SwimSwam's 2028 Recruiting Database: Pricing a Swimmer Before the Contract Exists

1:57.8 in the women's 200-meter butterfly. For most viewers, that is a sequence of digits that scrolls past the bottom of a television screen. For a head coach at an NCAA Division I program, it is a data point that can decide how to allocate part of a four-year scholarship fund to a seventeen-year-old athlete.

SwimSwam, the specialist swimming news outlet founded in 2026 by Olympic gold medalist Mel Stewart, has launched a product named the "2028 Recruiting Database." In its raw form, it is a public spreadsheet: athlete names, high schools, club affiliations, personal best times by event, and — most importantly — the university each athlete has committed to for the high school graduating class of 2028. Anne Lepesant, SwimSwam's long-time recruiting writer, operates and updates the database.

I once assumed a table like this was just a reference tool. I was wrong. After several seasons following American college swimming, I have come to see that a recruiting database does more than record reality — it helps produce it. A name that appears on the list, gets shared, gets quoted in coaches' group chats, acquires a market price. And once that price is set publicly, it is very hard to bring it down.

Context: a market with no exchange but with listed prices

To understand why a recruiting database qualifies as an event worth analysing, you need the structure of American college swimming. The NCAA divides competition into several divisions; Division I concentrates the money, scholarships, facilities and media attention. Swimming and diving sits in the group of sports that use equivalency scholarships. Each school holds one total scholarship fund, and the head coach has full discretion to divide it: one full award to a single elite athlete, or half an award each to eight different swimmers.

According to NCAA published figures, Division I women's swimming and diving carries fourteen equivalency scholarships; the men's side carries nine point nine. This is the crux. Because the fund is not split evenly, every award is a capital allocation decision, and every capital allocation decision needs input data.

The recruiting calendar sets the tempo. In Division I, coaches may contact prospects directly only from 15 June following the prospect's sophomore year of high school. Before that date, contact is tightly regulated. Official visits to campus may begin on the first day of junior-year classes. National Letters of Intent have two main signing periods: an early period in mid-November and a regular period in mid-April.

For the graduating class of 2028, the direct contact window opens on 15 June 2026. This is the most notable detail in the whole story of the SwimSwam database. The database is published and updated before the official contact window opens, which turns it into a valuation instrument before the market is permitted to trade. Coaches cannot call. They can read. And they read very carefully.

Three structural changes over the past decade have heated this market. The NCAA transfer portal launched in 2026, allowing athletes to register transfer intent publicly. From 2026, transfer rules permitted immediate eligibility, removing the requirement to sit out a year. Also in 2026, name, image and likeness rights were recognised, opening a new revenue stream for athletes. More recently, the settlement in the House v. NCAA case proposed replacing the scholarship-limit model with a roster-limit model, with a proposed cap for Division I swimming and diving in the region of thirty athletes per team.

Each of those changes raised the value of information. Each of them also raised the value of whoever holds the database.

Core analysis: how a spreadsheet becomes a market

One thing needs saying plainly: a recruiting database is not a ranking. It is a registry. But in a market short on transparency, a registry functions as a ranking, because order of appearance, number of interested programs and timing of commitment all carry information.

When an athlete commits to a school, that record tells the rest of the market three things: that this athlete has been valued highly enough by one programme to receive an offer, that this athlete is no longer in the free supply pool, and that the programme in question has spent part of its scholarship fund.

In swimming, an athlete's value is measured in time, but time is only the crude input. What sets real value is the gap between the athlete's personal best and the national championship qualifying standard, and how much room remains to close that gap at seventeen years old. An athlete at 1:57.8 at seventeen and an athlete at the same time at fifteen are entirely different assets. The second has two more years of development ahead; the first has largely reached the biological ceiling of the age group.

This is why I attach a limitations section to every analysis. A recruiting table records performance; it does not record maturation rate. It records swim times; it does not record arm span, height-to-span ratio, metabolic endurance or training background. And it certainly does not record the hardest thing to measure: the capacity to endure four years of high-volume training without collapsing through a shoulder injury.

On cost, the figure worth putting on the table is the price of one year at a private Division I school, typically somewhere between sixty thousand and more than eighty thousand US dollars, covering tuition, room, board and mandatory fees. Multiplied across four years, a full scholarship has a nominal value in the hundreds of thousands. When the fund is split into half awards, the institution's contribution halves, but the family still carries the rest. That is why families follow this database even more closely than coaches do.

The data pipeline behind the spreadsheet also deserves scrutiny. High school athletes' results come mainly from two sources: the national governing body's meet results system and school-level competitions. Aggregator platforms such as SwimCloud collect, standardise and cross-reference both sources into comparable athlete profiles. SwimSwam adds one more layer: the reporter layer. Lepesant verifies commitments through multiple channels, cross-checking with club coaches and sometimes with the athletes themselves.

When an editor says no, I learn to listen to the data. I first wrote that line after a piece was rejected for being too complicated for a general readership, and I still keep it as a professional principle. In this case it means: the value of a recruiting table lies not in whether it is right, but in whether it is cross-verified.

There is a phenomenon I call early commitment. Over the past two decades, the average commitment age in college swimming has fallen sharply. More and more athletes decide before entering their junior year of high school. Early commitment is not a sign of early maturity; it is a sign that the information market is running faster than the physical market. When every coach reads the same table and sees the same name at the top, the pressure of collective action produces a race to the bottom on timing.

I do not argue with emotion; I present a chain of data. And the chain here reveals a paradox: the more public information there is, the less time there is to evaluate. A coach gets three months instead of three years, and in those three months must rely on a third-party aggregation. That is the point where the database shifts from a descriptive tool to a generative one.

From my own observations across several seasons, large-conference programmes tend to close their recruiting classes earlier and concentrate on a handful of priority events. Smaller programmes compensate by targeting athletes whose development curve is incomplete — that is, targeting the part of the data the public table does not display. In other words, the competitive edge lies in reading what the table does not write.

The counterintuitive angle: the database is run by a party with industry interests

This is the part I consider most important and most easily overlooked.

The 2028 recruiting database is not published by an independent regulator. It is published by SwimSwam and operated by Anne Lepesant, a person holding a key position within the very industry the database describes. That does not make the data wrong. But it does give the data an implicit promotional function: the more attention the table receives, the higher the reputation of the publisher, and the greater its commercial value.

Being right too early is also a form of rejection. I have been through that with a forecasting model returned because it ran ahead of when readers were ready. Here, the question is not whether the table is accurate, but what the table measures and what it omits.

Technically, the table measures very well the things that are easy to measure: times, placings, schools, commitment dates. Systemically, it reflects a narrow definition of value: an athlete's value equals their personal best at seventeen. That definition ignores three factors that decide success at college level.

The first is relay depth. A championship team does not win with four star athletes; it wins with sixteen relay legs of comparable quality. The second is the development trajectory in the third and fourth years. Many athletes make large performance jumps at nineteen and twenty, once the physical base is thick enough to absorb heavy training loads. The third is retention. In college swimming, the number of athletes who leave a programme before graduating is not small, and the free-transfer wave since 2026 has only pushed that number higher.

A database cannot forecast retention, because retention depends on team culture, on the athlete-coach relationship, on the ability to handle academic pressure. Those are variables with no column in the spreadsheet.

The second problem is correlation being mistaken for causation. An athlete's appearance in a reputable database does not make them swim faster. It is the swimming fast that puts them in the database. That sounds obvious, but the consequences are not: once the table becomes the shared reference point, programmes start recruiting to the table rather than to their own specific tactical needs.

I have seen this in another field. My research on spectator-free matches during the pandemic taught me a methodological lesson: removing a single variable — crowd noise — changes outcomes, but not in the way raw data predicts. With recruiting tables, the variable removed from the model is the training process, and that is the largest variable of all.

The third problem is self-reinforcement. An athlete at the top of the table draws more attention, therefore more racing opportunities, therefore more data, and therefore stays at the top of the table. An athlete ranked more modestly gets overlooked, with fewer opportunities and less data. The database does not create inequity, but it amplifies inequity that already exists.

Every transfer deal is a problem waiting for a solution. And in swimming, the problem usually has more than one unknown.

The blind spot: relays and the third year

If forced to name a single largest gap in any swimming recruiting database, I would point at the relays.

In college swimming, team championship points are distributed across many rounds. Relay legs account for a large share of total points, and they reward depth rather than isolated peaks. A team with four outstanding athletes in four different events can lose to a team with twelve decent athletes in those same events.

The recruiting table ranks individuals. Coaches need rosters. The gap between those two needs is exactly where competitive advantage is created.

Based on my experience following competitions over the years, championship programmes tend to share one trait: they recruit athletes who sit outside the top of the table but whose season-over-season improvement index runs above average. These athletes often come from small clubs, race few national meets, and so have personal bests that understate their true level.

This is the kind of information no public database can fully supply, because it requires longitudinal individual data across multiple seasons, together with training context and competition schedule. A recruiting table records a moment. A championship programme needs a curve.

The match is over, but the data is still playing stoppage time.

What to watch in the next cycle

When the contact window for the class of 2028 opens on 15 June 2026, I will track three specific signals.

The first is the average time from a name appearing on the table to a commitment. If that interval keeps shortening, the market is operating under information pressure rather than quality pressure. That is an early indicator that transfer risk will rise in the following years.

The second is the number of commitments coming from small clubs that receive little media attention. A rising share means programmes are trying to read the hidden data rather than buy the visible data. A falling share means the table has become the sole yardstick, and tactical diversity in college swimming will narrow.

The third is the effect of the roster-limit model, should it be applied. When a roster cap replaces a scholarship limit, the incentive to split the scholarship fund changes. If a team may keep only about thirty athletes, granting a half award to a slow-developing athlete becomes far more expensive than concentrating capital on an athlete who already meets the national qualifying standard immediately.

Amid the noise of the stands, I choose to sit with the numbers. But I sit with the numbers in order to see what the numbers leave out.

SwimSwam's 2028 recruiting database is a good and useful product. It gives the public access to a market that used to operate in the dark. But readers should remember that every database has a boundary — the line between what is measured and what matters. In swimming, that boundary sits where the stands fall silent: five a.m. training sessions, months of shoulder rehabilitation, and a fifteen-year-old nobody has heard of who will stand on a podium at twenty-two.

Data does not feel. People do, and that is the part we have not yet put into the spreadsheet.