Serve Accuracy Fell at 80% Workload but Held at 60% in 15-Player Tennis Trial

A Turkish laboratory study set each player a personal workload, then measured serve accuracy at 60 and 80 percent of it. Serve accuracy survived the moderate session and fell away at the hard one.

Serve accuracy in male university-level tennis players dropped significantly only once workload reached 80 percent of each player individual maximum, according to a repeated-measures trial published on 18 September 2026 in BMC Sports Science, Medicine and Rehabilitation.

Fifteen trained male players took part. Their mean age was 22.33 years, with a standard deviation of 1.44, and they carried 2.6 years of tennis experience on average, standard deviation 0.5. Mean height was 1.78 metres, mean body mass 75.40 kilograms, and mean body mass index 23.75. The work was approved by the ethics committee at Bingol University, which also provided the testing facilities.

Rather than impose one fixed session on everyone, the researchers first ran the HIT-TURN test to establish each player maximal performance. Moderate intensity was then set at 60 percent of that individual value and high intensity at 80 percent. Players were assessed at baseline, at the moderate workload and at the high workload, each on a separate day, with 72 hours of recovery between visits.

Three outcomes were tracked at every visit: rating of perceived exertion, blood lactate concentration, and Serve Hit Points, the study measure of serve accuracy. Analysis used one-way repeated-measures ANOVA with Bonferroni-adjusted post hoc comparisons.

Fatigue Was Built to a Personal Scale, Not Guessed

Most fatigue studies in racket sport either run players to exhaustion or impose an identical drill on a squad with very different fitness. This trial did neither. By anchoring both workloads to an individual maximum from the HIT-TURN test, it aimed to put every player under comparable relative strain rather than comparable absolute work.

That matters for how the serve accuracy result should be read. A 60 percent session is not a fixed number of minutes or a fixed heart rate here. It is 60 percent of what that particular player could produce, which makes the comparison between conditions cleaner than the comparison between players. That is a genuine design strength, and it also narrows what these numbers can say about any single athlete.

What Happened to Serve Accuracy at 80 Percent

The Serve Hit Points measure differed significantly across the three conditions, at p less than 0.001. The post hoc pattern is the substance of the paper. Serve accuracy at high intensity was significantly lower than at baseline and significantly lower than at moderate intensity, both comparisons at p less than 0.001.

The contrast that did not reach significance is equally informative. Baseline against moderate intensity showed no significant difference in serve accuracy, which means a session at 60 percent of individual maximum left the skill measure intact in this group.

So the degradation was not gradual across the range tested. It appeared at the hard condition and not at the moderate one, at least at the two workloads the design sampled.

The Lactate and Effort Numbers Behind the Drop

Blood lactate concentration and rating of perceived exertion both increased significantly across the three conditions, each at p less than 0.001. That is the internal check the design needed: the workloads were not just labelled moderate and high, they registered as progressively harder in both a physiological and a subjective marker.

Without that confirmation, a null result at 60 percent could simply mean the moderate session was too easy to count as a session at all. With it, the more defensible reading is that players were genuinely loaded at 60 percent and still served as accurately as they did when fresh. Work on lactate accumulation rate and threshold power in 182 cyclists and triathletes shows how differently individual athletes sit on that curve.

Why the Missing Effect Sizes Matter

The published record for this trial reports significance levels but no group means, mean differences, confidence intervals or effect sizes for the serve accuracy outcome. The page also carries a note that it is an early-share version subject to further edits before the final version of record.

A p value below 0.001 says a difference is unlikely to be chance in this sample. It does not say how many serves out of a set the players actually missed. Readers who want to know whether the serve accuracy loss is a handful of points or a collapse cannot get that from what is currently reported, and should treat the magnitude as unknown rather than large.

What 15 Players Can and Cannot Settle

Fifteen participants is a small sample even for a within-subject design, and every one of them was a trained university-level man with a mean of 2.6 years of experience. Nothing here extends to women, to juniors, to recreational players or to touring professionals, whose serve mechanics and conditioning differ substantially.

The protocol is also a laboratory one. Players hit under controlled conditions after a prescribed workload, not in the middle of a match with rallies, changeovers, scoreboard pressure and tactical choices. The authors are explicit that more ecologically valid match-play protocols are the next step. Comparable questions about measuring fatigue consistently surfaced in a review of countermovement jump fatigue testing across 117 studies, and in work on neuromuscular fatigue and rowing technique in 19 elite rowers.

What the Serve Accuracy Finding Means for Practice

The useful reading is about ordering, not alarm. If this pattern holds, technical serve work has a better chance of landing when it sits before the hardest conditioning of the session rather than after it, and serve accuracy measured at the end of a brutal block is measuring fatigue as much as technique.

• Place precision serving early in a session, or on a day separated from the heaviest conditioning work.
• Treat serve accuracy numbers collected after hard work as a fatigue reading, not a technical baseline.
• Recognise that a moderate session did not degrade serve accuracy here, so skill practice after easier work looks reasonable.
• Do not read a single laboratory protocol as a prescription for match preparation.

This is general information rather than individualised coaching or medical advice, and it describes a 15-man laboratory study rather than a consensus. Players managing an injury or a medical condition should work from their own clinician and coach. For a sense of how modest performance margins behave in small trials, see the FLM report on caffeine and 30-metre sprint times in female athletes.