A twenty-two-year-old striker is a prospect. A twenty-two-year-old goalkeeper is a risk somebody has decided to take.
That difference is not sentiment, and it is not conservatism for its own sake. Goalkeepers age differently to everybody else on the pitch, they develop on a different timetable, and the shape of their careers looks nothing like the curve that governs the other ten players. Everyone in the game knows this in a vague way. Very few recruitment decisions actually reflect it.
Which is a problem, because the position is where the age question has the highest cost. Get it wrong with a winger and you have wasted a squad slot. Get it wrong with a keeper and you have wasted three seasons, because you only have one of them and he plays every week.
The Curve Is Flatter, Not Just Later
The standard finding for outfield players is a peak somewhere around twenty-six or twenty-seven, with a fairly brisk decline after thirty and very little top-level football left by thirty-four. Sprint speed, repeated high-intensity efforts and recovery between them all fall away, and no amount of accumulated wisdom compensates for a full-back who can no longer get back.
Goalkeepers do not follow this. The peak arrives later, usually somewhere between twenty-eight and thirty-two, but the more important point is that the curve is much flatter on both sides. A good keeper at twenty-nine is often still a good keeper at thirty-four. Peter Shilton passed a thousand league appearances and was still playing professionally at forty-seven. Edwin van der Sar won a Premier League title at forty and retired as a first-choice keeper rather than a squad one. Brad Friedel was playing Premier League football into his forties. Gianluigi Buffon did not stop until forty-five.
None of those are freak cases in the way that a forty-year-old outfielder would be. They are the extreme end of a distribution that genuinely extends that far.
Why the Position Ages Differently
Two things are going on, and they pull in the same direction.
The first is straightforwardly physical. A goalkeeper covers a fraction of the distance an outfielder covers and performs a fraction of the high-intensity sprints. The cumulative load on the body across a fifteen-year career is far lower, which means the systems that degrade fastest in outfield players are the systems a keeper leans on least. Explosive power still matters enormously, but explosive power in short, isolated bursts holds up much better than the ability to produce it forty times in ninety minutes and then do it again on Wednesday.
The second is that goalkeeping is more cognitive, relative to its physical demands, than any other position. Positioning, angles, reading a cross before it is struck, knowing when to come and when to stay, organising a back line, recognising the body shape of a striker about to shoot. All of that improves with repetition and none of it declines in your thirties. A keeper is accumulating the thing that matters most at the same time as he is losing a little of the thing that matters second most.
That trade sits in his favour for a remarkably long time.
The Skills Do Not Age Together
Treating goalkeeping as one ability is where most of the confusion comes from. It is at least four abilities, and they run on separate clocks.
Reflex shot-stopping ages best. Reaction time does slow with age, but the decline between twenty-five and thirty-five is small, and better positioning means an older keeper is often reaching shots more comfortably than he did at twenty-five, when he was arriving at them late and making them look harder.
Command of the box ages worse. Claiming a cross in traffic requires a standing jump, timing and a willingness to take contact in the air, and vertical leap is one of the first things to go. This is the phase you tend to see first: a keeper who used to come for everything starts staying on his line, and the punches outnumber the catches.
Lateral explosiveness sits in between, and tends to hold until the mid-thirties before falling away with some speed.
Distribution is the interesting one, because it barely ages at all — and that is precisely why it ends careers. Passing does not decline. But the standard required of it has moved so far in fifteen years that a keeper who built his game in 2010 without a reliable left foot or the composure to play out under a press is finished long before his body says so. The skill that finishes most modern goalkeepers is not one they lost. It is one they never had.
The Late Start Problem
The other half of the curve gets discussed far less, and for American football it matters more.
Keepers arrive late. Part of that is developmental — the coordination, size and decision-making the position demands genuinely take longer to assemble than the skill set of a winger. But most of it is structural. A squad carries three or four keepers and plays one. There is no rotation, no fifteen minutes off the bench to get a feel for the level, no easing in during a cup tie against lower-league opposition, because the cup tie is usually the backup’s only start of the season.
The result is a bottleneck. A twenty-year-old midfielder gets three hundred minutes and a manager’s judgement about whether he can handle it. A twenty-year-old keeper gets a bench seat and a guess. Gianluigi Donnarumma playing for Milan at sixteen is remembered as remarkable because it is remarkable, not because it is the start of a trend.
So the pre-peak part of the curve is not really a talent curve at all. It is a curve describing when clubs are willing to find out, and clubs are willing very late. A keeper who does not get a settled starting job until twenty-six has not necessarily developed slowly. He has more likely spent four years being denied the only thing that develops him.
The Cliff, Not the Slope
Here is the part that catches recruitment departments out. The curve is flat for a long time and then it is not flat at all.
Outfield decline is usually gradual and legible. You can watch a player lose half a yard across two seasons and plan for it. Goalkeeper decline is much more likely to arrive as a step change: a hip or groin injury that takes the dive out of one side, a run of three high-profile errors that turns into a confidence problem, a new manager who wants the ball played out and a keeper who visibly cannot.
Iker Casillas going from unquestioned to questioned in the space of about eighteen months is the archetype. Joe Hart losing the City job and the England job at twenty-nine, an age at which most keepers are entering their best years, is the version where it happens early.
Planning for a slope when the reality is a cliff is how clubs end up buying a replacement in a panic in August.
Telling Decline Apart From Bad Luck
None of this is easy to measure, and the honest version of the analysis has to admit that.
Goalkeeper shot-stopping numbers are noisy. A keeper faces something in the region of a hundred to a hundred and fifty shots on target across a league season, which is a small sample by any standard, and post-shot expected goals minus goals conceded — the best public measure available — carries weak year-to-year correlation as a result. A keeper can post a strong figure one season and a poor one the next without anything about him having changed. Two bad months is not evidence of a thirty-three-year-old declining. It is barely evidence of anything.
This is the same trap people fall into wherever a headline number is quoted without the sample behind it. The questions you would ask before trusting any advertised figure — how many observations produced it, over what period, and what happens to it when you extend the window — are exactly the questions to ask about a save percentage. It is the reasoning a careful reader already applies to football tips, and it applies just as well to a goalkeeper’s season. A number is only as good as the thing that generated it.
The practical version: judge shot-stopping over three seasons, not one. Judge aerial command and distribution over three months, because those you can actually see, and they degrade for reasons that show up on video rather than in a percentage.
What It Means for Recruitment
Put the two halves together and the recruitment argument builds itself.
If keepers peak between twenty-eight and thirty-two and hold that level into their mid-thirties, a twenty-nine-year-old is not an ageing signing. He is a keeper at the front of his best years, and clubs that treat that age as a warning sign are misreading the position entirely.
But the same maths cuts hard the other way past thirty-two. At that point you are buying the flat tail of a curve, paying for the reputation earned during the peak, and taking on the cliff risk without any of the upside. Add the adaptation cost of a new league, which for a keeper means new defenders, new communication, unfamiliar crossing patterns and often a new language to organise a back four in, and the first season is frequently written off before the decline even starts.
This is the case against MLS clubs importing keepers in their thirties, and it has held up well. The signings rarely deliver more than a solid year or two, they cost more than the domestic alternative, and — the part that compounds — they occupy the only starting spot available, which is precisely the reps a twenty-four-year-old American keeper needs in order to become the twenty-nine-year-old who would have been better and cheaper.
The bottleneck that makes keepers arrive late is a bottleneck clubs create, and every veteran import tightens it.
The Short Version
Goalkeepers peak later than everyone else, stay near that peak far longer, and fall off it faster when they finally go. The parts of the job that require athleticism decline on schedule, the parts that require judgement improve for years afterwards, and the part that ends most careers is a technical standard that moved while the keeper stood still.
Twenty-nine is not old for a goalkeeper. Thirty-four is not necessarily old either. But twenty-four is not young, and treating it as though it were is the reason so many good keepers never find out how good they were.