Can Size Get You an NHLer?

Justin Azevedo
November 05 2013 07:55AM

Via Sean Mort

You always hear about how size is an asset in the game. But do players with size actually have demonstrably better careers than their counterparts?

With the help of reader Byron Bader (who compiled the raw numbers and provided some analysis), we're going to take a look and see if that's the case or not.

The Numbers

There's simply too many players for us to include everyone ever in this sample, so we've cut it down to players drafted between 1990 and 2012. There was a total of 3289 draftees in that amount of time, and this is how they break down:

  Below 5'10" 5'10" - 6'0"  6'1" - 6'4" Above 6'4''
Total Draftees 80 1020 2034 155
NHLers 5 72 147 8
Percentage of NHLers 6.25% 7.06% 7.23% 5.16%
Average Points of Players in Top 1000 532 510 507 560

For the purpose of this exercise, "NHLers" are guys who are top-6 forwards and top-4 defensemen.

Observations

  • Teams are much more willing to spend a draft pick on a bigger guy than a smaller guy, even if the smaller guy has better scoring stats. Based on the percentage of NHLers to come from each group, that seems like it may be a mistake.
  • At first blush, it may seem as though drafting bigger players does come with an increase of points, but the small sample (only 8/1000 players) means the results can be skewed one way or another easily. Same with the smaller players. 
  • It is interesting that huge players, although they are a small percentage of total players, have the best average points. To me, this doesn't so much speak to value of size as it does the value of other skills: there is no player that I observed who was 6'4" and over who didn't have at least adequate NHL-level skating ability and also appeared on the top 1000 list. 
  • Based on that, drafting a player simply for size while ignoring his lack of skating ability and/or offensive upside will only lead to a wasted pick.
  • The sample is very small but it's interesting that there's an entire 2% separation between the huge group (smallest percentage of NHLers) and the big group (highest percentage of NHLers).
  • Based on these numbers, it seems like it may be hard for the average GM to strike a balance between finding a player who has a high likelihood to make the NHL (typically in the mid-range) and a player who can score at an above-average rate (typically on the extremes).
  • It seems to me that if you're looking at 4 players in the CHL (one in each size bracket) who have identical scoring stats (i.e., like 45P in 50GP) you want to take the guy who's between 6'1" and 6'4. Highest likelihood to make the show.
  • However, if the player is an elite scorer, obviously take the 6'4 and up player.
  • Small players are seemingly undervalued a significant amount. They have a very good chance at making the NHL compared to their peers and they score at a high pace - but only 2.4% of draftees were in that bracket.
  • Because of this, I can't think of any reason why a GM would pick a player for his size (even though he has other deficiencies) over a player who can score.

Conclusion

It's no secret that teams covet players with size. Everyone is trying to find "the next Milan Lucic" or "the next Zdeno Chara". Unfortunately, the stunningly obvious reality is that not every player with size will make any sort of an impact at the NHL level. So then why do GMs continue to reduce their odds at finding NHLers?

It's simple: the pursuit of the outlier is a much more attractive option to the average person than it is to settle for the medial. To me, that doesn't really make sense - the point of an NHL draft is to maximize the amount of NHLers you can get your hands on. The farther away you go, the harder it is to find an NHLer. Thus, draft players who are already equipped with point scoring and skating ability before you draft for size.

A9d138d0e612f28cd46f9b7057ed715d
Justin is a 22-year-old Flames fan who also happens to be pursuing a double major at the University of Calgary. He has played hockey at high levels, enjoys wearing shorts and tends to drink far too much Grasshopper. Please don't hate him.
Avatar
#1 prendrefeu
November 05 2013, 08:14AM
Trash it!
2
trashes
+1
12
props

Just writing/speaking from a purely sociology and psychology perspective, in most societies there is general favouritism given to people who are taller than those who are average or shorter stature, despite there being absolutely no correlation between height, skill level, nor likelihood of success in their performance.

This applies to most situations in life, but it's most obvious in the workforce: an employer is more likely to hire the taller candidate and give that candidate preference for upgrades to their employment despite that person's ability to perform their job.

... and to close things off, I'll drop in a quote from my Social Psychology professor back in my undergraduate studies:

"After centuries of conclusive studies and information that could help us as a species be more successful in our goals, we can only come to one conclusion: PEOPLE ARE IDIOTS."

Avatar
#2 SoCalFlamesFan
November 05 2013, 08:25AM
Trash it!
2
trashes
+1
20
props

I'm not a statistician, but by sampling only "scorers" in the top two lines and not defensive minded players, where size (or playing big) is a factor is not this whole stat misleading. True small players should be able to accelerate and deke on average better than larger players thereby having more NHL ready skills a team is made up of 4 lines. Drafts don't always work out. Fleury is not a 4th liner type of player. He, like when he "returned", was like your stats project,a top two line player or nothing. A small player then is a homerun or nothing. A larger player can be demoted to a defensive role. A larger window of usefulness.

Avatar
#3 SeanCharles
November 05 2013, 08:32AM
Trash it!
2
trashes
+1
5
props

@SoCalFlamesFan

I agree. Third lines are integral to a teams success in the NHL nowadays.

I can understand leaving 4th liners out, but leaving top-9 forwards and bottom pairing dman might have misrepresented the data.

Many failed top-6 and top-4 prospects with size end up in these positions...

Avatar
#4 mattyc
November 05 2013, 08:52AM
Trash it!
2
trashes
+1
6
props

Yeah - interesting idea, but what SoCal said - you're introducing a lot of bias by subjectively deciding who qualifies vs. who doesn't. Perhaps you could assign a threshold (>100 games played?).

It would also be important to see some confidence intervals on those numbers. Is the 6.25% from the shortest group actually (statistically) significantly different from the tallest group?

Avatar
#5 @Gingras34
November 05 2013, 09:02AM
Trash it!
1
trashes
+1
3
props

Keegan Kanzig. That is all.

Avatar
#6 Byron Bader
November 05 2013, 09:03AM
Trash it!
1
trashes
+1
1
props

I agree with your points about bottom line players. This initial analysis, however, was deliberately looking at the offensive aspect of drafted players. The taller player certainly has a little more versatility. If he doesn't work out as a stud scorer, you can move him back to the 4th line. That's probably why they're so alluring. But I would say that it's not all that hard to trade for a guy in the back two lines. I wouldn't waste a high draft pick on a tall guy that scored at half the clip as a short or average size player just because he's got that "hockey body" though.

What's interesting is I looked at the top scorers from the CHL (just the top 15). And the vast majority of them were either short or average height. Only Mario Lemieux was in that perfect tall range. Perhaps if shorter players were given as much opportunity as taller players to make the NHL, rather than being completely discounted because of their size, there would be more of them in the dance.

Avatar
#7 SmellOfVictory
November 05 2013, 09:29AM
Trash it!
2
trashes
+1
7
props

"Small players are seemingly undervalued a significant amount. They have a very good chance at making the NHL compared to their peers and they score at a high pace - but only 2.4% of draftees were in that bracket."

Selection bias. Teams are less likely to pick a smaller player, so in order to be drafted the players have to be standouts in their developmental leagues. This is not the case for average or large players. If anything, the fact that a smaller percentage of small players end up as NHLers, in spite of the lower number of draftees, is indicative of the real challenges that exist in turning a small player into an NHLer.

Avatar
#8 Parallex
November 05 2013, 09:30AM
Trash it!
2
trashes
+1
2
props

I probably would include 3rd line forwards as part of the sample population but it's perfectly acceptable to leave 4th line and bottom pairing guys off of it. Those guys are essentually replacement level players and thus are easy (and more importantly cheap) to replace.

Avatar
#9 Parallex
November 05 2013, 10:24AM
Trash it!
1
trashes
+1
1
props
SmellOfVictory wrote:

"Small players are seemingly undervalued a significant amount. They have a very good chance at making the NHL compared to their peers and they score at a high pace - but only 2.4% of draftees were in that bracket."

Selection bias. Teams are less likely to pick a smaller player, so in order to be drafted the players have to be standouts in their developmental leagues. This is not the case for average or large players. If anything, the fact that a smaller percentage of small players end up as NHLers, in spite of the lower number of draftees, is indicative of the real challenges that exist in turning a small player into an NHLer.

Well... yes and no. There is a bit of selection bias in that statment because you're right they are standouts in their developmental leagues. But it's not by extension indicative of any true challenges that exist in turning a small player into an NHLer. It would stand to reason that if they have a very good chance at making the NHL compared to their peers and they score at a high pace then if teams were to expand the "range" of smaller players taken (meaning more then just the top standouts) it would eventually reach a point where the chance at making the NHL and the scoring rate would eventually even out.

Now as too where that point is I don't know... but the data does lead itself to the conclusion that smaller players are under-valued at the draft table and bigger players over-valued.

Avatar
#10 MK
November 05 2013, 11:10AM
Trash it!
0
trashes
+1
3
props

Looking at the Flames projected line for tonights game against min. I do not see Sven in the line up. Is he injured or Hartley decided to bench him?

Avatar
#11 SmellOfVictory
November 05 2013, 11:24AM
Trash it!
1
trashes
+1
3
props
Parallex wrote:

Well... yes and no. There is a bit of selection bias in that statment because you're right they are standouts in their developmental leagues. But it's not by extension indicative of any true challenges that exist in turning a small player into an NHLer. It would stand to reason that if they have a very good chance at making the NHL compared to their peers and they score at a high pace then if teams were to expand the "range" of smaller players taken (meaning more then just the top standouts) it would eventually reach a point where the chance at making the NHL and the scoring rate would eventually even out.

Now as too where that point is I don't know... but the data does lead itself to the conclusion that smaller players are under-valued at the draft table and bigger players over-valued.

How is that not indicative? With small players they take only the absolute best players they can find (guys who destroy their respective leagues), which means there's no real "chaff" to separate out from the wheat, so to speak. Despite this, they still have a lower success rate than players of other sizes. If you started drafting twice as many small players, you'd almost certainly see that success rate drop drastically, unless you honestly think there is a wealth of small undrafted guys who would turn into top 6/4 NHLers.

Think of it this way: if NHL teams only picked the standout players at each size range, as they have with small players, all percentages would be a lot higher. The only reason the 5'10-6'4 range is anywhere near the less than 5'10 success percentage is that the longshots taken are almost exclusively guys with some size. If you were to expand the number of below 5'10 draftees (e.g. doubling it) I would bet money that you'd see a substantial drop in success percentage for that size range.

Avatar
#12 vowswithin
November 05 2013, 11:25AM
Trash it!
0
trashes
+1
1
props

@MK

He benched him again....

Avatar
#13 SVENSANITY
November 05 2013, 11:31AM
Trash it!
2
trashes
+1
0
props

This is a hard argument to make with just the numbers alone. You have a very obvious third variable effect occurring here. Your hypothesis is essentially: Is there a difference in NHL quality players based on physical height & weight of the players. The null hypothesis is that there is no difference. The alternate hypothesis is that there is a difference between NHL quality in players physical height and weight.

The third variable problem which effects the directionality could be that smaller players may or may not be insulated by larger players playing on the same line.

So for example let's use St. Louis in Tampa at the height of his career 2004: He may have had line mates like B. Richards (6'1, 200), Stillman(6'0, 200), Lecavallier (6,4, 208), Freddie Modin (6'4, 220).

All of these individuals are both gifted offensively and large in stature which would in effect create more space for St. Louis to be effective. Is it possible that St. Louis could have been as effective without large line-mates? It certainly is possible. However, this third variable effect forces you to stick with the null hypothesis in this case as there is not enough evidence of a effect in either direction to reject the null.

Avatar
#14 kittensandcookies
November 05 2013, 11:32AM
Trash it!
2
trashes
+1
1
props

God I'm so pedantic. Data is a plural. Datum is the singular form.

Hey how about those Flames?

Avatar
#15 piscera.infada
November 05 2013, 11:35AM
Trash it!
0
trashes
+1
0
props
vowswithin wrote:

He benched him again....

Actually, I heard he has a some sort of hand injury - if true, this seems precautionary.

**I can't remember where I heard it exactly, but I think it was Peter Maher on the Fan this morning. Then again, I can't find anything about it, maybe I made it up, I don't know

****Nope, totally made that up, my bad.

Avatar
#16 schevvy
November 05 2013, 11:42AM
Trash it!
1
trashes
+1
3
props

Sven is sitting again. Hartley is turning into Randy Carlyle. This is bad

Avatar
#17 Byron Bader
November 05 2013, 11:53AM
Trash it!
0
trashes
+1
0
props

As there are questions to why only 1st 2nd liners were included, I'll provide some more insight on the bottom feeder players. Below is a more detailed chart based on what the player turned out/turning out to be (I apologize if it doesn't work exactly right... first timer). For this I went through every player and ranked them so you have this: L1 - Bonafide stud; L2 - Good NHLers; might not be top tier guys buy typical 50-60 point getters with long careers; L3 - Career 3rd liner with a long career; L4 - 4th line goon or 3rd liner with a short career; L5 - Played a few games before disappearing; L6 - Never made a dent.

What stands out is that basically the taller you are the more likely you were to play in the show (as showed above). But as has been suggested and as you can see by the numbers you can take a tall player and change him into a grinder or 6th d-man if you need to whereas this is much less likely for short players. But, again, you can find a fighter or a big dman to eat up 5 minutes of ES time any day... it'll cost you a 5th rounder and you're all set.

1990-2012 Draft Short Average Tall Huge Total Count 80 1020 2034 155 3289 L1 (Studs) 0 18 65 2 85 L2(Good NHLer) 6 58 84 8 156 L3(Career 3rd Liners) 4 79 172 16 271 L4(4th Line Goons) 2 37 99 10 148 L5(Played a few NHL games) 1 12 41 2 56 NHLer Totals (omits L5) 12 192 420 36 660 % of NHLers 15% 19% 20% 23% 20%

The main point I want to make is... if we could get a mulligan, take Cammarata and not Keegan Huge next time. I almost cried when they made that pick.

Avatar
#18 Byron Bader
November 05 2013, 11:55AM
Trash it!
0
trashes
+1
1
props

Holy hell. That didn't work at all. Anybody have any insights how to get a functioning table in the comments.

Avatar
#19 Derzie
November 05 2013, 11:58AM
Trash it!
1
trashes
+1
1
props

If a player is too small or too big, they need to be exceptional to make the NHL. The outliers on either end will have higher than average points in general, the 'average' small and big guys don't make it and do not dilute the stats. Bottom line is, if your are small or a monster, be great or go home.

Avatar
#20 Parallex
November 05 2013, 12:03PM
Trash it!
0
trashes
+1
4
props

@SmellOfVictory

Yes that's right... I believe I said exactly that actually. I just don't see where from the data you're drawing the conclusion that it indicates any real challanges (over and above the generic challanges faced by everyone).

On the line-up stuff. WTF! Bouma upped? Baertschi Scratched? Jackman Playing? Backlund still on the 4th Line? Madness I say, Pure Madness... This is almost the worst thing Hartley could do. If he just demoted Sven to the fourth line at least he and Backs would have each other. Now Baertschi doesn't even play, Backs is stuck playing with the chum, Monahan has a worse linemate, and Bouma has to play over his head. Insanity.

Avatar
#21 Byron Bader
November 05 2013, 12:06PM
Trash it!
0
trashes
+1
0
props

@SVENSANITY

An excellent point about Marty in Tampa. Not to mention that he plays with Stamkos (6'1'', 200) and Purcell (6'3'', 200) a lot of the time now. I think it's pivotal that a short player plays with taller players so that they can create some space out there. A line of short studs would be interesting. I don't see it going well at all.

Avatar
#22 jonahgo
November 05 2013, 02:41PM
Trash it!
0
trashes
+1
1
props
SmellOfVictory wrote:

"Small players are seemingly undervalued a significant amount. They have a very good chance at making the NHL compared to their peers and they score at a high pace - but only 2.4% of draftees were in that bracket."

Selection bias. Teams are less likely to pick a smaller player, so in order to be drafted the players have to be standouts in their developmental leagues. This is not the case for average or large players. If anything, the fact that a smaller percentage of small players end up as NHLers, in spite of the lower number of draftees, is indicative of the real challenges that exist in turning a small player into an NHLer.

i was thinking the exact same thing. if the population under consideration is "potentially draftable players" (which it should be if we are considering decisions during the draft), then there is a very high attrition of under 5-10 players in this study's sample due to the fact that gms don't select marginal talents when they are small.

Avatar
#23 SmellOfVictory
November 05 2013, 03:24PM
Trash it!
1
trashes
+1
0
props

@Parallex

I see it as indicating challenges to smaller players due to the fact that only the absolute elite tiny players are drafted in the first place, yet they still have a lower success rate than the average and larger players.

Avatar
#24 Jeff Lebowski
November 05 2013, 11:45PM
Trash it!
0
trashes
+1
0
props

I wonder about Olympic ice? This is my team for Sochii

St. Louis - Crosby - Stamkos, Tavares - Toews - Giroux, Couture - Bergeron - Richards, Duchene - Getzlaf - Perry, Carter - Thornton

Keith - Doughty, Bouwmeester - Pietrangelo, Vlasic - Subban, Weber

Price Luongo Smith

Avatar
#26 jvuc
November 09 2013, 01:34PM
Trash it!
0
trashes
+1
0
props

The smaller players tend to available during UFA or via trades. Skilled big men rarely make it to the UFA market.

Comments are closed for this article.