Acceleration speed profile for football players
How to profile every player’s acceleration and top speed from the GPS data you already collect, without a single sprint test, and how to use it without being misled by the data.
Every coach in football wants faster players. Far fewer know exactly how fast each of their players can accelerate, because finding out has traditionally meant a dedicated sprint session, timing gates or radar, and a maximal effort from every player on the same day. In a congested fixture schedule, that test often never happens.
A 2021 proof-of-concept study led by Jean-Benoît Morin offers a different route. Instead of testing players, it builds each player’s acceleration-speed (AS) profile from the GPS data already collected in normal training. This article expands on our carousel: what the profile is, how it is built, what the research found, and how to use it without being misled by poor data.
Why acceleration decides matches
Football is an intermittent sport, but its decisive moments are often fast and short. In an analysis of goals in the German Bundesliga, Faude and colleagues (2012) found that a straight sprint was the most frequent action in goal situations, performed by the scoring or the assisting player in 45% of the goals studied. Most of those sprints are not 40 m dashes from a standstill. They start from a jog, a turn or a run, and they are won in the first few metres.
That is exactly why the traditional sprint test is an imperfect fit for football. A straight-line sprint from a static start tells you a lot about a player’s mechanical capacities, but it:
- Needs a dedicated session, with timing gates, radar or a laser, and a set-up that takes time out of training.
- Needs a true maximal effort from every player, on the day, when some are fatigued, carrying a knock or not motivated to sprint flat out.
- Isn’t football-specific: players rarely accelerate from a standing start in a straight line in a match.
The in-situ approach flips the problem around. Players already accelerate hard, at every speed, many times a week. If the GPS units are recording those efforts anyway, the information is already there.
What the acceleration-speed profile actually is
The AS profile describes the maximal acceleration a player can produce at each running speed. At low speed a player can accelerate hard; as they get faster, their capacity to keep accelerating falls, until it reaches zero at their maximal speed.
There is a neat reason that relationship is a straight line. During a maximal sprint, a player’s speed rises along a well-described exponential curve (Samozino et al., 2016). If speed follows that curve, acceleration falls in direct proportion to the speed already reached. Plot the best accelerations against speed and you get a line with two anchor points:
- A0: where the line meets the acceleration axis, the theoretical maximal acceleration at zero speed (m/s²).
- S0: where the line meets the speed axis, the theoretical maximal running speed (m/s).
If those terms sound familiar, it’s because they mirror the sprint force-velocity profile. That method, from Samozino and colleagues (2016), uses the same exponential speed curve plus body mass and air resistance to estimate F0 (maximal horizontal force) and V0 (maximal velocity). The two methods describe the same underlying qualities in different units:
| Sprint force-velocity profile | In-situ acceleration-speed profile | |
|---|---|---|
| Data source | One or two maximal sprints, usually 30–40 m, from a standing start | GPS data from normal training and matches |
| Force / acceleration capacity | F0, maximal horizontal force (N/kg) | A0, maximal acceleration (m/s²) |
| Speed capacity | V0, theoretical maximal velocity (m/s) | S0, theoretical maximal speed (m/s) |
| Context | Straight line, static start | Real football: rolling starts, after changes of direction |
| Testing burden | A dedicated session and full effort on the day | None beyond wearing the GPS unit |
How the profile is built from GPS data
Morin et al. (2021) designed a simple procedure that any sports scientist with reliable GPS data can reproduce. The key idea is to throw away almost all of the data and keep only the efforts that represent the player’s limit at each speed.
Starting the analysis at 3 m/s removes the slow, low-intent movement that dominates any training session. Splitting the speed range into 0.2 m/s bands and keeping only the two highest accelerations in each band leaves around 50 points per player, each one a near-maximal effort at that speed. A straight line through those points, refitted once the outliers are removed, gives the profile.
What the research found
Morin and colleagues applied the method to 16 professional players at an Italian top-flight club, using GPS data collected passively during normal training. The results were strikingly consistent:
| Finding | Result | What it means |
|---|---|---|
| Linearity | R² above 0.98 for every player | The straight-line model fits real football data almost perfectly |
| Maximal speed (S0) | 9.2–9.5 m/s | In line with sprint values reported in professional footballers |
| Maximal acceleration (A0) | 7.2–7.7 m/s² | Comparable to the force capacities seen in sprint force-velocity testing |
| Reliability | Random error under 8% between two 2-week blocks | Stable enough to track changes across a season |
| Individual differences | Large, even between players doing the same training | A squad average hides very different profiles |
The authors’ conclusion was that A0 and S0 describe the same qualities as F0 and V0, but measured in a genuine football context, with no extra testing session needed.
Get the data right, or the profile is wrong
The method is only as good as the data behind it. A pilot study supervised by Morin (Komino, 2022) compared GPS-derived profiles in 10 players against classic 30 m sprint force-velocity profiles, and the results show both the promise and the limits:
- Top speed agreed well. S0 and V0 were strongly correlated (r = 0.95), even when the GPS profile came from a single sprint.
- Maximal acceleration was less precise. 45 minutes of 7v7 did not provide enough data to estimate A0 well.
- Volume matters. Around 50 maximal efforts, including hard accelerations at high speed, are needed for a robust profile.
The reason is visible in Figure 2. S0 is pinned down by the points near top speed, which most players reach at some point in a week. A0 is extrapolated from the steepest part of the line, so if a player never accelerates flat out from moderate speeds in the sessions you capture, the line tilts and A0 is underestimated.
A practical data checklist
| Requirement | Target | Why |
|---|---|---|
| Collection window | At least 5 full sessions (about 2 weeks) | Covers each player’s full speed range |
| Maximal efforts | ~50 near-maximal points, including near top speed | Short small-sided games alone are not enough |
| GPS quality | Reliable, high-frequency units (18 Hz in the original study) | Acceleration is very sensitive to noise |
| Consistency | Same unit for each player, same processing every time | Changes should reflect the player, not the device |
| Method | Track A0/S0 or F0/V0, not a mix | The values are similar in meaning but not interchangeable |
How to read a player’s profile
Two players can share the same top speed and still be very different footballers. The line shows where.
Because the profile is a straight line, you can read off a player’s maximal acceleration at any speed with one simple equation: acceleration = A0 × (1 − speed ÷ S0). Using the example profile from Morin et al. (2021):
| Running speed | Situation in a match | Maximal acceleration available |
|---|---|---|
| 0 m/s | Standing start (theoretical) | 7.9 m/s² (A0) |
| 3 m/s | Jogging, then reacting to a ball over the top | 5.3 m/s² |
| 5 m/s | Running, then breaking into a sprint | 3.6 m/s² |
| 7 m/s | Already sprinting, trying to find another gear | 1.9 m/s² |
| 9.19 m/s | Top speed | 0 m/s² (S0) |
Two other numbers are worth knowing. The slope of the line (A0 ÷ S0) describes how quickly a player’s acceleration capacity fades as they speed up. And A0 × S0 ÷ 4 gives an index of maximal acceleration power per kilogram, the AS equivalent of the maximal power value in force-velocity profiling. Both come straight from the same two numbers, with no extra data.
Using the profile in training
Morin and Samozino (2016) argued that sprint profiles are most useful when they individualise training rather than simply rank players. The same logic applies to the in-situ profile. Read it alongside the coach’s eye and other data, then decide:
| What the profile shows | What it may suggest | Training ideas to consider |
|---|---|---|
| Lower A0, normal S0 | Acceleration capacity is the limiter | Heavy resisted sprints, short accelerations, strength and power work |
| Normal A0, lower S0 | Top-speed capacity is the limiter | Flying sprints, longer maximal efforts, high-speed exposure |
| Both low | Overall sprint capacity needs work | A blended block, with progressive maximal sprint exposure |
| A0 or S0 drops in-season | Possible fatigue, detraining or low exposure | Check load and sprint exposure; reinstate high-speed work |
| After injury | Profile compared with the pre-injury baseline | Use the gap to guide return-to-play progressions |
Because the profile updates with every block of data, it is also a monitoring tool. Re-profiling through the season shows whether players are keeping their speed qualities when the focus shifts to fixtures, and it gives an objective marker for players returning from injury, without asking them to complete a maximal test before they are ready.
Take-home messages
- Collect enough data: at least 5 full training sessions (about 2 weeks) to cover each player’s full speed range.
- Capture true max efforts: profiles need hard accelerations at every speed, including near top speed. 45 minutes of 7v7 was not enough.
- Prioritise data quality: use reliable, high-frequency GPS and the same unit for each player.
- Don’t mix methods: A0/S0 and F0/V0 measure similar qualities, but the numbers aren’t interchangeable. Track one method over time.
- Re-profile through the season: use changes to individualise training, monitor players and guide return-to-play.
Check the Dublin dates, or ask EPI whether your existing qualification counts towards Phase 2 entry.
References
- Faude, O., Koch, T., & Meyer, T. (2012). Straight sprinting is the most frequent action in goal situations in professional football. Journal of Sports Sciences, 30(7), 625–631.
- Komino (2022). Unpublished pilot study comparing GPS acceleration-speed profiles with 30 m sprint force-velocity profiles, supervised by J.-B. Morin. Université Jean Monnet, Saint-Étienne.
- Morin, J.-B., Le Mat, Y., Osgnach, C., Barnabò, A., Pilati, A., Samozino, P., & di Prampero, P. E. (2021). Individual acceleration-speed profile in-situ: A proof of concept in professional football players. Journal of Biomechanics, 123, 110524.
- Morin, J.-B., & Samozino, P. (2016). Interpreting power-force-velocity profiles for individualized and specific training. International Journal of Sports Physiology and Performance, 11(2), 267–272.
- Samozino, P., Rabita, G., Dorel, S., Slawinski, J., Peyrot, N., Saez de Villarreal, E., & Morin, J.-B. (2016). A simple method for measuring power, force, velocity properties, and mechanical effectiveness in sprint running. Scandinavian Journal of Medicine & Science in Sports, 26(6), 648–658.
