent of scouting and performance. (Correlation coefficient: an astounding minus-0.267.) This shows how much weight is valued, much more so than height.
But when teams miss badly on judging a prospect, they often undervalue weight:
It's really odd: On one hand, scouts value weight; on the other hand, they are missing on guys who are, on average, five pounds heavier than the field.
But this tells us that, while players slip in the draft for their lack of height, it's usually OK to be skinny. One possible reason: We say taller players have a "frame" to grow into, while we say shorter players have lower ceilings (no pun intended).
Meanwhile, the worst value picks are also heavier than the field. But if you look at weight relative to height, these guys are just as dense as the average pick -- just a bit taller. So this indicates that teams are often fooled by height, while missing out on shorter guys who have some muscle on them.