In one study, a 4v1 rondo on a 6×6 m pitch, a square six metres per side, nearly doubled the middle defender’s acceleration events over 30 seconds compared with a 6v2 on the same pitch.

Measure magnitude to know how hard each acceleration is, count events for repetition volume, and record starting speed for match-speed exposure. Pick whether the session needs harder accelerations, more accelerations, or accelerations that start from higher running speeds before choosing player numbers and pitch size. A 4v1 can make each acceleration harder without producing the most events or any actions that begin from a fast run.

A 4v1 gives the middle defender repeated chances to leave a near-stationary position, close a passing lane and stop again. Players need a larger exercise to accelerate or brake after a fast run.

Pick the output before the format

The 4v1 raised how hard the defender accelerated. How often and from what speed depended on player numbers and pitch size.

A coach who records only event counts knows whether a drill produced more starts and stops but cannot tell whether each one was harder. A coach who records magnitude without starting speed can identify a hard acceleration but cannot tell whether the player was already running when it began.

Magnitude and frequency separate

Silva and colleagues tracked 20 elite adult male players over ten weeks with 10 Hz GNSS (satellite) trackers in a 2026 rondo study. The researchers compared eight-minute 4v1, 4v2 and 5v2 drills and characterised every acceleration above 3 m·s⁻² by magnitude, duration, starting speed, ending speed and distance. ¹

The 4v1 produced greater acceleration magnitudes and higher ending speeds than both larger formats. Accelerations also lasted longer and covered more distance than in 4v2, while the 4v1 produced greater deceleration magnitudes than the alternatives.

Action frequency increased with player numbers and pitch size, even though the strongest individual accelerations appeared in 4v1. All formats produced accelerations beginning and decelerations ending at near-stationary speeds below 4 km·h⁻¹.

In 4v1, the defender left a low speed, accelerated harder, and reached a higher ending speed over a short distance. The study did not show any long-term acceleration capacity. Its abstract gives no per-format event counts, so no universal repetition target can be set.

Across 42 football-training studies, reducing player numbers generally increased acceleration and deceleration demands. Those higher demands could mean more events or harder events; the review did not separate the two. Six GNSS brands and varying event thresholds prevented direct numerical comparison between those studies. ²

The middle defender’s count profile

Baydemir and colleagues analysed only the players defending in the middle in a repeated-measures study. Twenty elite male youth players completed 4v1 and 6v2 bouts on natural grass, lasting 15 or 30 seconds with a work-to-rest ratio of 1 to 5. The study did not vary the order of the conditions across players, so fatigue or order effects cannot be ruled out. ³

The table compiles the 30-second averages for every pitch size. Zone 1 covers events beginning in the 0–6.9 km·h⁻¹ speed band and Zone 2 covers 7–12.9 km·h⁻¹. These zones classify the player’s speed at the moment a software-detected event begins, not the acceleration magnitude itself.

At 6×6 m over 30 seconds, 4v1 produced 3.70 Zone 1 acceleration events against 1.90 in 6v2. The difference of 1.80 events was significant, with an adjusted p value below 0.001 and a Cohen’s dz of 2.24, a very large effect. The corresponding Zone 1 deceleration gap was also 1.80 events.

The 4v1 counts generally fell as the area expanded from 5×5 to 8×8 m. The 6v2 counts generally rose as the area expanded from 6×6 to 10×10 m.

One defender in a tight area has to keep restarting as the ball moves around a short perimeter. Two defenders can share nearby coverage when the area is small, while a wider area stretches the pair and asks each player to travel farther as the ball circulates.

The count study used proprietary software whose vendor did not disclose the detection threshold or minimum duration, so its counts cannot be merged numerically with Silva and colleagues’ results for accelerations above 3 m·s⁻². One study measured acceleration magnitude while the other counted events. Their numbers cannot be combined across formats and bout lengths.

Match speed remains missing

A professional under-23 monitoring study compared rondos, small-sided games, technical work, compensation drills (running drills used to top up the physical load small games miss) and matches across four training weeks. Matches produced more high-intensity accelerations than rondos with an effect size of 1.47, a large effect, and more high-intensity decelerations with an effect size of 0.53, a moderate effect. ⁴

Accelerations beginning above 25 km·h⁻¹ appeared only in compensation drills and matches. Rondos and small-sided games produced more actions from low starting speeds because their limited space prevented players from reaching the running speeds available in competition.

A separate study of amateur squads tracked 19 training weeks. Across those weeks, three weekly sessions produced 1.6 to 2.1 times as many high-intensity accelerations as a match but only a third to a half of a match’s sprint distance. ⁵

A tight 4v1 delivers repeated short accelerations from near standing but leaves the player without any efforts that begin from a fast run.

Keep the measurement stable

The count study exported events from software whose detection threshold could not be independently checked. Clubs should keep the same device model, firmware, software, threshold and minimum effort duration when comparing rondo formats, and the same player should wear the same unit whenever possible.

A Catapult Vector S8 validation found an overall acceleration error of 0.29 ± 0.14 m·s⁻² against motion capture and excellent between-device agreement. Error still increased at the highest acceleration values, and some of the authors worked for Catapult, which also helped fund the study. ⁶

An official-match comparison of GNSS and optical tracking found excellent agreement for total distance but weaker agreement for high-speed and sprint counts, with agreement scores (ICC, where 1.0 is perfect) of 0.659 and 0.640. Both systems tracked the same match, yet they did not return interchangeable event totals. ⁷

The STATSports Apex validation also warned that validity from one model cannot automatically be extended to another. A club changing hardware or software should establish a new baseline before treating a count difference as a change in the drill or player. ⁸

Run a controlled club comparison

Start with 30-second 4v1 bouts at 5×5 and 6×6 m. Keep the middle defender, passers, touch rule, ball supply, surface, recovery and device unchanged, then reverse the order in a later session so fatigue does not always favour one condition.

Compare session averages rather than one bout. Record event count, acceleration magnitude, ending speed and starting-speed band separately. A rise in event count means the player started and stopped more often. Greater magnitude and higher ending speed confirm that those actions were also harder.

A successful short-burst block can still leave the starting-speed bands concentrated below 5 km·h⁻¹. The remaining session then needs an exercise with enough distance for players to accelerate or brake after entering above 20 km·h⁻¹.

The magnitude findings come from elite adult males across all drill positions. The event-count findings come from elite male youth players defending in the middle, in a study that always ran the conditions in the same order. The evidence does not establish long-term adaptation, injury prevention, improved decision-making, or one universal pitch size.

After the rondo, check the starting speed of the defender’s hardest actions. If every hard action begins near standing, add a drill in which the defender accelerates and brakes after a fast run.