The number that built the super-shoe era is 4 percent. It is real, it is peer-reviewed, and almost nobody reading this will ever collect it. The study that produced it tested high-caliber runners at 14 to 18 kilometers an hour, which is roughly 4:35 to 6:08 per mile, the pace band between a sub-three-hour and a sub-two-and-a-half-hour marathon. Eighteen of eighteen runners improved, by an average of four percent, with no drop-off across those speeds. It is a clean result. It was also funded by Nike, and it was measured at a pace most age-groupers will never race.
Run the same shoe slower and the number comes apart. An independent group, working from a university grant and a crowdfunding campaign with every pair bought at retail, put the Vaporfly against a weight-matched plain shoe at the speeds real recreational runners hold. At 12 kilometers an hour, about an 8:03 mile and a 3:31 marathon, the saving was 1.6 percent. At 10 kilometers an hour, a 9:39 mile and roughly a 4:13 marathon, it fell to 0.9 percent, and at that speed the result was no longer statistically distinguishable from zero. The authors were blunt about the pattern: below about 13 kilometers an hour, the benefit is speed-dependent and shrinking.
Now place most age-groupers on that curve. In the 2021 Boston Marathon, a competitive field, only 21 percent of finishers broke 3:15. At London, 10 percent did. The median marathoner in the United States runs about 4:25, which works out to a hair over 10 kilometers an hour, sitting right at the bottom of the curve where the saving is small and statistically silent. The 4 percent lives at a pace that most of the people buying the shoes for it will never run.
The reason is mechanical, and it is the part that makes the whole thing make sense. The foam in these shoes is a spring, and a very good one, returning around 86 percent of the energy you put into it. But a spring only gives back what you load into it. A fast runner strikes the ground at close to three times body weight and compresses that foam fully. At jogging-to-tempo pace the impact is closer to 2.1 to 2.3 times body weight, not enough to wring the full return out of the foam, and the stiff carbon plate is tuned for a faster turnover than a slower runner provides. You have to hit the ground hard enough to collect the rebound, and most age-groupers simply don't.
It also matters which super shoe. When that same independent lab tested seven carbon-plated models head to head, the spread was wide. Nike's Alphafly and Vaporfly and the Asics Metaspeed landed around 2.5 to 3 percent. But the Hoka and Brooks models in the test were statistically no better than a plain racing flat. The category name promises a lot that several shoes wearing it don't deliver, and the response is individual on top of that: even among trained runners the original study ranged from 1.6 to 6.3 percent, and at slow paces the average sits close enough to zero that some runners get nothing, or slightly worse. If you are going to spend the money, race in them in training first and find out which side of that line you're on.
None of this makes super shoes a waste, and the honest version of this story has to say so. They are light, and a lighter shoe saves a little economy at any pace, independent of the foam. They appear to reduce muscle damage and keep the legs fresher late in a race, which matters far more over 26.2 miles than in a five-minute lab trial, though it is worth knowing that evidence is largely Nike-funded. And a 2026 study found meaningful savings even at very slow speeds with no pace drop-off, which cuts against everything above, with the caveats that it was run by a brand against an ordinary trainer rather than a weight-matched flat. The fair conclusion is not that the shoes don't work. It is that the headline number is not yours, and the quieter benefits are smaller and less certain than the marketing implies.
So buy them if you like them, for the cushioning, the fresher legs, and the modest real edge. Just don't buy them for a 4 percent that was measured at a pace you don't run, and don't mistake them for the upgrade that fixes your race. The shoe gives a 2:30 marathoner something it cannot give a 4:25 one. The training that turns a 4:25 into a 3:50 is still where your time is hiding, and no foam will hand it to you.
The takeaways
- The famous 4 percent was measured at 14–18 km/h, sub-3-hour marathon pace, in a Nike-funded study. At that speed every runner improved.
- Run the same shoe slower and the saving fades: ~1.6 percent at 12 km/h and a statistically insignificant 0.9 percent at 10 km/h — the pace most age-groupers actually race. That study was independent and bought its shoes at retail.
- Most marathoners live at the bottom of that curve. Only ~21 percent of Boston and ~10 percent of London finishers broke 3:15; the US median is about 4:25.
- Not all super shoes are equal: in an independent seven-shoe test, the Hoka and Brooks models were no better than a plain racing flat, and individual response varies widely. Race in them before you race on them.
- They still help in quieter ways (light weight, fresher legs late), so buy them with clear eyes, not for a 4 percent that isn't yours. The training that makes you faster is the real upgrade.
Sources
- Hoogkamer, Kram et al., A Comparison of the Energetic Cost of Running in Marathon Racing Shoes, Sports Medicine 2018 (the "4%"; Nike-funded; 14–18 km/h)
- Joubert, Dominy & Burns, Running Economy at Slower Running Speeds, IJSPP 2023 (independent; 1.6% at 12 km/h, 0.9% ns at 10)
- Joubert & Jones, Seven carbon-plated racing shoes compared, Footwear Science 2022 (Hoka/Brooks ~0% vs Nike/Asics ~2.5–3%)
- That Triathlon Show with Dustin Joubert — #322 and #383
- Marathon pace distribution — average marathon times (US median ~4:25)
- Black, Jones et al., Highly Cushioned Shoes and Muscle Damage, MSSE 2022 (Nike-funded)
- Advanced Footwear at Slower Speeds, Sports Medicine Open 2026 (counter-finding; brand-affiliated; trainer control)