Thoughts on Estimating Maximal Acceleration and Max Sprinting Speed

A recent article by Buchheit & Eriksrud (2024) got me thinking about (i) maximum sprint acceleration and (ii) velocity estimation with athletes. Buchheit & Eriksrud (2024) proposed two tests to estimate “maximal locomotor function” in football: (1) 40-m straight-line sprint and (2) change-of-direction tests (m5-0-5, 10-0-5, 15-0-5) measured using the 1080 tether device (which provide instantaneous velocity-time trace) (Figure 1).

Figure 1. 40-m straight-line test (A) and 15-0-5 change-of-direction test (B). The COD setup is similar for the m5-0-5 and 10-0-5 tests. Images taken from Buchheit & Eriksrud (2024).

The study’s objective was to find reliable variables to estimate maximal locomotor function involving the abilities to accelerate, max speed sprint, and maximally decelerate, which can be easily performed in a practical setting without causing too much burden to the athletes and training process.

Although they haven’t done any factor analysis, the results of this study are pretty interesting. As a side note, I would love it if the authors started providing code and data so the others could reproduce their results or maybe analyze the data using a different approach, particularly for an open journal such as this.

I will not go into the full review of the paper (you can freely access it) but will focus on one main finding. Namely, manifested/observed peak acceleration (0.5sec interval with the highest acceleration) was HIGHER in the COD tests (particularly the m5-0-5 test) compared to the 40-m straight-line run (6.28 – 6.35 ms-2 vs. 5.70 ms-2).

The authors proposed the following explanation, which I agree with:

“Re-acceleration during the second phase of the 15-0-5 (phase1b) was slower than in the first phase and comparable to the linear sprint (Figure 6). This could be due to the difficulty of starting from a lateral position while also adjusting for the momentum generated in the first phase of the test. The reason why Max Accel was slightly higher during the 15-0-5 than during the 40-m sprint (Tables 6 and Figure 6) may be related to a pacing strategy, where players adapt their acceleration strategy to the distance/time available, with shorter distances/times requiring greater acceleration early in the run.”

Taking this together with my PhD project of estimating sprint acceleration-velocity profile (Jovanović, 2023; Jovanović et al., 2024), in which we demonstrated issues in estimating maximum acceleration, I will propose a few practical solutions (i.e., speculations that need more validation) for practitioners interested in estimating maximum acceleration and maximum velocity (with non-sprinters). I will focus on these two qualities and leave COD and deceleration qualities aside (for now).

To do so, I will split the discussion into two segments: (1) task design and (2) measurement design.

IMPORTANT CONSIDERATION

I have sent this article to Håkan Andersson for a review and critical feedback, and he mentioned that the difference in max acceleration between the straight-line sprinting and COD tasks in Buchheit & Eriksrud (2024) might be due to the tether device (i.e., 1080) having a 3kg (~5% BW) horizontal resistance in the sprint condition, and assisted resistance in the COD task. Ideally, this study should be repeated with the laser, rather than the tether device

Since Håkan mostly works with sprinters, he is not necessary recommending splitting the sprint assessment in separate tests estimating max acceleration and max speed.

Task design

How do we design tests to estimate max acceleration and max velocity? So far, we have simply performed 30-40m linear sprints, measured with either timing gates, laser, or tether devices, from which we extracted observed and/or estimated (i.e., using a mono-exponential equation) maximum acceleration and maximal velocity values.

At least for non-sprinters, there might be better approaches for estimating these two qualities. As indicated by Buchheit & Eriksrud (2024), it appears that athletes perform some type of pacing strategy. Similarly, we tend to use more isolated tests when we try to estimate different performance or physiological qualities. For example, estimating lactate threshold and VO2max usually demands two separate and specific tests. Can this also be the case when evaluating these two sprint traits?

Practically, we might separate the tests to estimate (i) max acceleration and (ii) max velocity by performing more repetitions of very short sprints (i.e. 5-10m) and flying sprints (Figure 2). In both tests, the athlete can focus on maximizing either acceleration or max velocity.