The World Cup: Where Being Old Might Be Good
England never really had a chance in Brazil. And not for the reasons they normally ascribe to their failures at international tournaments—penalty kicks, not being Germany, etc.
No, England were done pretty much the moment manager Roy Hodgson decided to leave Ashley Cole off the national team. That’s not to say that Cole’s absence was the difference between England’s winning and losing (although he has been one of, if not the, best left back in the world for the better part of the last decade). But given how Hodgson filled out the rest of the 23-man squad—Raheem Sterling (19), Ross Barkley (20), Luke Shaw (18), and Alex Oxlade-Chamberain (20)—it was clear he was spearheading a youth movement within the Three Lions. And Cole was 33.
In 2010, then-manager Fabio Capello saddled England with the third-oldest squad in South Africa, behind only Brazil and Australia. England finished second in a group they figured to walk through, then went out meekly after being dismantled by a much younger and more dynamic Germany—die Mannschaft was about 3 years younger than England on average—in the Round of 16.
For 2014 Hodgson seemed determined to learn from his predecessor’s mistake. England headed to Brazil as the ninth youngest team in the tournament. And in their opener against Italy, a somewhat youthful England line-up showed an attacking flair that had largely been absent from many of their cautious displays that marked World Cup qualifying. From an aesthetic standpoint, Hodgson’s strategy worked. A young (okay, youngish—five of the six youngest players were on the bench) and energetic England went right at and created plenty of chances against the Italians. From a results standpoint, however, it didn’t matter much. England lost 2-1.
A few days later England played Uruguay. Lather, rinse, repeat. In the post-mortem, age seemed to be forgotten as a national narrative convenience. England didn’t fail even to get out of its group because the team was too young (inexperienced, naieve). England failed because it was England. But what if it was something else? What if age did have something to do with it?
It’s important first to note that the range of squad ages at the World Cup is pretty narrow, such that the difference between ‘young’ and ‘old’ isn’t terribly large to being with.
If you take the average age of every squad going back trough the 1990 World Cup, the first and third quartiles are between 26.08 years and 27.65 years. That means half of all the teams since Italy ’90—that’s 104 squads—are all within about a year and a half of each other. There aren’t many (okay, any) countries fielding 11 35-year-olds.
In fact, if you don’t count goalies, there are only 10 players that are 35 or older in the entierty of the 2014 World Cup. That’s 1.56% of the cummulative rosters.
So what? Right. Age is just an Arlo Whiteism. It’s a stat you drop into a broadcast because it’s a number and you have it. But that there’s rarely any context with it is part of the problem. Are young teams surviving matches in challeging heat/humidity conditions better? Do younger legs fair better the longer the tournament goes? Or does experience actually matter when you’ve got to navigate three games and one slip-up could send you home?
Before the 2010 World Cup in South Africa, then-Italian manager Marcello Lippi did the opposite of what Hodgson did this year. He named a squad littered with veteran players that helped him win the 2006 edition in Germany. As a result, Italy was old. The squad’s average age was a little over 28.7. And some of the players who were instrumental in helping the Italians win in 2006—Fabio Cannavaro 36, Mauro Camoranesi 33, Gianluca Zambrotta also 33—were really old.
Lippi took a bit of a beating from the Italian media for not incorporating more youth into the squad. His response was fairly reasonable: “The World Cup is about seven games over a month,” said the Azzurri manager, “We don’t necessarily need all the players to be 24 years old.”
Turned out Lippi was overly optimistic. Or simply wrong. Maybe both.
Italy’s Geezurri didn’t even make it out of what should have been a relatively easy group. They finished winless and dead last in a group containing New Zealand, the side widely regarded to be the worst to have qualified for the 32-team field. At least Italy’s AARPers we mercifully spared the full seven-game slate.
So being old at the World Cup, that’s bad.
Funy thing about that. That old Italian sqaud in South Africa? It was almost exactly the same age as the side that won everything in Germany four years prior (28.72 years in 2006 vs. 28.74 years in 2010).
So being old at the World Cup, that’s good.
Yes. Or maybe. It’s hard to say, but it might actually be better to be older. Not old, but older. And looking at every elimination round match from every World Cup back to 1990, there’s evidence to maybe support that. That’s a lot of equivocating, I know.
Taking the Round of 16 matches, it’s close to an even split. Which is to say that older teams have 23 wins and 25 losses. Clearly that’s not going to indicate anything other than age doesn’t appear to be a factor at all. But the Round of 16 is likely going to be the least all-things-being-equal of the knockout rounds. Match-ups are based off of group finishing position and some groups are weaker than others, just by the nature of the draw. So you’re most likely to end up with mismatches where one team is clearly better than the other and being older or younger doesn’t mean much.
However in the quarterfinal and semifinal rounds combined for those same six competitions, older sides are 24-12. That’s a .667 wining percentage. It’s not a huge sample—36 games—but it’s not tiny either.
It’s worth pointing out that a few of these differences are so miniscule as to be irrelevant. For example in 2002, South Korea—average age 27.13—beat Spain—average age 27.11. The older team won, but that difference tells you nothing really. And besides anyone who saw that match might recall South Korea won because of some really questionable calls in what was maybe the second-worst officiated match in World Cup history. In fact South Korea were only playing Spain because they had been the beneficiares of the worst officiated match in World Cup history in the previous round against Italy.
There are any myriad of factors that go in to winning. And when the margins are so slim, things like referee decision, dumb luck deflections, or individual moments of skill might have way more to do with determining wins and losses than something like age (where age is maybe a proxy for experience or maturity or some other hard to quantify quality).
Moreover the same average age could look different for different teams depeding on the distribution. How much do goalkeepers, who tend to be older, affect a given team’s average? With Luke Shaw, Alex Oxlade Chamberlain and Ross Barkley spending most of the first two decisive games on the bench what does it matter that England is “young” if Hodgson is still leaning heavily on his more experienced players? Isolating age effects, if any exist, just isn’t something that can be done simply by looking at average squad age.
Still for the most part, the average age of the finalists to have contested the last few World Cup falls within a remarkably narrow range, with most of the teams hovering right at 28 years-old give or take a few tenths. That is just above the third quartile mark.
West Germany : 27.692
South Africa ’10
The winning Spain and Brazil (’02) sides do skew a bit young, but even then only Spain is noticably below the mean of 26.84 years for the entire seven-Cup set of 208 teams (2014 teams are in the total pool of squad ages, but clearly we can’t put their results into the knockouts becase those matches haven’t been played and I’m not as clairvoyant as I used to be). It’s part of the reason that average age never seems to be a much-mentioned characteristic of World Cup winners. There’s nothing remarkable about teams who are, for the most part, a little bit older than average, especially when it’s usually true of both teams contesting the final.
But the roll call of World Cup semi-finalists is a pretty tight clique already. So this might not be capturing anything significant about age at all. Did Germany win its 1990 semifinal against England because of its 27.9-average-age-ness or because of its German-ness? England were actually the older side; shouldn’t that have been an advantage? Or was it nothing more than the match went to penalties and England are a bit shit at penalties?
The fact that most of the finalists from our sample average a touch under 28-years-old might just be reflecting something about their national team programs; specifically, that the countries that are good enough to win the World Cup are generally so full of wold-class talent that players are not likely to break in at 17 and are probably surplus to requirements by the time they’re north of 30. So you end up with a squad with some players between 22 and 25 and even more between 26 and 29. Brazil and Italy are the two cup-winningest countries in the history of the competition—they’ve claimed 9 of the 20 contested tournaments—and, since 1990, they have squads that average 27.907 and 27.995 years respectively.
This isn’t a claim about causality. It might be nothing more than a curiosity resulting from the fact that the range of ages just isn’t very large. England didn’t lose its first two games in Brazil because its average age heading to Brazil was only 26.56 years old. That’s stilly. They lost because they couldn’t stop Luis Suarez, something lots of teams have failed to do by the way. But there might be something about having an average age right about 28 that is indicative of some other factors that are significant in picking up wins or avoiding losses at the World Cup.
So, even if it was just a question of hedging bets, Hodgson might have been smart not to leave Ashley Cole at home.
A note on the chart and the numbers:
First, the chart was coded with D3. There’s a pulldown on the top left that allows you to select the year. It only goes back to 1990 for a few reasons. But generally, there are so many factors that change over time—fitness, sports nutrition, the mix of federation representation, skill level, socio-economic variables, etc.—that getting meaningful comparisons across generations is difficult. So I quasi-randomly chose 1990 as a cut-off. Even that is problematic in that the field of teams increased from 24 to 32 from ’94 to ’98.
But you can select any year from the pull down and get the average age of every squad in that tournament. If you mouseover any individual bar you’ll get a tooltip identifying the country and their age. It was spelled out explicitly in the piece, but it should be obvious that, save for 1998, the range of average squad ages at the World Cup is really pretty narrow.
Second, the numbers you are seeing for average ages on TV and elsewhere are likely inaccurate. If you watched the USA v. Ghana match might have noticed either a graphic or heard commenters make reference to Ghana having the youngest squad in Brazil with an average age of 24.9 years. The Ghanaians are indeed the youngest country at the World Cup, but that number is imprecise. The average age of Ghana’s actually closer to 25.5 years.
Why the difference? ESPN (or whatever service feeding ESPN if they didn’t do it in house) likely took each player’s age—just the year; so 22, 27, 21, etc.—on the start date of the tournament and used that for their calculation. And indded if you go to Wikipedia, grab each Ghanaian player’s age and average them out you get 24.91.
But for the most part, players don’t have their birthday the day the tournament starts so they aren’t exactly 22 or 27 or whatever when the World Cup kicks off. Think about it: for every player whose birthday was the day after the tournament started, it’s just as likely there is one whose birthday is 364 days away; or for every player whose birthday is three months after, there is likely one whose is nine. Average the two out—three months and nine months—and you get half a year between them.
As a result almost every squad’s age is underestimated by about 6 months. Wikipedia also lists player birthdays and, sure enough, using those, the Ghanaian squad’s average age turns out to be 25.45 years; or almost exactly half a year older than the number ESPN is using. This is a table of all of the differences between just using the raw age year and the more precise birthday to calculate the age for every squad since 90 (the NA’s means the team didn’t qualify that year). You can see how all of the values are very close to .5 (half a year).
It might seem a little nitpicky, but A) six months is a measurable percentage of a soccer player’s career (it’s basically the bulk of another season) and B) using birthday can change the order slightly. For example, Portugal goes from being second oldest to fourth oldest squad in Brazil. Colombia goes all the way from the bottom half (youngest) to top half (oldest). Okay it’s only a move from 17th to 15th, but you do get different outcomes when you use the more precise method.