Episode 20 – Beyond Watts per Kilo: What the Data Can’t Tell You About Future Pros (interview with John Wakefield)

Talent identification in cycling sounds simple — find the rider with the best numbers and sign them. In the latest Cycling Science Podcast, John Wakefield, Head of Scouting at Red Bull BORA-hansgrohe, explains why it’s anything but.

There was a time when finding a future professional cyclist came down to two questions: what are this rider’s results, and what’s their watts-per-kilo? On the latest episode of the podcast, John Wakefield — performance coach and head of scouting for the Red Bull BORA-hansgrohe rookie programme, and co-author of the 2022 “Compound Score” paper — made a convincing case that both questions, on their own, now mislead more than they reveal.

An arms race for the young

The backdrop is a peloton that has transformed in barely five years. “It’s become an arms race for the younger generation,” Wakefield says. Where teams once sent a 23-year-old to a Grand Tour for two weeks of experience, they now send 20-year-olds to win. Every junior has a power meter, a manager, and access to almost the same equipment and data as the World Tour. The result is a development pipeline under enormous pressure to find the next Pogačar before anyone else does.

The Compound Score, explained

That pressure is what the Compound Score was built to handle. The idea is simple and elegant: take a rider’s five-minute maximal power and multiply it by their watts-per-kilo from that same effort. A 70 kg rider producing 450 W at five minutes is doing roughly 6.4 W/kg — multiply the two and you get a compound figure of around 2,900. The point is that it rewards both absolute power (the engine that drives a breakaway or holds the front of the bunch) and relative power (the climber’s currency) in one number, rather than forcing you to choose between them.

Crucially, the framework hasn’t stood still. Wakefield and colleagues have since added a durability dimension — testing riders not when they’re fresh, but after a substantial workload — and a “mixed” compound score across several effort durations. As he put it, the under-23 and under-19 ranks have “moved on,” and the tools have had to move with them.

Where the numbers run out

The most striking part of the conversation was Wakefield’s candour about what data cannot do. VO₂ max, long treated as the gold-standard selector, “is not it anymore,” he argues. Confusing maturation timing with genuine talent is, in his view, another trap. And no spreadsheet can tell you whether a teenager will cope when the wheels come off — because, statistically, they will. “You can’t, at 17 or 18 years old, say a kid is going to be a GC rider,” he says. “If someone’s doing that, they’re selling a story.”

The biggest single reason promising riders don’t make it isn’t physiological at all. It’s “wanting too much too soon” — skipping steps, chasing a World Tour contract at 18, and getting lost in the system before their potential is realised. The antidote, Wakefield believes, is a strong off-bike education and support structure: everything from handling a setback to, quite literally, knowing how to submit an invoice.

The practical takeaway for coaches

For coaches working with talented juniors and none of a World Tour team’s resources, Wakefield’s advice is refreshingly low-tech. Be patient. Track relative versus absolute power consistently over time rather than chasing a single lab number. Don’t put every kid in the lab. And be careful what you tell them — “what you tell a kid, they believe,” so support beats pushing, every time. He’s wary of riders who become “robots,” reading their bodies off a screen: ask some of them what 400 watts feels like in the legs, and they can’t tell you.

Will AI take over?

And the future everyone asks about? Wakefield expects AI to make scouting faster and decisions better-informed — sifting three riders in the time it once took to assess one. But he’s firm on its ceiling: “AI only tells you what you teach it. I definitely don’t feel that the human side is gone.” The machine can shortlist. For now, a human still picks the champion.

---

Listen to the full episode on Spotify or watch on YouTube. The Compound Score paper (Leo, Spragg, Wakefield & Swart, 2022) is published in the Journal of Science and Cycling.