*2.2. Procedures*

The study was conducted in February of 2019, in the middle of the 2018/2019 basketball season. Testing was performed in each of the clubs playing/training courts. All tests were performed during a single testing session lasting approximately 180 min per participant. Participants attended the sessions in larger groups and rotated between the testing sessions (see below). The subjects were instructed to refrain from any physical activity for at least 24 hours before testing. Testing started with anthropometric measurements, after which the subjects performed a warm-up (5 min of low intensity running, 8 repetitions of dynamic stretching and body weight activation exercises). After the warm-up, subjects were randomized into 4 groups to complete four testing stations in a random order: (i) jumping, (ii) COD, (iii) hip and trunk dynamometry; and (iv) hip and trunk range of motion (ROM). There was a 5-minute rest between the testing stations. Each test began after the subject reported that they felt comfortable with the task (no more than 3 practice trials were taken by a player).

#### 2.2.1. Functional Leg Length

Functional leg length was defined as the distance between anterior-superior iliac spine (ASIS) and the ground, and was measured with laser distance meter (LD 420, Stabila, Hungary). The subjects were instructed to stand with their bare feet on the ground, heels next to the wall, with feet separated hip distance apart. During left leg measurement subjects were instructed to put their left hand on the right shoulder and vice versa. Three measurements on each leg were performed and the mean from these three repetitions was taken for further analysis.

#### 2.2.2. Single-Leg Countermovement Jump

SLCMJ was performed on a force platform (Type 9260AA, Kistler, Winterthur, Switzerland). Subjects were informed to step on the force platform with their testing leg and hands on their hips. The opposite leg was slightly flexed at the knee, but was not touching the shin of the tested leg. Swinging with the opposite leg during jumping was not allowed. The subjects were instructed to jump as high as possible, with the countermovement depth being self-selected, land on two legs and a hold balanced position for 3 s. The jump would be accepted if all of the above-mentioned instructions were met. Three trials on left and right leg were performed with 30 s of rest between each trial. Peak force (N), peak power (W), and highest jump height (m) for each leg was taken for further analysis.

#### 2.2.3. Single-Leg Horizontal Jump

Tape measure was used to measure Single leg horizontal jump (SLHJ), which was performed on the basketball court su fficing the standards of international basketball federation (FIBA). Subjects were informed to put their testing leg with toe at the starting line and hands on the hips. Subjects performed a countermovement to self-selected depth before pushing themselves into the horizontal jump, landing onto both legs and holding a balanced position for 3 s. Three trials on each leg were performed with 30 s of rest between each trial. The distance (m) was measured to the nearest 0.01 m with a tape measure. The longest jump from each leg was taken for further analysis.

#### 2.2.4. Single-Leg Triple Jump

Single leg triple jump (SLTJ) was measured similarly to SLHJ. Subject performed three consecutive horizontal jumps in which they landed onto both legs and held a balanced position for 3 s. The jump would be accepted for if all of the above-mentioned instructions were met. As with the SLHJ, the longest of the three jumps was taken for further analysis for each leg.

#### 2.2.5. Single-Leg Lateral Jump

Tape measure was used to measure Single leg lateral jump (SLLJ). Subjects were informed to put the inner (i.e., medial) edge of their feet at the starting line and hands on their hips. When a subject was ready, he performed a countermovement to self-selected depth before pushing himself into lateral jump landing onto both legs and holding a balanced position for 3 s. Three trials on left and right leg were performed with 30 s of rest between each trial. Distance (m) was measured to the nearest 0.01 m with a tape measure. Longest jump from each leg was taken for further analysis.

## 2.2.6. Trunk Strength

Trunk strength assessment was done according to Markovic et al. [26] protocol. A trunk dynamometer (S2P Ltd., Ljubljana, Slovenia) with a bending beam load cell (model 1-Z6FC3/200 kg, HBM, Darmstadt, Germany) was used to measure trunk flexion, extension, and lateral flexion isometric strength. Isometric strength was measured as peak torque (Nm/kg) of the best one second interval during maximal voluntary contraction. All of the output variables were normalized with subject's body mass. During trunk extension measurement, subjects were standing with back turned towards the dynamometer with sensors on level of scapular spine and hands crossed on their shoulders. During trunk flexion measurement, subjects were standing turned face towards dynamometer with sensors on the same level as during extension and arms floating in the air to prevent their contribution. During trunk lateral flexion, subjects were standing sideways to the dynamometer, positioned so that their spine was in neutral position. The hand closer to the sensor was placed on their opposite shoulder and the other one was placed on opposite hip. During all trunk strength measurements, subjects were in their training shoes, standing with feet hip width distance apart, while a rigid strap was tightly fastened across pelvic girdle to achieve good fixation. During every trial subject was verbally encouraged to reach his best performance and hold it for 3–5 s. Each task was done three times with 60 s of rest in between. The best result for each task was taken for further analysis.

## 2.2.7. Hip Strength

Hip strength assessment was modeled based on Markovic et al. [27] protocol. A multipurpose dynamometer (Muscleboard, S2P Ltd., Ljubljana, Slovenija) was used to measure hip flexion, extension, abduction, adduction, internal and external rotation strength. Peak force values were multiplied by lever arm (leg length, in meter) to calculate hip torque. Isometric strength was measured as peak torque (PT) (Nm/kg) during one second interval of maximal voluntary contraction and rate of torque development (RTD) (Nm/ms) was measured in 100 ms interval as Δ torque/Δ time value. During all of the measurements, the o ffset of the sensor was performed with the relaxed leg, but with the subject having minimal contact with sensor (~5% MVC). Then, the subject was instructed to reach MVC as fast as possible. During all measured actions (except hip rotations), the distance between the mid-part of the aluminum brace of sensor and medial malleoli was set to 5 cm and a rigid strap was tightly fastened across pelvic girdle to achieve good fixation. Hip flexion and extension were measured unilaterally while the remaining tasks were performed bilaterally. During hip flexion (Figure 1A) subject was sitting on dynamometer with hands on the ground and the tested leg extended in knee with hip flexion (~30◦), while the non-tested leg was on the ground with knee flexion of ~90◦. During hip extension (Figure 1B) subject was laying prone on elbows on the ground, with the tested leg extended, non-tested leg was in the 90◦ knee flexion, resting on dynamometer surface. During abduction and adduction (Figure 1C) subject was sitting with legs hip apart, knees fully extended and hip in ~30◦ flexion. During hip internal and external (Figure 1D) rotation, the subject was kneeling on all fours with knees and hip in 90◦ flexion with knees hip apart. The best out of three results from each movement was taken for further analysis.

**Figure 1.** Position of subjects during isometric hip strength testing. (**A**) flexion; (**B**) extension; (**C**) abduction/adduction; and (**D**) internal/external rotation.

#### 2.2.8. Range of Motion

Passive hip range of motion (ROM) during flexion, extension and internal/external rotation was measured with a digital inclinometer (Baseline, Fabrication Enterprises Inc., White Plains, NY, USA) and abduction/adduction with a handheld goniometer (Baseline, Fabrication Enterprises Inc., White Plains, NY, USA). All of the measurements were performed by the same measurer to minimize error. For flexion and extension, the inclinometer was aligned between the tested side femur trochanter major and lateral condyle. Hip flexion ROM was measured with the subject in supine position (the non-tested leg was extended in knee and hip fixated) and the knee of the tested leg in extended position. The tested leg was then moved and kept extended during whole movement until first pelvic movement. During hip extension ROM, subject was in prone position, and tested leg was kept in knee flexion (~90◦) during whole movement until first pelvic movement. Hip internal/external rotation ROM was measured with the inclinometer located in the center of vertically positioned (with pendulum) tibia with the subject in pronated position and knee in 90◦ flexion. Internal and external rotation were performed until the first pelvic movement, with hand-stabilization on the pelvis. Hip abduction and adduction ROM were measured with goniometer with stationary arm pointed toward the opposite anterior superior iliac

spine and the movable arm pointed toward the patella of the tested leg. During abduction, the subject was laying supine with both legs in neutral position (start) from which the tested leg was moved into abduction until first pelvic movement (finish). During adduction, the subject was supine with non-tested leg laying from table in ~30◦ abduction and tested leg in neutral position (start) from which leg was moved into adduction until first pelvic movement (finish). Mean result from three attempts on each leg was taken for further analysis with all of the results expressed in degrees (◦).

For the trunk lateral flexion ROM, the subject was standing barefoot with feet hip width distance apart with his back and heels touching the wall. Starting position was measured with a tape as distance between the middle finger (hand on the wall) and the floor. Subjects performed lateral flexion, sliding downwards with the hands without breaking contact with the wall and lifting their heels of the ground. At the end of movement, the end position was measured, and the di fference between starting and end position was calculated. Mean result from three attempts on each side was taken for further analysis with all of the results expressed in meters (m).

#### 2.2.9. Change of Direction

For COD performance, the T-test (Figure 2) was used, as outlined by Semenik [28]. All of the COD testing was done on the basketball court su fficing the standards of international basketball federation (FIBA) using photocells timing gates (Brower Timing Systems; Draper, Utah). At the beginning of test, the subject was standing 30 cm behind the start/finish line where photocells were placed. Subject was instructed to sprint (filled line up in Figure 2) from the start to the first cone, touch the tip with one hand, shu ffle (dashed lines in Figure 2) to the cone opposite of the touching hand and touch the tip of the lateral cone, then shu ffle to the third cone and touch it with the first hand, shu ffle back to the middle cone touching the tip with the second hand and then pedal back (filled line down in Figure 2) to the finish line. One-minute recovery was given between each trial. The best out of three trials was taken for further analysis.

**Figure 2.** Schematic representation of the T-test.

#### *2.3. Statistical Analyses*

Due to the unknown reliability of the strength assessment with a novel MuscleBoard device (S2P Ltd., Ljubljana, Slovenia), particularly in view of RTD, we checked the intra-session reliability of the strength outcome measures that were used as potential predictors for COD performance. We used (a) intra-class coe fficient correlation to assess relative reliability, (b) absolute typical error and relative typical error (expressed as coe fficient of variation) to assess absolute reliability, and (c) paired-sample

two-tailed T-tests to check for systematic bias. We used the data from our larger study involving 115 basketball players, who had completed the exact same protocol for strength assessment, the only difference was the lower number of repetitions (2 compared to 3 in our study) for each task. We used guidelines of Koo and Li [29] for reporting Intraclass Correlation Coefficient (ICC). Based on the 95% confident interval of ICC estimate values < 0.5 = poor, 0.5–0.75 = moderate, 0.75–0.90 = good, and > 0.90 = excellent reliability.

We used multiple regression analysis to model prediction of COD performance (dependent variable) with asymmetries (independent variable). Independent variables were grouped into five categories of asymmetries (anthropometric asymmetries, lateral hip and trunk strength asymmetries, hip and trunk strength ratios, hip and trunk ROM asymmetries and jumping asymmetries)—a total of 33 potential independent variables.

Inter-limb asymmetry was calculated with the following equation [30]:

$$Asymmetry \left( \% \right) = \left( \frac{stronger-waveer}{stronger} \right) \times 100$$

Shapiro–Wilks tests were performed to assess the normality of distribution of independent variables; only 24 % (8/33) were considered to be normally distributed given an alpha level of *p* > 0.05. Step-wise regression analysis was performed for overall CODS performance using all independent variables. An analysis of standard residuals was carried out, which showed that the data contained no outliers (s standardized residuals minimum: −1.71, standardized residuals maximum: 2.51). A collinearity test indicated that multicollinearity was not a concern (minimum tolerance: 0.75, maximum VIF:1.33). The data met the assumption of independent errors (Durbin–Watson value: 2.26). Also, Breuch–Pagan (10.43, *p* > 0.05) and Koenker (8.32, *p* > 0.05) test indicated the homoscedasticity of the model. Statistical significance was set at alpha level of *p* ≤ 0.05 and all statistical procedures were conducted using the Statistical Package for the Social Sciences for Windows (v.26.0; SPSS Inc., Chicago, IL, USA). Mann–Whitney U test was used to for testing statistical significance in difference between hip PT and RTD, and left and right leg strength ratios.
