Malcolm Gladwell, in his best-selling book Outliers writes “Those born in the last quarter of the year might as well give up on hockey…” Why? There is
…an iron law of Canadian hockey: in any elite group of hockey players – the very best of the best – 40 percent of the players will have been born between January and March, 30 percent between April and June, 20 percent between July and September, and 10 percent between October and December
A typical major junior (CHL) roster – the level Gladwell talks about in his book – will follow this pattern. But Gladwell doesn't give any statistics about the NHL.
It turns out that NHLers birthdays are fairly evenly distributed across the first three quarters of the year. January-April players fare worse than Gladwell's law predicts, July-September and October-December players fare much better. Here are the actual numbers:
Canadian and US born players: Final 8 teams, 2010 rosters
Month born:
| Jan-Mar | Apr-Jun | Jul-Sep | Oct-Dec | |
| MTL | 8 | 6 | 9 | 6 |
| PIT | 11 | 4 | 11 | 5 |
| CHI | 11 | 9 | 4 | 6 |
| VAN | 5 | 12 | 7 | 2 |
| SJS | 5 | 6 | 11 | 4 |
| DET | 7 | 4 | 3 | 7 |
| BOS | 5 | 11 | 10 | 0 |
| PHI | 14 | 9 | 4 | 3 |
| TOTAL | 66 | 61 | 59 | 33 |
Source: www.nhl.com.
Hockey is a physical game. Size matters. When players are 4 or 8 or 10, birth month and size are highly correlated. So early-year children have a better chance of making an elite team. When the CHL teams draft 15- and 16-year olds, more of the larger, early-born players make the cut.
But by the time players are ready for the NHL draft, any correlation between birth month and size has pretty much disappeared. Younger players have more potential for growth and development during their time in juniors. If my numbers are right, and the first-quarter effect has largely disappeared by the time players reach the NHL, then players born in the later in the year must have a better chance of making it from the CHL to the NHL.
Players on 60 CHL teams, 29 AHL teams, European franchises and the US college system are all vying for a few hundred NHL spots. A player has to have talent, the right environment (lots of ice time), and luck (no concussions) to earn a regular space on an NHL roster.The overwhelming majority never will.
The losers from the current system are not December babies, whose dreams of making the NHL are crushed at an early age. It is the January babies and their families, whose size advantage early on gives credibility to their NHL dreams, who sacrifice so much chasing on-ice glory.
But Gladwell's iron law of Canadian hockey holds for one group of elite players: on this year’s Leafs team, 40 percent of the North American players were born between January and March.
If you would like to play with these numbers yourself, here's the file:
Worthwhile Canadian Initiative 
Great post Frances! My sister is a teacher and she’s well aware of the fact that children who are born just after the cut-off date for Grade 1 doing much better than children born just before (December babies), but it’s nice to see effect observed elsewhere.
Oh, and my sister actually thinks the December baby struggles most in high school, rather than elementary. In her opinion, December babies are much less mature than their classmates in those later years, and struggle because of it.
Going back to your hockey example, one might expect to see the same effect in the transition between junior hockey and the NHL. A December baby draftee will be almost a year younger than a January baby draftee when moving up to professional hockey, and dealing many of them must find it much more difficult to deal with all the challenges, lacking the extra year of maturity.
I always feel inadequate when my co-bloggers do posts using real data, when I’m just waving my arms around.
Thoughts: is it obvious that all the effects are due to social conventions like when we start the school year? Nearly all animals and birds have a solar cycle. (The silly robins are trying to build a nest above my patio again, just like last year). Humans don’t. Is this because humans are Africans, and came from the tropics? Might there be incipient solar cycle effects in the Northerly or Southerly regions? Or could there be something like the effect of vitamin D through sunlight on their pregnant mothers?
I wish I could remember where I read something on this recently (Marginal Revolution?) Someone using (Swedish?) data on children’s weight on age and birthdate? Of course, when food was more seasonal, that could have had an effect too.
Kosta, interestingly, the cut-off date for NHL draft eligibility is September, so the players who are oldest at the time of the draft are those born in October.
Births aren’t evenly spaced throughout the year. I couldn’t find recent data on birth spacing, but there is a bit of the spike in the hockey birth dates exactly nine months after Christmas/New Years.
Some ambitious parents who know about the effect Kosta describes plan to have their children in the first half of the year. If this is true of potential hockey parents the first quarter affect may be overstated. My sense is that in middle-class Canadian families, more birth are in the first half of the year.
There are also studies like the one Nick described that correlate birth month and health. One theory is vitamin D during pregnancy – the optimal time to be pregnant in terms of fetal development is during the summer when there’s lots of vitamin D.
One thing I didn’t do was correlate birth month with performance in the NHL, but if anyone would like to add +/- ratios, position, on-ice minutes or anything else to the spreadsheet above and mail it to me at frances_woolley@carleton.ca I would be thrilled! (e.g. is the first quarter effect larger for defencemen, for whom size is more important?)
Frances,
So can we assume that the US and European farm systems skew toward the end of the year? If they do, then it would indeed seem that Canada’s system unduly privileges early birth months. And why does there seem to be full corrections for Q2 and Q3, while Q4 babies are still at a huge disadvantage?
Nothing that a birth month hockey tax credit for Conservative leaning suburban voters can’t solve.
Makes you wonder if a lawsuit challenging the Jan. cut-off date isn’t far off. Or separate leagues/divisions.
J. Powers – I excluded all of the European players as their developmental system is so different. The numbers above are Can/US born players only, which all use a Dec 31st cut-off up until the NHL draft (Sept cut-off).
As Nick pointed out, we can never be sure whether or not there’s full corrections for Q2 and Q3, because we’ll never know for sure when the pool of potential hockey players were born (Canadian births by month won’t do it, because the pool of players comes from a particular demographic – it’s a rare single parent who can cope with the demands of having a kid in high-level hockey, and the faces in the NHL are still overwhelmingly white. Places where there is abundant natural ice and not much else to do, e.g. small town Saskatchewan, are overrepresented. Yes, I know, Jerome Iginla and Ray Emery, but Iginla’s grandparents helped out a lot, his mom dedicated her life to his career and Ray Emery’s step-dad was very involved in his hockey development).
I have no idea why the Q4 babies are still at such a disadvantage. Perhaps there’s fewer of them. Here’s my humble opinion: you have to have truly extraordinary amounts of talent plus the right home environment to make the NHL. A July/Aug/Sept baby with that talent/environment will still make the cut for triple A. Why might an Oct/Nov/Dec baby not do so? He has to overcome a greater developmental disparity. And hockey is a mental game. Even if you have the hockey skills, will you have the confidence? Can you hold your own in the locker room?
By the way: another source of hockey data: http://www.hockeydb.com
There are also differences in who has children when, via Gelman, http://www.stat.columbia.edu/~cook/movabletype/archives/2009/09/who_has_babies.html
Good read!
Is there any money to be made in publishing a book entitled:
Talking numbers with Malcolm Gladwell ? Mr. Gladwell might want to buy all rights and copies for a hefty price.
Thanks.
Are there other sports in Canada with other cutoffs? For example, if I was born in Q4, I might find it more fun to play a sport where the cutoff is October 1, for example. Perhaps school sports rather than club sports have a different cutoff and thus lure athletes of different ages?
Steve: and if there were other sports with different cutoffs, that would also let us test the cutoff date hypothesis against the (say) vitamin D/seasonal veggies hypothesis.
Steve, I know that soccer is definitely by calendar year (Dec cut-off) as are pretty much all school based sports. Track is calendar year. The only non-Dec 31st cut-off I can find is the Sep 15th NHL draft cut-off. John Tavares is breaking all of Gretzky’s OHL records right now because as a post-Sept 15th player he has an extra year in Juniors.
It would be interesting to look at players born Sep 1-14th v. Sept 16th-30th to see if that particular cut-off (and the extra year in juniors) makes a difference. If anyone out there in cyberspace wants to write that paper, please let me know at frances_woolley [at] carleton.ca.
Digging around on-line I’ve found much better player databases – this is what draws people into the econ of sports, super-cool data! There are databases which include player height, which could be used to tease out some of the sunshine effects, since height is an excellent correlate of overall health/environment.
One possible experiment that one could do is see if there’s a difference in the Q1 premium pre- and post- intro of the birth control pill (and much greater control over timing of births) in 1962. Again, please let me know if you plan do write a paper on this, because I’m thinking it would be fun to do!
Interesting reading. Clearly having a year end cut off for a sport where size matters, and where the season doesn’t start on 1-January, favors children born earlier in the year. I have a 1-March child, and a 21-November child, and can see the effect. The data seems very good even up to the NHL level. While the later draft cut off may offset this effect a bit, it clearly doesn’t compensate fully. I suspect this is because the kids that were initially smaller often found another sport or simply gave up. At 8 years old, it is frustrating to play against kids who are almost 3 years older than you because they were born Jan 2 and you were born Dec. 30th (my birthday).
I’m confident we’d have more kids stick with many league sports if we lined age cut offs up better with the start of the season.
Todd