Freelance Friday: Birthday Bias in the NBA
By Guest Post
Freelance Friday is a regular series at The Nylon Calculus where we solicit and feature basketball analytics writing from the wider community. Today’s article comes to us from Jasper Wu and concerns the “Relative Age Effect” in the NBA. Jasper is the founder of ballandone.com and can be found on twitter @Ball_andOne
The Relative Age Effect (RAE) is a term used to describe the phenomenon of higher participation rates in the upper echelons of sport and academia by those born earlier in the year. Malcolm Gladwell found this effect, which he dubs “accumulative advantage”, to be prevalent among professional Canadian hockey players because the cutoff date for youth hockey leagues in Canada is January 1st. RAE has also been found to exist in the MLB, NHL and youth soccer. The premise behind this bias is straightforward; there is a huge physical developmental difference between young kids born in January vs those born in December of the same year. However, these kids usually play in the same age group in youth leagues. Naturally, kids born earlier will tend to perform better because of their physical advantage, thus being identified as talented and receiving more specialized training and attention. This small reward from being born earlier continually compounds, which results in a disproportionate number of athletes with early-year birthdays at the professional level.
More from Nylon Calculus
- Nylon Calculus: Reviewing preseason predictions for the Western Conference
- Nylon Calculus: Reviewing preseason predictions for the Eastern Conference
- Nylon Calculus: LeBron James and the slam dunk aging curve
- The Whiteboard: Why has the NBA seen so many 50 point games this season?
- Nylon Calculus: 15 early season predictions for the NBA Western Conference
I wanted to see whether or not this “Birthday Bias” exists in the NBA in terms of getting drafted. Is the NBA biased towards players born during certain times of the year?
To address, this question, I collected data from the last 25 NBA drafts. In total, this includes 1460 draft selections from 1992 to 2016. I also noted the date of the actual draft in each year so I could calculate how old a player was when he was drafted.
Here is the breakdown of births by month for the general U.S. population. These percentages are roughly the same for other countries as well:
And here is the distribution by month of all NBA drafted players in the last 25 years:
The most well represented months for drafted NBA players relative to the general population are clearly February and March, and the worst is December. This strongly implies that Birthday Bias does exist in the NBA as players born earlier in the year are more likely to be drafted.
To confirm that this bias is more than just randomness, I did a linear regression on the difference between % of Drafted players and % of General Population according to month.
The regression’s p-value is small (p=.0109), meaning that the coefficient is almost certainly significant. We can therefore conclude that the month a player is born in affects his chances of getting drafted.
Specifically, a player born in February or March has a 3-4% better chance of getting drafted than someone born in December. This is interesting, but somewhat counterintuitive. A significant majority (roughly 70%) of players drafted in the last 25 years are from the United States, but the cutoff for most U.S. Kindergartens is early September – meaning that a kid born October 1st has to wait almost an entire year to start Kindergarten. One more year of growth relative to his peers. This explains why August is the least represented month among drafted players – a kid born in August just beat the cutoff, so he was probably among the youngest and weakest physically in his class year. However, we should then expect to see a huge jump in September or October representation, which does not happen. An over-representation of February birthdays would imply that February 1st should be the cut-off for grade school or youth league enrollment, but this is not the case in the U.S.
What, then explains the proportionately large number of February birthdays drafted? At first, I thought it might be because non-US players may have a February cutoff in their countries, so an overwhelming majority of them would have February birthdays. But looking at all drafted players born in February, roughly 72% were American – right in line with the percentage for other months.
Another theory I came up with relates to the fact that basketball has very few barriers to entry, the most important of which is weather. Whereas in hockey or baseball you need equipment, to practice basketball all you need is a $20 ball, a hoop and good weather. Let’s assume that kids first start playing sports at age six. An October baby who just turned six can’t practice basketball unless he has access to a (crowded) indoor court, so he turns instead to all-weather sports like swimming and track. On the other hand, a February baby turns six right as winter is ending, so outdoor sports like soccer and basketball are his first sports. This theory is pure speculation at this point of course.
One thing I thought about was that the February spike might be related to the basketball season date range of roughly November to February. If we treat the end of basketball season as a cutoff date, then the birthday month proportion makes a lot more sense. In reality though, most basketball leagues (like the AAU) also have a cutoff of late August/September so those born in October should still have an advantage. However, I’m sure there is some explanation I haven’t thought of yet.
Here are some other interesting findings I came up with my age data. First, the average age of a draftee increases as we move later into the draft. On average, younger guys are drafted earlier, and older guys are drafted later.
This makes sense. The youngest players in the draft every year are “One-and-Dones” who are also the most talented, hence they leave school early. Older players are most likely fringe NBA guys who stayed in school trying to improve their stock further.
Second, the average age of a drafted player has trended downward over the last 25 years. Note the downward trend from 1992-2005 as more and more teams were drafting right out of high school, and note the spike after 2005 when the One-and-Done rule was implemented.
Lastly, I wanted to see if birthday bias exists for really good players (at least made an all-star team or all-NBA team). To define “really good”, I went through all 1460 players drafted in the last 25 years and marked the ones who made at least one All-star appearance or All-NBA team. I did not double count players if they made multiple all-star/all-NBA teams or Hall of Fame. Using Percentage of players drafted according to month as my baseline, I subtracted percent of all star players drafted, according to month.
The above chart takes into account the proportionate number of all players drafted according to month. For example, 8.22% of drafted players were born in June. The 1% value for June above means that roughly 8.22% + 1% = 9.22% of really good players were born in June. We can see that February and March are relatively the most common birth months for really good players, and January is the worst.
The Relative Age Effect is clearly present in the NBA to some degree. Looking at the birthdays of all players drafted in the last 25 years, we find that there are a disproportionate number of February and March birthdays relative to birth trends of the general population. This difference is statistically significant, meaning that it is likely more than pure chance.
Explaining why most drafted players are born in February is harder. Roughly 70% of drafted players are American, and the cutoff date for most schools and youth leagues in the US is September. This means that the oldest players are guys born in October, not February. Maybe this bias has something to do with the timing of good weather or end of basketball season in schools. I have yet to think of a thorough explanation though.
Additionally, we found that the average age of a draftee increases as draft pick number increases, which makes sense. Also, teams have recently been trending towards drafting younger players on average. Finally, despite the larger number of players drafted in February and March, those two months are still the highest relative birth months of NBA all-star or all-NBA players.
So there you have it. If you want to maximize your future kid’s chances at getting drafted, try to have him in February. I’m not really sure why February is the magical month, but there is no denying that the NBA is skewed towards February-born players.