Specificity in Change of Direction Training: Impact on Performance Across Different Tests Study Notes

Study Overview: Specificity in Change of Direction (COD) Training

• Original Title: Specificity in Change of Direction Training: Impact on Performance Across Different Tests.

• Authors: Michael Keiner, Konstantin Warneke, Josua Skratek, Bjorn Kadlubowski, Konstantin Beinert, Andreas Wittke, and Klaus Wirth (2025).

• Institutional Affiliations:

  • Department of Training and Exercise Science, German University of Health & Sport, Ismaning, Germany.

  • Institute of Human Movement Science, Sport and Health, University of Graz, Graz, Austria.

  • Department of Strength & Conditioning, DSC Arminia Bielefeld, Bielefeld, Germany.

  • Department of Strength & Conditioning, VFL Bochum, Bochum, Germany.

  • Department of Sport Science, German University of Health & Sport, Ismaning, Germany.

  • Institute of Sport Science, University of Applied Sciences Wiener Neustadt, Wiener Neustadt, Austria.

• Primary Objective: To evaluate the specificity of COD training by determining whether targeted training protocols (Triangle, 505, or Curved Sprint) yield superior adaptations when the training exercises align directly with the testing exercises, compared to standard soccer practice.

Introduction to Change of Direction (COD) in Soccer

• Physical Demands of Soccer: The sport is defined by intermittent submaximal running interspersed with high-intensity sprints. Recent longitudinal data from the English Premier League and Spanish LaLiga show significant increases in the volume of high-intensity running and sprint performance.

• COD Definition: Non-linear sprints involve accelerating, decelerating, and altering running direction in response to game demands. These are preplanned directional shifts requiring balance, speed, and control.

• Technical Requirements: Efficient COD movement depends on physical elements and technical skills, specifically proper foot placement and body posture.

• The Problem of Transfer:

  • Different COD tasks can be improved via plyometric training, strength training, or linear sprint training.

  • Transfer effects diminish as the complexity (sharpness and frequency of directional shifts) of the COD task increases.

  • Research indicates a shared variance of only $0 \text{--} 79\%$ between various COD tests. Because this shared variance is frequently below $50\%$, different COD tests may actually assess independent, discrete skills.

• Study Hypothesis: A specific training protocol matching the profile of the outcome test (angle, distance, speed, acceleration, and deceleration) will result in superior adaptations compared to divergent training or regular soccer practice.

Methodology and Subjects

• Study Design: A 4-arm, 4-week randomized controlled intervention conducted during the preseason.

• Participants:

  • Initially recruited: $84$ elite youth soccer players from 5 teams in the highest German junior division.

  • Final analysis sample: $n = 77$ players (Ages $12 \text{--} 19$ years).

  • Exclusion criteria: Injury/illness during testing or missing more than $2$ training sessions.

  • Exclusions: $n = 4$ due to illness; $n = 3$ due to personal/school commitments.

• Group Assignments:

  1. Triangle Sprint Training Group: $n = 21$.

  2. 505 Sprint Training Group: $n = 21$.

  3. Curved Sprint Training Group: $n = 17$.

  4. Control Group: $n = 18$ (Participated only in regular soccer training; non-randomized due to team constraints).

• Athlete Maintenance: Subjects maintained regular soccer training (4 days/week) and weekend matches. Monitoring of sleep and nutrition was not systematic, but athletes received annual briefings and were instructed to follow elite-level health standards.

Testing Procedures

• Environmental Controls: Tests were performed on artificial grass at the same time of day to minimize circadian effects. No fatiguing sessions were allowed within $3$ days of testing.

• Familiarization: Two sessions held $1$ week prior; subjects completed each test $4$ times ($2$ per side).

• Measurement Tools: Witty Time Kit double timing gate system (Microgate, Bolzano, Italy). Starting points were marked with a cap $0.75\,m$ from the gate to prevent early triggering by body/hand movement.

• Warm-up: Standardized 15-minute routine including $5$ minutes of submaximal running, running drills (high knees, butt kicks, lateral steps), movement preparation (lunges, hand walks), and $3 \times 20\,m$ acceleration runs.

• Tests Conducted:

  • Triangle Test: Two turns at $60^{\circ}$ angles; total distance of $10\,m$. Speeds measured at $5\,m$ and $10\,m$. Reliability: ICC $0.88 \text{--} 0.91$.

  • 505 Agility Test: One $180^{\circ}$ cut. Athletes accelerate for $10\,m$, time is measured for the entry and exit of a $5\,m$ zone after the turn. Reliability: ICC $0.81 \text{--} 0.93$.

  • Curved Sprint Test: Following the arc of the penalty area on a standard soccer field ($17\,m$ total sprint) with a split at $8.5\,m$. Reliability: ICC $0.89 \text{--} 0.93$.

Training Protocol

• Scheduling: Conducted on Match day plus 2 (MD+2), Match day minus 4 (MD-4), and Match day minus 2 (MD-2).

• Structure: Supervised by a strength and conditioning coach; sessions occurred before team training with a $30$-minute break.

• Volume/Intensity: $6$ repetitions ($3$ per side) of maximal intensity COD sprints assigned to the specific group focus. Subjects rested for $2$ minutes between sprints.

• Total Duration: Approximately $30$ minutes per session including warm-up.

Statistical Analysis

• Software: SPSS 25.1; visualization via Loffing (2022) syntax.

• Thresholds: Significance set at p < 0.05.

• Tests:

  • 1-way ANOVA for baseline differences ($T1$).

  • Repeated-measures ANOVA (Time and Group $\times$ Time interaction).

  • Post hoc Tukey tests for group differences.

  • Effect sizes: Partial eta squared ($\eta_p^2$) and Cohen’s $d$.

• Classification of Effects:

  • $\eta_p^2$: Large ($\ge 0.25$), Moderate ($\ge 0.1$), Small ($\ge 0.01$).

  • Cohen’s $d$: Large (> 0.8), Medium (> 0.5), Small (> 0.2).

• Smallest Worthwhile Change (SWC): Defined as $0.5$ of the between-athlete SDs, reflecting a Cohen effect size of $0.50$.

Results: Reliability and Group Comparison

• Intraday Reliability (Total Group):

  • Triangle Sprint (Right/Left): ICC $0.97/0.94$; CV $1.86/1.76\%$.

  • 505 Sprint (Right/Left): ICC $0.97/0.97$; CV $2.14/2.00\%$.

  • Curved Sprint (Right/Left): ICC $0.99/1.00$; CV $1.17/1.10\%$.

• Anthropometric Data (Means):

  • Triangle: Age $15.0 \pm 1.6\,years$; Mass $61.9 \pm 12.6\,kg$; Height $1.74 \pm 0.10\,m$.

  • 505: Age $15.3 \pm 1.7\,years$; Mass $62.3 \pm 13.7\,kg$; Height $1.73 \pm 0.13\,m$.

  • Curved: Age $15.4 \pm 2.2\,years$; Mass $60.1 \pm 13.2\,kg$; Height $1.71 \pm 0.12\,m$.

  • Control: Age $14.0 \pm 0.6\,years$; Mass $58.9 \pm 7.7\,kg$; Height $1.71 \pm 0.06\,m$.

Key Performance Outcomes

• Triangle and 505 Groups: Demonstrated significant (p < 0.05) performance improvements across all COD tests (Triangle, 505, and Curved Sprint).

• Curved Sprint Group: Improved significantly only in the Curved Sprint test.

• Control Group: Showed no significant gains in any test (p > 0.05).

• Specific Interaction Effects: The "Time $\times$ Group" interaction effect was significant for the Curved Sprint Left variable ($F = 3.63, p = 0.002, \eta_p^2 = 0.14$), but was not significant for most other tests.

• Post Hoc Findings for Curved Sprint Left:

  • Triangle group showed better performance enhancements than the Curved and Control groups ($d = 0.93$ and $1.19$ respectively).

  • 505 group showed better enhancements than the Control group ($d = 0.82$).

Discussion and Practical Applications

• Mechanisms of Improvement: Improvements are likely driven by enhanced technique efficiency, experience, and increased feedforward control. Significant directional changes require specific braking strategies (e.g., $180^{\circ}$ turns) that are not present in curved sprints.

• Efficiency of Sharp Cuts: Training methods involving cuts over $120^{\circ}$ (Triangle and 505) seem to provide a broader stimulus that transfers to curved sprinting, whereas the reverse is not true. Curved sprint training lacks the deceleration and sharp acceleration phases needed for the 505 and Triangle tests.

• Responders Analysis: Even when the target training exercise matched the test, only about half of the subjects were classified as responders. This suggests that technique training alone may be insufficient for all elite youth athletes and may need supplementation with strength or reactive strength training.

• Limitation in Elite Populations: Players in high-performance systems possess high baseline COD ability, resulting in a flatter progression curve and limited adaptation reserves.

• Recommendations for S&C Coaches:

  • Implement COD drills that closely mirror the specific movement demands of the sport’s target tasks.

  • Prioritize hard cuts/sharp directional shifts if specifically aiming to improve technique and efficiency in movement patterns relevant to high-intensity match play.

  • Note that defining "sport-specific" remains difficult due to differing tactical and technical demands in team sports.