The "Age-Bias" Effect is Alive and Well

[Karl K]

Some time ago, JohnR put me onto a well done study on the age-bias effect in youth soccer.

Here is the link to the article; it was broken last I looked, but it may come back.

http://www.psychologie.uni-bonn.de/...sch/ageeffx.htm

This is a well done academic article about what the authors call a "systematic discrimination" against kids born later in the competition year.

They studied professional leagues in Germany, Brazil, England, France, Australia, Japan and found that an abnormally large percentage of the current professionals were born in the first few months of their youth age group, with an abnormally small percentage born late in the year.

Reason is simple. Coach looks at a group of little kids, says I'm going to pick the best players out there because they are the "prospects." They tend to be fairly old for their age groups, because a few months make a big difference among little kids.

These chosen kids then not only retain their relative age advantage throughout their youth soccer years, but they get the better coaching, positive reinforcement, etc. so that the early decision to label them as "better players" becomes a self-fulfilling prophecy.

I sent the link to this article last year to one of the directors at my club, a guy who had a 10-year professional career, and here is what I wrote in my email to him:

"I bet if you were to do a study of ODP teams, which work on the calendar year, you would see this phenomena show up dramatically, especially since the evaluation there is done so haphazardly and infrequently."

I wrote that over a year ago.

Well, the relative age bias effect is alive and well.

First, check out the birthdays for the Region II team of '88s heading to Italy for competiton in mid-April. See:

http://www.region2.com/2003boysTeam88Italy.pdf

Note:

--Only 5 of the 16 are currently u14s according to the school year 8/1/88 cutoffs.

--10 of the 16 have birthdays is the first half of the calendar year.

--9 of THOSE 10 have birthdays in the first QUARTER of the calendar year.

I sent this info along to JohnR, and he put me onto the lastest Top Drawer article on the U15 National team. Check out the new 24 player pool for the U15s. See:

http://www.topdrawersoccer.com/Nati...1048099599/view

12 with birthdays in the first quarter, 8 in second quarter, 2 in third quarter, 2 in fourth quarter. There is even a (slight) tendency toward older kids within each quarter! Five January birthdays, 4 April birthdays, etc.

The effect is just huge for this group.

There is often talk about players falling through the cracks because of lack of money or because of lack of access...

But maybe, just maybe, the BIGGEST bias in youth soccer is simple reverse-agism. Maybe it's not the prejudices some think -- the bias against kids who are poor, are from the inner city, and aren't privileged suburbanites...

Instead, it seems that simmering below the surface is an unconscious discrimination against kids who are, in fact, simply just "too young" at the time.

[CSheard]

I've noticed this as well. [geek alert:] Over the past couple of years, I've kept a running file of the US national pool players, from veterans all the way down to 15 year olds (based upon actual caps, training camp call-ups, awards, playing with the national youth teams, other indicators of potential, etc.), and consistently it always comes to my attention as I enter birthdays that the predominant dates are early in the year (even seemingly more often in January itself). Although you point out the likely age-bias, it always made me wonder if players/parents weren't deliberately faking their ages to play a year down, in order to have the physical advantage. I'm sure at some point, birth certificates or passports or other identification is used to actually verify player ages, which would obviously prevent such fakery, but it always felt inconceivable that so many of our players were born in January.

[JohnR]

Karl -

Repeating myself from another post, but ...

Think about it this way. The USYSA took 20 kids from the January - June 1988 pool. Took 4 kids from the July - December 1988 pool.

I betcha that, on the whole, players #5 through #10 in the July - December 1988 pool are better long-term prospects than players #15 - #20 in the January - June 1988 pool. The first group are Top 10 players within a given 6-month period, while the second group are hugging the bottom of the Top 20.

But the first group is out and the second group is in.

Which kids do you think are likelier Bradenton candidates, the first group or the second group?

Yes, I'd say that this issue has significant affect on the player pool for future U.S. national players.

Of course, this subject being a hobbyhorse of mine, I would say that. :)

[JohnR]

Originally posted by CSheard
I'm sure at some point, birth certificates or passports or other identification is used to actually verify player ages, which would obviously prevent such fakery, but it always felt inconceivable that so many of our players were born in January.

I admire your skepticism! A man after my own heart.

But no, it's not fakery. A few months in age makes a huge difference in performance when you are evaluating a bunch of otherwise evenly matched (i.e., top) players.

The age effect holds true across European and South American countries, too.

[CSheard]

Just crunched actual numbers from my woefully un-scientific files: out of 372 players for whom I have birthdates (and that presupposes that I have the dates correct - I'm constantly coming across conflicting dates in media articles): 124 were born in the months of January, February and March, when "mathematically", only 93 should have been born in that first quarter. I say mathematically, because I don't know if actual birth rates stay straight line throughout the calendar year.

That said, I believe my local soccer club breaks the ages off as of Aug. 1, as opposed to January 1, which would mean my two children born in May will be the younger kids. Oh well - they'll just have to practice twice as much and prove they are better than the bigger kids.

[JohnR]

CHeard -

You will find that on the better "A" teams at the U9 through U11 level, with the August 1st cutoff, there are about 3x as many kids born in the August - October time period than born in May - July.

Some of these teams, you can spot instantly. During warmups, just from the size & physical maturity. We'll say, "Ah, an autumn team." Whereas we'll sometimes mistake the "B" teams for teams of another year down, because the B teams tend to made up of June birthdays -- just a couple of months from the next year's cutoff.

The odds aren't stacked in your kids' favor. Still, not an impossible task. On my son's current team, the three strongest players are born in February, April, and July. So, stranger things have happened.

[Karl K]

My kid has an early August birthday, which puts him at the beginning of the competition year for the school-year scheme we use in the USA.

When you see him with his teammates, he is among the larger and more mature kids.

But when he has gone to ODP tryouts and training sessions, where the cutoff is 1/1 and he is a "late" birthday in the competition year...well, guess what??

In that context, he is amongst the smaller, less developed kids.

As the song goes, you always have to ask, "compared to what?"

[ripmstr]

It seems to be the kids who have been chosen for '98 and '90 ODP in our region in FL with late birthdays are the ones who are well into puberty.

[The Wanderer]

How many players are we missing out on because of this? Any ideas?

When we say that the majority of the players come from the first half of the even numbered years, do we know how big the whole pool was at that time? For example, was there a larger number of candidates available from the first quarters of the year when compared with the last ones?

[whip]

I really hope that here in USA coaches look at the situation from a complete different prospective: T-A-L-E-N-T.......so far the hiring of young soccer players have prove to be a complete success as a prove of it "TAYLOR TWELLMAN" has prove the theory that there homegrown talent here in USA

[JohnR]

Originally posted by The Wanderer
How many players are we missing out on because of this? Any ideas?

When we say that the majority of the players come from the first half of the even numbered years, do we know how big the whole pool was at that time? For example, was there a larger number of candidates available from the first quarters of the year when compared with the last ones?

?

The law of large numbers means that, from 6 month period to 6 month period, there won't be a lot of variation in the quality of (say) the top 25 players.

OK, it is true that when you're looking at the top 5 players you could run into a situation where the best kid in one 6-month period isn't as good as the 5th best kid in another period. But for ODP purposes, when filling a 24-man roster, one period should be much the same as the next.

Here's my take on the situation. Of the top 75 players in the U.S. over each 2-year period, the ODP national teams miss 25 of them due to the age effect. (For purposes of this exercise, I'll assume that ODP identification is otherwise perfect ... of course, it's not, so the true numbers of top players missed is lower, but I'll ignore that complication here.) Maybe one of these 25 kids who failed to make national ODP will make MLS before the age of 21.

That leaves 50 kids. Of those 50 kids, 25 are born in even-numbered years and are put on the professional fast track via Bradenton. Maybe 7 or 8 of these will make MLS before the age of 21.

The others, born in the odd-numbered years, don't go to Bradenton. Maybe 3, 4 of them will make MLS before the age of 21.

In summary, right now 10 to 12 of those best 75 kids in a two-year period might be a pro before the age of 21. But if all of them were treated as well as the early-year Bradenton kids, the number would be 20 to 25.

So that's my best estimate, that if you cured the age effect and made Bradenton available every year, rather than every other year, you'd double the amount of young U.S. pros.

[The Wanderer]

Originally posted by JohnR
?The law of large numbers means that, from 6 month period to 6 month period, there won't be a lot of variation in the quality of (say) the top 25 players.

I agree, however, I was referring to the quantity of players. If there were more available players born in the first 6 months of a year when compared to the last six months, that would obviously skew the data. It's probably not true anyways, just a ?.

[beineke]

Originally posted by The Wanderer
I agree, however, I was referring to the quantity of players. If there were more available players born in the first 6 months of a year when compared to the last six months, that would obviously skew the data. It's probably not true anyways, just a ?.

Are you suggesting that kids born in Oct or Nov would be less likely to play soccer, anyway? This is certainly possible, since being the oldest in their grade makes them more likely to get drawn onto school sports teams instead of club teams ...

[The Wanderer]

Originally posted by beineke
Are you suggesting that kids born in Oct or Nov would be less likely to play soccer, anyway? This is certainly possible, since being the oldest in their grade makes them more likely to get drawn onto school sports teams instead of club teams ...

Exactly. Kind of like when we hear about how few black coaches there are in college football without considering how many qualified candidates there are available.

[JohnR]

Originally posted by beineke
Are you suggesting that kids born in Oct or Nov would be less likely to play soccer, anyway? This is certainly possible, since being the oldest in their grade makes them more likely to get drawn onto school sports teams instead of club teams ...

The kids born in October and November certainly do play club soccer. Indeed, they are the dominant players in club soccer, given their age advantage.

Look up most top youth teams in the U9 through U13 ages (and probably later, too) and you'll see many late-year birthdays. More kids born in the fourth quarter of the year than in the first quarter.

Yes, it is possible that some of these club-soccer youth stars do give up soccer for other team sports when they hit high school. But by then, they've already been given the hint by failing to make ODP teams that maybe they should consider another sport.