Who was measured, how many, which sport? Five common skeleton claims put to the evidence test — from the speed myth to the equipment promise.
The most honest question in skeleton is not "what is fast?" but "how do I actually know that?" — because this sport is small, its study base is thin, and a considerable part of what circulates as knowledge is a model, a federation figure or a transfer from another sport. That is not an accusation, it is a property you have to work with.
Three questions are enough to classify almost any skeleton claim: who was measured? how many? and was it even skeleton? The last is the sharpest — several of the most robust numbers in the sport come from collision sports or from mixed bobsleigh-and-skeleton squads.
Take an example of a good number: at one large sliding track, 43,200 runs were evaluated over four years, with an injury in roughly 0.5 per cent of runs [1]. Large sample, clear method — but it covers bobsleigh, luge and skeleton at one track, so it is not skeleton-isolated [1]. That is what a robust number with honest small print looks like. And that is how you check the five claims you keep meeting in this sport.
Status: practice, not measurement. These values come from federation and media reports, not from a peer-reviewed measurement series — which is why they consistently appear without a source number in this guide corpus. They are probably roughly right and still not a data basis. Use them to explain the sport to someone, never to justify a training decision.
Status: partly, and strongly track-dependent. The relationship between push time and finish time varies across tracks from roughly 0.57 to −0.14, averaging about 0.21 [2] — so on some tracks a better start pays practically nothing. Small print: this evaluation sits at performance-analysis level, not as a core journal study [2]. The counterpoint comes from simulation: once mass, start, drag and friction are optimised, the result rests solely on the ability to find the best trajectory [3].
Status: plausible, inconsistently evidenced — and not from skeleton. What is described as a predictor is isometric neck strength, not neck circumference; individual studies in rugby players found a roughly 45 per cent lower concussion rate, other reviews found strength gains without proven injury reduction, newer meta-analyses find no clear picture — and the optimal dosage is explicitly unresolved [4]. All of this data comes from collision sports [4]. For skeleton, add this: "sled head" is not a formal diagnosis, and a peer-reviewed skeleton-specific measurement series is missing [5]. Train your neck anyway — but as a sensible measure with an open effect size, not as a promise of protection.
Status: the wrong way round. The trajectory model explicitly recommends maximising the combined mass of athlete and sled within the rules [3]. And the rulebook is unambiguous: sled up to around 43 kilograms (men) or 35 (women), total weight around 115 or 92 kilograms, ballast exclusively on the sled [6]. Small print: "every kilo of body weight is a kilo less sled" is a derivation from those two sources, not a study recommendation — and weight work belongs under sports medical supervision anyway.
Status: modelled, not measured in the field. For skeleton runners a friction coefficient of around 4.6 to 4.8 thousandths was calculated, with the flattest runners producing the lowest value and the rocker influencing friction more strongly than the cross-sectional radius [7]. That is a numerical model, not a field measurement [7]. The same applies on the aerodynamic side: the finding that athlete and sled form one system comes from institute reports and a preprint, which is why no drag coefficients are quoted anywhere in this corpus [8]. Material counts — but the numbers behind it are models, not a stopwatch.
Thin evidence is a reason for more caution, not less. Where the effect size of a measure is open — as with neck training and concussion risk [4] — you must not budget it as protection: no training programme justifies additional run volume or an earlier return after a head injury.
And one sentence stands above all: the fact that robust physics exists for line and steering does not make it an instruction. How exactly shoulder and knee movements influence the rotation of the sled is still only partly understood — line and steering work belongs exclusively at the track, with qualified coaches and in controlled progression.
Attach a single line to every number you take into your training: who was measured, how many, which sport. Three items, ten seconds of effort. After a month you will notice that a surprisingly large share of your convictions stands naked in that line — and those are exactly the candidates you spend neither money nor runs on. This is not an academic exercise, it is budgeting.
By three items: who was measured, how many, and whether it was skeleton. A large, cleanly collected figure such as the evaluation of 43,200 runs with roughly 0.5 per cent injury rate [1] holds up — with the note that it covers one track and all three sliding disciplines [1]. If those items are missing, it is a story.
Because they come from federation and media reports and not from a peer-reviewed measurement series. Throughout this guide corpus they are therefore deliberately carried without a source number — as a practical illustration, not as a data basis for training decisions.
No, but it is no promise of protection. What is described as a predictor is isometric neck strength, not neck circumference; individual rugby studies showed roughly 45 per cent fewer concussions, other reviews no effect, and the optimal dosage is unresolved [4]. The data comes from collision sports [4], and for "sled head" a skeleton-specific measurement series is missing [5].
Most likely the injury rate from the track evaluation across 43,200 runs [1] — large sample, clear method, and the limitation is stated openly [1]. On the performance side the trajectory simulation is the strongest work, because it was validated against real descent times [3].
Because the sport is small and field measurements in an ice track are demanding. That is why central statements come from simulation and calculation: the runners' friction coefficient from a numerical model [7], the weighting of descent factors from a trajectory simulation [3], the aerodynamics from institute reports and a preprint without citable drag coefficients [8].