Two recent preprints shift the credit for racing-shoe speed from the carbon-fiber plate to the midsole foam, and change what’s worth weighing when you pick a pair.
- In a comparison of four shoes identical in geometry and mass, modern foam and the plate produced the exact same gain: 1.30% each.
- Foam’s advantage shows up once you stop artificially equalizing weight, because modern foam is lighter than traditional EVA: roughly 1.3% versus 0.8%.
- A second study measured two racing foams: nearly identical vertical stiffness, but shear stiffness almost double in one of them. That’s the property that handles lateral push, and no spec sheet mentions it.
- A racing shoe’s usable life comes down to foam fatigue. The carbon plate barely degrades at all.
Shoe launches spend most of their airtime on one thing: the carbon-fiber plate buried in the midsole. Two recent papers, both preprints and so not yet peer-reviewed, arrive from opposite directions and land on the same conclusion. The piece everyone talks about isn’t the one doing most of the work. That’s the foam around it.
The Part Everyone Talks About vs. the Part Doing the Work
A modern racing shoe pairs a tall midsole of light, highly elastic foam with a curved geometry that helps the foot roll forward. A stiff plate, usually carbon fiber, runs through that foam and boosts resistance to longitudinal bending. Marketing built its whole story around the plate because it’s the easiest piece to show off. The foam stays a black box that brands publish almost nothing about.
The problem: those three elements always arrive together. Buy a shoe with a plate and you’re also buying the new foam and the tall geometry, and from the outside no one can tell which of the three is driving the gain. Separating them takes an experiment that holds everything constant except one variable at a time.
What Changes When You Isolate the Variables
Foam, Plate, and Geometry in a Controlled Comparison
That’s what a University of Massachusetts Amherst team did in a preprint posted on bioRxiv. Fourteen trained runners ran on a treadmill in four versions of the same racing shoe, identical in midsole height and geometry, differing only in foam type and whether the plate was present. The three experimental versions were supplied by the manufacturer, worth keeping in mind when weighing the results. Masses were equalized with adhesive weights, since a lighter shoe improves running economy on its own.
The numbers: starting from the version with traditional EVA (ethylene-vinyl acetate) foam and no plate, adding the plate cut metabolic cost by 1.30%. Swapping in PEBA (polyether block amide), the polymer behind modern midsoles, produced the exact same gain: 1.30%. The two intermediate versions, each with only one modern element, came out indistinguishable. With both technologies together, the gain rose to 1.93%, less than the sum of the parts.
So far the two elements weigh the same. The reversal comes when you drop the artificial weight-matching: modern foam weighs about 30 grams less per midsole, while the plate adds roughly 20. Redoing the math for shoes as they’re actually sold, the authors estimate foam’s contribution at around 1.3% against roughly 0.8% for the plate. Half a percentage point of difference, produced without touching the materials, just by no longer correcting for their weight.
The authors caution that the comparison holds for those two foams and that plate specifically; stiffness and elasticity vary widely even within the same polymer family.
Shear Stiffness, the Property No One Talks About
The second paper, a preprint from a Stanford group, took apart two racing shoes and pulled samples from their midsoles, testing them in tension, compression, and shear. Shear stiffness measures how much foam resists a sideways push: picture pressing your palm into the midsole and sliding it sideways, instead of pushing straight down.
The gap between the two measurements is the interesting part. In compression, the two foams are nearly identical, 268 and 299 kilopascals: under body weight, they behave the same way. In shear, one comes in at almost double the other, 219 versus 117 kilopascals. Two shoes that support you identically in a straight line can give way very differently to the side, and no spec sheet says so.
Why It Matters on Turns and Uneven Ground
There’s a physical reason this number stays low across every racing foam: lightening a cellular material drops its shear resistance faster than it drops compression resistance. It shows up in the energy return measured at Stanford too, around 85-90% in compression and only 74-76% in shear. Energy spent deforming the foam sideways comes back a lot less.
In a straight line, this barely matters. It matters on a tight turn, on broken ground, when your foot lands on a sudden change in surface: the authors point to exactly these conditions for reduced stability and a higher injury risk. If your race has a lot of direction changes or uneven footing, lateral hold stops being a footnote, and it’s worth pairing your shoes with an ankle trained to handle lateral give.
How Long a Racing Shoe Actually Lasts
The Stanford preprint adds a point on shoe lifespan. Under repeated loading, the carbon plate barely degrades structurally, while the foam undergoes viscoelastic fatigue and gradually loses elasticity and energy return. The performance benefit, the authors report citing existing literature, tends to fade within 150-300 miles of use, roughly 240-480 km. A shoe stops performing like new well before it looks worn out.