Review Reports - Understanding Gameplay Acceleration Ability, Using Static Start Assessments: Have We Got It Right?

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This study aimed to examine whether sprinting initiated from static and pickup starts share similar motor qualities. The topic is interesting; however, several aspects require improvement.

Please see other specific comments…

INTRODUCTION

P1L35 - “In team sports, acceleration is often performed from a rolling start, termed a ‘pickup’.” – A reference is needed…

METHODS

The inclusion and exclusion criteria are not clearly described. What was the rationale for selecting these participants? Why were athletes from these specific sports chosen? Were previous or present injuries considered in the selection process?

P3L112 - “The rest between trials was a minimum of 5 minutes.” – Why 5 minutes? References?

P4L143 - “An a priori sample size estimate was conducted to determine the number of participants required for the correlational analysis. Using G*Power, a conservative estimate was derived using the correlation-based model with an anticipated correlation of 0.70, an alpha level of 0.05, and a power of 0.80. This approach yielded a minimum requirement of 14 participants to detect a statistically significant correlation.” – I believe this should be moved to the ‘Subjects’ section.

DISCUSSION

P6L174 - “Of primary interest was determining whether static start and pickup acceleration were similar motor qualities and therefore could be assessed and trained in the same manner. Primary findings were: 1) there was a great deal of unexplained variance between static and pickup acceleration amax R² = 11.6-39.6%; 2) the shared variance between pickup acceleration entry velocities was 16.8%; and, 3) no systematic visual trends were observed between static start and pickup acceleration ranking of athletes.” – The discussion should not be a replication of the results...

Limitations? The fact that participants from different sports were selected may lead to biases in the results, as well as the phase of the season during which the data were collected.

Future research?

Author Response

Comment 1: P1L35 - “In team sports, acceleration is often performed from a rolling start, termed a ‘pickup’.” – A reference is needed…

Response: Thank you for your feedback. I have added a reference describing the prevalence of pickup acceleration, and clarified that 'pickup' acceleration is a coaching term. The amended sentence is highlighted in yellow on the document.

Comment 2: The inclusion and exclusion criteria are not clearly described. What was the rationale for selecting these participants? Why were athletes from these specific sports chosen? Were previous or present injuries considered in the selection process?

Response: Thank you for your feedback. Thank you for your feedback. The text has been amended to read " The criteria for inclusion were: 1) male athletes aged between 16 and 35 years of age; 2) engaged in competitive sporting activities; 3) at least one year of structured sports performance training; and 4) injury-free for at least 12 months prior to the study.”

Comment 3: P3L112 - “The rest between trials was a minimum of 5 minutes.” – Why 5 minutes? References?

Response: Thank you for your feedback. The text has been amended to read "Haugen et al. [23] suggested rest times ranging from 2 minutes (for 20 m accelerations) to 7 minutes (for elite-level 40 m sprints), so a rest interval of at least 5 minutes was chosen to allow for full recovery." The passage has been highlighted in yellow in the text.

Comment 4: P4L143 - “An a priori sample size estimate was conducted to determine the number of participants required for the correlational analysis. Using G*Power, a conservative estimate was derived using the correlation-based model with an anticipated correlation of 0.70, an alpha level of 0.05, and a power of 0.80. This approach yielded a minimum requirement of 14 participants to detect a statistically significant correlation.” – I believe this should be moved to the ‘Subjects’ section.

Response: Thank you for your feedback. The passage has been moved as asked.

Comment 5: P6L174 - “Of primary interest was determining whether static start and pickup acceleration were similar motor qualities and therefore could be assessed and trained in the same manner. Primary findings were: 1) there was a great deal of unexplained variance between static and pickup acceleration amax R² = 11.6-39.6%; 2) the shared variance between pickup acceleration entry velocities was 16.8%; and, 3) no systematic visual trends were observed between static start and pickup acceleration ranking of athletes.” – The discussion should not be a replication of the results...

Response: Thank you for your feedback. This section has been removed from the text.

Comment 6: Limitations? The fact that participants from different sports were selected may lead to biases in the results, as well as the phase of the season during which the data were collected.

Response: Thank you for your feedback, the text has been amended to read

Thank you for your feedback, Yes, potentially that could be a limitation; however, if we were to only choose one sport, the results would only be generalizable to one sport and therefore a major limitation. Having a random selection of sports enables greater generalizability of the results. Next steps would be to identify the sports in which pickup acceleration is an essential component of athleticism and the mechanical determinants of pickup acceleration, which are activity- and sport-specific. We have mentioned that in future research.

The text has been amended to read Several limitations to this study must be noted. First, the athletes tested were from a variety of sports at different points in their competitive seasons at the time of data collection, potentially influencing the results. However, it needs to be noted that having a random selection of sports/participants enables greater generalizability of results.

Comment 7: Future research?

Response: Thank you for your feedback. The text has been amended to read " In the future researchers should determine the mechanical differences between entry velocities, the technical skills that determine proficiency, and the physical factors that influence performance differences across different activities/sports.”

Reviewer 2 Report

Comments and Suggestions for Authors

Thank you for the chance to review your manuscript outlining the study assessing the relationship between static and pick up accelerations in team sport athletes (males). This is a straightforward and valuable investigation that the authors have undertaken; it has a broad range of impact on the testing of team sport athletes.

Line 32 consider adding, given an amount of time, or per unit of time, etc. vs. just 'over time' (this is minor, but it will read smoother and more accurately)

Line 50 focus of this investigation vs. article

Introduction is solid; touches on related literature from gait/rehab as well as team sports; points posed lead clearly to the purpose and hypothesis.

Line 118 What is meant by randomization being done via an Excel spreadsheet? Please be more specific with the reader.

Did participants have any familiarization period? Any sense of the number of trials that were repeated?

Line 135 Provide justification for the 0.5 Hz selection; concern a cut-off of 0.5 Hz would attenuate signal velocity and rapid accel changes

Lines 150-151, 154 italicize r and p, and R²

Line 156 italicize a_max as it appears in line 136

Table 1 italicize the p values

References - some are the journal style while others are not; run a check

Author Response

Comment 1: Line 32 consider adding, given an amount of time, or per unit of time, etc. vs. just 'over time' (this is minor, but it will read smoother and more accurately)

Response: Thank you for your feedback, the text has been amended to read 'per unit time'.

Comment 2: Line 50 focus of this investigation vs. article

Response: Thank you for your feedback, the text has been amended accordingly.

Comment 3: Line 118 What is meant by randomization being done via an Excel spreadsheet? Please be more specific with the reader.

Response: Thank you for your feedback. The text has been amended to read "The trial entry velocity order was allocated randomly using an Excel spreadsheet."

Comment 4: Did participants have any familiarization period? Any sense of the number of trials that were repeated?

Response: Thank you for your feedback. The text has been amended to read "Prior to data collection, athletes were led through a standardized warmup consisting of sprint-specific mobility drills, dynamic stretches, and familiarization with all equipment, including the LED pacing system and LPE setup."

Seven trials were removed, The text has been amended to read If participants failed to maintain the pace set by the LED, the trial was discarded (n=7).

Comment 5: Line 135 Provide justification for the 0.5 Hz selection; concern a cut-off of 0.5 Hz would attenuate signal velocity and rapid accel changes.

Response: Thank you for your feedback; it is well thought out. To follow up on your concern about the 0.5 Hz filter, we must tell you that our decision was based on the specific objective of extracting the underlying center of mass (CoM) velocity profile rather than step-to-step velocity fluctuations observed in the raw velocity data during sprint running. As shown in the attached plots, the raw filter exhibits oscillations due to the underlying braking and propulsive phases of each step. Filters with higher cut-offs (2 and 5 Hz) maintain these fluctuations, resulting in velocity curves that visually and numerically depend on the athlete’s instantaneous position within the stride. This is problematic for our analysis, as one of our key outcome measures was velocity at specific landmarks (2, 5, 10 m) following the selected pickup point, which was sensitive to stride timing rather than CoM velocity. At lower cut-off frequencies (1 Hz), the filter still exhibits noticeable oscillations due to individual step accelerations, especially at low speeds, where the fluctuations distort the smooth monotonic rise characteristic of CoM velocity during early acceleration. This is possibly due to the nature of our measurement technique, combined with a moving transition into acceleration.

Comment 6: Lines 150-151, 154 italicize r and p, and R²

Response: Thank you for your feedback, this has been corrected.

Comment 7: Line 156 italicize a_max as it appears in line 136

Response: Thank you for your feedback. I ran a search and all a_maxabbreviations in the text follow the same format. If I missed it please clarify which sentence.

Comment 8: Table 1 italicize the p values

Response: Thank you for your feedback, this has been changed.

Comment 9: References - some are the journal style while others are not; run a check

Response: Thank you, the document has been switched to a numbered reference style, and the bibliography has been updated.

Reviewer 3 Report

Comments and Suggestions for Authors

This manuscript investigates the relationship between acceleration ability initiated from a static start and a "pickup" (rolling) start, which is highly relevant to team sports. The study's finding that these are relatively distinct motor abilities challenges traditional assessment practices and is a valuable contribution. However, several methodological and reporting details need to be addressed to strengthen the scientific rigor and clarity of the manuscript.

Comments

The rationale for the chosen pickup entry velocities is insufficient and potentially flawed:

The study selected and because they "aligned closely with the sprint velocity thresholds defined by GPS unit manufacturers". This is a weak justification. The physiological or mechanical relevance of these manufacturer-defined thresholds to actual gameplay is not clearly established or discussed. A more robust justification, perhaps based on the typical velocity distribution of accelerations in the specific sports of the participants, did the authors checked that?
The LED pacing system (LED Rabbit) is a critical methodological component. The manuscript needs to provide more detail on how the system ensures the athlete matches the light speed and how deviations were identified before being discarded, because this is not clear. The authors could also report the number of discarded trials to establish the efficacy and standardization of the pacing method.
The authors mentioned that a entry velocity was excluded due to "greater line sway in the tether". This exclusion highlights a potential limitation of the equipment (1080 Sprint) for higher-velocity pickup starts, and this should be thoroughly discussed as a limitation of the study and its technology choice.

The extraction of key variables lacks sufficient detail:

The manual determination of the "pickup point" as a "sudden increase in velocity between and " is highly subjective. This introduces a significant risk of operator-dependent error. The authors either they must provide the inter- and intra-rater reliability for this manual determination method or justify its use over an objective breakpoint detection algorithm, did one considered?
The manuscript mentions that 2, 5, 10, and splits were extracted, but only was used in the analysis, why? Given the focus on acceleration mechanics, including an analysis of the time to cover the first of the acceleration phase (starting from the pickup point) would add significant, objective instructional value and improve the manuscript's completeness.
A order Butterworth lowpass filter was applied to the data. A cutoff frequency is exceptionally low for motion analysis, especially sprinting. This is likely to significantly attenuate and distort the true instantaneous acceleration peaks, which typically contain higher frequency content. The authors must rigorously justify this very low cutoff frequency with relevant literature or provide a sensitivity analysis showing its effect on values.

The discussion must differentiate the magnitude of the shared variance for each pickup velocity more explicitly. The static start's relationship to the pickup is moderate, while its relationship to the pickup is low (, ). This suggests that the faster the initial entry speed, the more distinct the motor ability becomes, which is a crucial, high-value insight that should be strongly emphasized.

The manuscript presents an important and relevant finding for strength and conditioning practice. The conclusion that static and pickup acceleration are relatively distinct motor abilities is strongly supported, particularly at the higher entry velocity. However, the rigor of the methods, especially the filtering cutoff and the subjective 'pickup point' determination, needs significant attention before acceptance.

Author Response

Note: All in-text amendments are highlighted in blue.

The rationale for the chosen pickup entry velocities is insufficient and potentially flawed:

Comment 1: The study selected 1.5 m/s and 3.0 m/s because they "aligned closely with the sprint velocity thresholds defined by GPS unit manufacturers". This is a weak justification. The physiological or mechanical relevance of these manufacturer-defined thresholds to actual gameplay is not clearly established or discussed. A more robust justification, perhaps based on the typical velocity distribution of accelerations in the specific sports of the participants, did the authors checked that?

Response: There are three reasons why we chose the entry velocities we used. First, The two entry velocities represent walking and jogging entry velocities which in gait research have different mechanical qualities e.g. walking double support phase, jogging no double support.

Second, the entry velocities selected allowed us to compare our findings to other research. For example, the Lacome (2014) article examined rugby and also provided thresholds for walking (0-1.9 m/s) and jogging (>1.9 m/s). Our values are also closely aligned with the article by Sonderegger, which used 1.6 (6 km/h), 3 (10.8 km/h), and 4.1 (15 km/h). The text has been amended to read. “These entry speeds were selected because they aligned with values observed in sport [4] and with similar values reported in related pickup acceleration research [12, 20, 21], allowing for comparison of our findings with those in other studies”

Comment 2: The LED pacing system (LED Rabbit) is a critical methodological component. The manuscript needs to provide more detail on how the system ensures the athlete matches the light speed and how deviations were identified before being discarded, because this is not clear. The authors could also report the number of discarded trials to establish the efficacy and standardization of the pacing method.

Response: Thank you for your feedback. The text has been amended to read "To ensure the athlete matched the LED pacing system, the participants were familiarized with the LED pacing system prior to testing. After each trial during testing, the researcher inspected the entry velocity on the 1080 Sprint tablet to ensure it was within 10% of the prescribed entry velocity. Seven trials were discarded given these criteria."

The text "During pilot testing, LED entry velocity accuracy was validated using high frame-rate (240 fps) distance–time video calibration at velocities up to 8 m/s. The authors found that LED-indicated pacing accuracy was between 0.2% to 4.5%, up to velocities of 8 m/s when compared to video-established velocities." was also added.

Comment 3: The authors mentioned that a 4.5 m/s entry velocity was excluded due to "greater line sway in the tether". This exclusion highlights a potential limitation of the equipment (1080 Sprint) for higher-velocity pickup starts, and this should be thoroughly discussed as a limitation of the study and its technology choice.

Response: Second, the LPE's signal recording at the 4.5 m/s entry was compromised, given that as athletes moved faster, the increased entry velocity led to greater vertical sway along the line. As a result, the signal presented with what Mangine et al [27] called “false peaks” on the signal, which made distinguishing the start or end point of some steps challenging. To abate this we used the same strategy Mangine used by integrating the aberrations into the previous step. When approached, the manufacturer suggested setting the tether resistance to 3 kg to mitigate this oscillation; however, the added resistance no longer simulates ‘free running’ and is likely to affect split times and a_max.

The extraction of key variables lacks sufficient detail:

Comment 4: The manual determination of the "pickup point" as a "sudden increase in velocity between 13 and 20 m" is highly subjective. This introduces a significant risk of operator-dependent error. The authors either they must provide the inter- and intra-rater reliability for this manual determination method or justify its use over an objective breakpoint detection algorithm, did one considered?

Response: Thank you for your feedback. The text has been amended to read “Previously, test-retest reliability over three testing occasions separated by at least seven days has been established [24]. Change in the mean was used to indicate systematic bias, coefficient of variation (CV%) to establish absolute consistency, and intraclass correlation coefficients (ICC) to reflect relative consistency. Breakpoint distances and breakpoint sample ICCs were >0.99, and CV% all under 10%, with some under 5%.”

Comment 5: The manuscript mentions that 2, 5, 10, and 20 m splits were extracted, but only a_max was used in the analysis, why? Given the focus on acceleration mechanics, including an analysis of the time to cover the first 5 m of the 30 m acceleration phase (starting from the pickup point) would add significant, objective instructional value and improve the manuscript's completeness.

Response: Thank you for your feedback. In the first instance, we thought that the split times would add significant value to the story we were trying to articulate. It was hypothesized that those who had higher static start a_max would have shorter split times. However, when we ran the correlations between static start acceleration a_max and the split times at the 1.5 m/s entry (r=0.33 to r=-0.39), and the 3.0 m/s entry velocity (r=-0.47 to -0.48), we found the correlations to be relatively small and the coefficients of determination (R²) between variables were less than 25%. This confirmed once more that static start a_max was a relatively independent quality to the split times. As such these results did not offer a great deal of new information to that which we already reported and discussed. For the sake of simplicity, we just reported the correlations between static amax and pickup acceleration a_maxat different entry velocities. However, if you would like us to add this to the manuscript, we will do so.

Comment 6: A 4th order 0.5 Hz Butterworth lowpass filter was applied to the data. A 0.5 Hz cutoff frequency is exceptionally low for motion analysis, especially sprinting. This is likely to significantly attenuate and distort the true instantaneous acceleration peaks, which typically contain higher frequency content. The authors must rigorously justify this very low cutoff frequency with relevant literature or provide a sensitivity analysis showing its effect on a_max values.

Response: Thank you for your feedback; it is well thought out. In response to your concern about the 0.5 Hz filter, we must state that our decision was based on the specific objective of extracting the underlying center of mass (CoM) velocity profile rather than step-to-step velocity fluctuations observed in the raw velocity data during sprint running. As shown in the attached plots, the raw filter exhibits oscillations due to the underlying braking and propulsive phases of each step. Filters with higher cut-offs (2 and 5 Hz, Figure 1) maintain these fluctuations, resulting in velocity curves that visually and numerically depend on the athlete’s instantaneous position within the stride. This is problematic for our analysis as one of our key outcome measures was velocity at specific landmarks (2, 5, 10 m) following the selected pickup point, become sensitive to stride timing rather than showing CoM velocity. At lower cut-off frequencies (1 Hz, Figure 2), the filter still exhibits noticeable oscillations due to individual step accelerations, especially at low speeds, where the fluctuations distort the smooth monotonic rise characteristic of CoM velocity during early acceleration. This is possibly due to the nature of our measurement technique, combined with a moving transition into acceleration.

Figure 1

Figure 2

Comment 7: The discussion must differentiate the magnitude of the shared variance for each pickup velocity more explicitly. The static start's relationship to the 1.5 m/s pickup is moderate, while its relationship to the 3.0 m/s pickup is low (r=0.34, p=0.06). This suggests that the faster the initial entry speed, the more distinct the motor ability becomes, which is a crucial, high-value insight that should be strongly emphasized.

Response: Thank you for your feedback. We agree with your comment and think it’s a great addition to the article. We have amended the discussion to add “The magnitude of shared variance between the static start and pickup conditions revealed additional insight. The shared variance between the static start and 1.5 m/s entry was moderate (r = 0.63, p < 0.001) but dropped to a low level at 3.0 m/s (r = 0.34, p = 0.06), indicating that kinematics diverges substantially from a static start as entry velocity increases. This pronounced reduction in shared variance shows that pickup acceleration at higher velocities differs fundamentally from static start acceleration, reinforcing that static starts do not generalize well to faster entry conditions and should be treated as a distinct motor task in both research and practice.”