A physics engine, and a verification tool enforced on top of it. PhysWall inverts a closed, published, non-linear physical law at a single measurement point — and refuses when the inverse is not unique. Seven laws, one engine, and the same refusal in all of them.
Fifteen metres, twenty handholds, eleven footholds, and a route the IFSC has specified to the millimetre since 2007 — hole spacing 125 mm to a 1 mm tolerance, panels 375 mm to 2 mm, identical at every sanctioned wall in the world.
What the stopwatch settles is average power against gravity. The IFSC wall is 15 metres everywhere, so a time and a body mass fix the work done and the rate it was done at.
Every other world record is a measurement of an athlete and a venue at once. Tracks differ, pools differ, courses differ. The speed climbing wall does not: same height, same five-degree overhang, same position for every one of the thirty-one holds, at every competition in the world.
A time on this wall is a measurement of a person. Run it backwards through the work done against gravity and you get power per kilogram — the figure cycling and running already use, so the comparison needs no translation.
men's record 4.64 s 31.7 W/kg women's record 6.10 s 24.1 W/kg cycling sprint, 5 s ~24 W/kg stair-running record ~25 W/kg
The men's figure is above any published sprint number. The obvious explanation is that a climber uses arms and legs together where a cyclist uses legs alone — and it is offered as an explanation rather than a measurement, because we have not found a measurement of it — which is not the same as there being none.
100 m sprint 9.5% marathon 8.8% 100 m freestyle 11.4% speed climbing 25% to 35%
That range is not imprecision. It is what the record progression actually does:
2017 5.48 / 7.38 34.7% 2021 5.20 / 6.84 31.5% 2023 4.98 / 6.24 25.3% 2025 4.64 / 6.10 31.5%
Four points, spanning 9.4 percentage points, on a route that has not changed. Whatever produces the gap is not a fixed property of the wall — a fixed wall against a moving gap means the gap is coming from who is climbing, not only from what they are climbing.
⚠ And four points is four points. A record is the extreme tail of a distribution, not its middle, and two records in one year can be set by two athletes who happen to be having a good season. The number to trust would come from full field results, not from the top of them.
Height is part of it and cannot be most of it. Elite women in this discipline are roughly 7% shorter than the men, not a third weaker. The holds, though, are bolted where they are bolted: a 155 cm climber and a 190 cm climber make the same twenty moves to the same holds, and one of them extends to four-fifths of their reach on every move while the other extends to two-thirds.
Which share of that gap is physiology and which is a wall built to one set of proportions.
Part of it already has an answer. Anthropometric studies of elite male climbers report heights by discipline:
lead 173.3 ± 6.3 cm boulder 176.1 ± 7.6 cm speed 177.6 ± 6.4 cm
Speed climbers are about a centimetre and a half taller than boulderers and four centimetres taller than lead climbers — in the direction you would guess. The competition records carry arm span alongside height for every athlete back to 1991, which would let anyone run the comparison properly; it has not been run here.
And two things stop that settling it. Speed climbers are also 3.5 kg heavier with broader shoulders, so the sorting may be on build rather than reach. And within speed climbing, height separates nothing: what distinguishes international athletes from national ones is lower fat mass, greater lean mass and greater skeletal breadth across the shoulders, upper arm and thigh. The same study reports that traits which matter in other climbing disciplines are not relevant in speed climbing at all.
So the 1.5 cm is a fact about who takes up the discipline, and not about who wins it.
Those figures are all men, and for women the literature disagrees with itself. One review concludes stature has less influence on female speed climbers. A study of five elite women found height correlating with race time at R²=0.89. A World Cup survey found female finalists significantly shorter and lighter than the semi-finalists they beat, while men showed no difference at all.
Five athletes, seven athletes, three conclusions. That is a sample size problem rather than a contested result — only about a third of participants in climbing research are women.
The other half is not published: move-by-move split times, which would separate a climber who is slower on every move from one who loses time on the three moves that are furthest for them. The video exists at every World Cup. The timing does not leave the finish button.
PhysWall was developed and architected by Gadi Zion.
Built on PhysWall — the same engine reads antenna bandwidth, conductor loss, bit erasure and heat limits. It answers what the measurement implies, and refuses when the measurement cannot say.⚠ Check this instead of believing it. Every number here reproduces from a source that is named, and the claims that turned out wrong are still printed next to what replaced them. The same engine runs all of these — it asks how much a measurement allows you to conclude, and refuses the same way in every field. The same engine runs all of these — it asks how much a measurement allows you to conclude, and refuses the same way in every field. How to check each one →