Humans Running at Stadiums and Beaches and the Accuracy of Speed Estimations from Fossil Trackways

1. What they did

Researchers often estimate how fast an animal was moving from its fossil trackway, using an equation proposed by Alexander in 1976. It says that speed rises with stride length in a predictable way once you allow for the height of the animal’s hips (hind limb height). The idea has been widely used on dinosaur trackways, but some studies have claimed the estimates could be off by a factor of 1.5 to 2.5. Humans were part of the data Alexander used, so testing the equation more thoroughly on people tests the method itself.

The authors tested it in three ways:

• Elite athletes. They gathered published stride length and speed measurements from World Championship and World Cup races (men’s and women’s 100 m and 200 m, and individual runners in the 4 × 400 m relay). Race starts and running on curves were left out.
• A beach experiment. They repeated Alexander’s own beach experiment with a much bigger sample: 14 people (seven women, seven men), mostly vertebrate palaeontology students at the Universidad Complutense de Madrid. On firm sand exposed at low tide at La Griega beach in Asturias, Spain, each person walked twice, walked fast once and ran twice, and the authors related stride length, hip height and speed (hip heights in the group ranged from 0.86 to 1.02 m).
• A fossil case study. They used their results to estimate the speeds of the people who left upper Pleistocene trackways at Willandra Lakes in southeastern Australia, 19,000 to 23,000 years ago, and compared them with earlier estimates by Webb (2007).

2. What they found

Athletes

• For a given stride length, 400 m runners were slower than 100 m sprinters, with 200 m runners in between. So the same stride length can go with different speeds, depending on stride frequency, which cannot be measured from a trackway.
• Athletes’ speeds were similar to or lower than the range Alexander’s equation predicts for people.

The beach experiment

• Walking data fell close to Alexander’s prediction; running data were more scattered.
• Walking changed to running when stride length reached about 2.06 times hip height, in line with earlier work.
• The equation fitted to the beach data was almost identical to Alexander’s.
• On average, observed speed was very close to predicted speed (a ratio of 1.006, with a standard deviation of 0.116). The most extreme cases were 38% faster and 35% slower than predicted.

All the data together

• Athletes and beach walkers and runners together fitted one simple curve that uses stride length alone (no hip height needed), with a very good fit (R² = 0.969). For people, this curve is almost the same as Alexander’s equation with a hip height of about 1 m.
• Predictions from this curve were usually within 10 to 15% of the real speed. Each type of running had its own pattern, with observed speeds generally lower than predicted at the longest strides.

Willandra Lakes

• Speeds from the new equation were generally lower than Webb’s (2007) estimates, which were made with a different method based on foot length, stride length and sex.
• The fastest trackway (T8), which Webb put at 10.3 m per second (sprinter speed), came out at about 7.2 m per second (range 6.2 to 8.2). This person was probably still sprinting, but at a speed more believable for someone who was not a trained athlete.

3. What they concluded

• Alexander’s method for estimating speed from stride length is generally very accurate for people, usually within 10 to 15% of the real speed. This is more accurate than earlier work suggested.
• Because people were part of the data behind the method, this strengthens the case that reliable speeds can be worked out from fossil trackways in general, although estimating the trackmaker’s hip height remains uncertain.
• For human trackways, very reliable speeds can be estimated directly from stride length, without knowing the person’s hip height.
• The method of Charteris et al. (1981), used by Webb (2007), overestimates speed in most of the Willandra Lakes trackways.
• Future work that also considers body weight could help test a proposal to refine Alexander’s equation using body mass.

4. How tracking was or was not used

The study is about reading speed from tracks. In the beach experiment, people walked and ran on firm, low-tide sand, as in Alexander’s original experiment, and their stride lengths were related to their speeds; the paper does not describe how strides or speeds were measured. In the case study, stride lengths measured from fossil human footprint trails (taken from Webb 2007) were turned into speeds. No tracks and signs identification or trailing was involved.

6. Other technologies, skills or innovations

• A simple equation that gives a person’s speed from stride length alone: speed (m per second) = 0.794 × stride length (m) to the power of 1.67. A stride here is the distance between successive prints of the same foot (two steps).
• The authors note that the same exponent (1.67) appears in all their equations, and suggest it is a basic property of two-legged movement, whatever the species, surface or training.