Effects of video filming angle on throwing form instruction: enhancing recognition of the “elbow dropped” position

Abstract

[Purpose] This study aimed to elucidate the usefulness of stick figure images as visual feedback in recognizing the “elbow dropped” position among baseball players and to identify the optimal filming angle for such recognition. [Participants and Methods] Study participants were 51 male baseball players aged 11–15 years who had been diagnosed with throwing-related shoulder or elbow injuries and who also exhibited the “elbow dropped” position. Shadow throwing was filmed from four angles (front, rear, ventral, and dorsal), converted into stick figure images using SPLYZA Motion, and presented in three formats: standard playback, frame-by-frame, and still images. The ease of recognizing the “elbow dropped” position was evaluated using a Visual Analog Scale. [Results] The stick figure images were significantly easier for participants to interpret than the actual throwing videos. The front and dorsal filming angles provided the clearest visualization, followed by the rear and ventral filming angles. [Conclusion] Filming from a front or dorsal angle is the most effective method for teaching baseball players to recognize the “elbow dropped” position. Stick figure images may enhance the players’ understanding of proper throwing mechanics.

INTRODUCTION

During the throwing motion, which consists of a whole-body kinetic chain, it is important to maintain not only upper limb girdle flexibility but also lower limb flexibility, trunk stability, and proper throwing mechanics^{1, 2)}. The kinetic chain is disrupted if any of these factors is compromised, which leads to inefficient throwing that places excessive stress on the shoulder and elbow joints of the throwing arm, potentially resulting in pain and throwing-related shoulder and elbow disorders³⁾. There are frequently multiple factors involved in the development of throwing-related shoulder and elbow impairments, including upper limb girdle function, trunk stability, lower limb function, and throwing mechanics. It is challenging to accurately identify the underlying pathology because symptoms typically appear only during the throwing motion⁴⁾. The throwing motion is generally divided into six major phases: the wind-up phase, early cocking phase, late cocking phase, acceleration phase, deceleration phase, and follow-through phase^{5, 6)}. One commonly reported fault in throwing mechanics is the so-called “elbow dropped” position⁷⁾, characterized by a decreased shoulder abduction angle from the late cocking phase to the acceleration phase. This position has been frequently associated with throwing-related injuries^{8, 9)}. Many studies suggest that muscle strength and flexibility deficits^{10, 11)}, as well as poor overall performance⁹⁾, contribute to this “elbow dropped” phenomenon. Thus, throwing disorders arise from a complex interplay of structural and functional issues, with poor throwing mechanics being a primary contributing factor^{8, 9)}. This makes proper throwing instruction a crucial treatment strategy for physical therapists. In orthopedic clinics, video recordings are commonly taken from multiple angles when providing throwing instruction to baseball players^{12, 13)}, and these recordings are used in visual feedback learning. However, it remains unclear which viewing angle is most effective for players to understand their own mechanics, instead of understanding provided by assessments solely from the physical therapist’s perspective. This study recorded the throwing motion of baseball players from four angles: front, rear, ventral, and dorsal. The SPLYZA Motion application (SPLYZA Co., Ltd., Shizuoka, Japan), which converts videos into stick figure representations, was then used to analyze the resulting footage. Stick figures have the advantage of being visually simple and easy to understand because they display only skeletal and joint movements in motion analysis, excluding detailed information such as clothing, facial features, or background. After I examined the effectiveness of stick figure animations, I presented participants with videos in the order of standard playback, frame-by-frame playback, and still images, and subsequently investigated which filming angle made it easiest for them to recognize the “elbow dropped” position.

PARTICIPANTS AND METHODS

In this study, the “elbow dropped” position was defined as a throwing motion in which the elbow is positioned lower than the line connecting the two shoulder joints during the late cocking phase⁷⁾. Study participants were 51 right-handed male baseball players (ages 11–15 years, there were no restrictions on playing position) diagnosed with throwing-related shoulder or elbow impairment and exhibiting the “elbow dropped” position. Participants were recruited from three orthopedic clinics in Tokyo where I work part-time. None of the participants were aware of their “elbow dropped” position during their throwing motion. Participants mean age was 12.8 ± 1.5 years (range: 11–15), mean height was 158.5 ± 7.4 cm (range: 145–172 cm), and mean weight was 53.7 ± 9.3 kg (range: 38–71 kg).

This study was conducted in accordance with the Declaration of Helsinki. The purpose of the study and details on collected data (all results anonymized and used solely for research purposes) was explained to both the participants and their parents. Consent was obtained from all parties, and all necessary ethical considerations were taken into account throughout the study.

First, shadow throwing videos were recorded on an iPad Air (Apple Inc., Cupertino, CA, USA) from four angles: front, rear, ventral, and dorsal (Fig. 1). Recorded videos were then converted into stick figure images using SPLYZA Motion (SPLYZA Co., Ltd.), a markerless motion capture analysis application installed on the iPad Air (Fig. 2). To assess the clarity of recognizing the “elbow dropped” position, participants were presented with both the original throwing videos and the stick figure videos (standard playback, frame-by-frame, and still images) for comparison. Additionally, the stick figure videos from all four angles were shown separately, and participants evaluated the comprehensibility of each angle using a Visual Analog Scale (VAS). The VAS used in this study consisted of a 10 cm straight line drawn on paper; the left end (0 cm) of the line indicated the viewed angle was incomprehensible and the right end of the line (10 cm) indicated it was very clear and fully comprehensible. Participants marked their responses with a pen.

Fig. 1.

Video filming directions (4 directions).

Fig. 2.

Stick figure images created with the SPLYZA Motion application.

For statistical analysis, the χ² test was used to compare the comprehensibility of the actual throwing videos and stick figure videos. A one-way ANOVA was performed to analyze differences in the comprehensibility of the four stick figure video angles. When a significant main effect was found, the Tukey–Kramer method was applied for post-hoc comparisons. The significance level for all statistical analyses was set at 1% (p<0.01). Statistical analysis was conducted using SPSS Statistics 22 (IBM, Armonk, NY, USA).

RESULTS

In the comparison between actual throwing videos and stick figure images, most participants indicated that it was easier to recognize the “elbow dropped” position in stick figure images than in actual videos (Table 1). This was true for all images taken from all four angles (χ² test; p<0.01). Stick figure image angles arranged in descending order of clarity according to VAS scores were the following: front view and dorsal view (equivalent clarity), rear view, and ventral view (Tukey–Kramer test, p<0.01; Table 2).

Table 1. Comparison of preferences for actual videos or stick figure images (n=51).

	Front view	Rear view	Ventral view	Dorsal view
Actual video (people)	2	1	4	2
Stick figure images (people)	49	50	47	49
Significant	*	*	*	*

χ² test, *p<0.01.

Table 2. Stick figure comprehensibility by filming angle (n=51).

	Front view	Rear view	Ventral view	Dorsal view
VAS (cm)	9.4 ± 0.7 (7.8–10)	8.3 ± 1.0 (6.2–10)	6.4 ± 1.2 (3.7–8.8)	9.3 ± 0.7 (7.2–10)

A one-way ANOVA & Tukey–Kramer test (post-hoc).

front view=dorsal view >rear view >ventral view (Tukey Kramer test, p<0.01).

Indicated values are mean ± standard deviation (minimum value−maximum value).

DISCUSSION

The results of this study indicate that filming from the front or dorsal angles are the most effective approach when using videos to supplement throwing instruction to elementary and junior high school baseball players who are unaware of their “elbow dropped” position. Additionally, stick figure videos are easier for players to understand than actual throwing videos.

Under the current Japanese health insurance system, medical reimbursement standards set by the Japanese Ministry of Health, Labour and Welfare define 20 minutes as one unit of clinical time. Some orthopedic clinics allocate two units (40 minutes) per session, but it is challenging to dedicate extensive time to video recording due to the need to conduct evaluations, treatments, and throwing instruction within this limited time frame. SPLYZA Motion is an effective feedback tool for clinical throwing instruction because it allows for markerless motion analysis and rapidly converts videos to stick figure images.

One reason why stick figure images were evaluated as easier to understand may be that elementary and junior high school students struggle to clearly recognize the elbow position in regular throwing videos due to excessive visual information. This supposition is supported by some study participants who commented that they were unsure where on the elbow they should focus and that the bulk of shoulder and elbow muscles made it difficult to accurately determine the location. In contrast, stick figure images simplified the body structure, making the left shoulder–right shoulder–right elbow alignment clearer and allowing participants to more easily recognize elbow height. Additionally, the 3D mode function of SPLYZA Motion may have contributed to a better three-dimensional understanding of the throwing motion.

Regarding differences in clarity based on filming angle, the front and dorsal views were the easiest for participants to understand, followed by rear and ventral views. The “elbow dropped” position is characterized by a decrease in shoulder abduction angle from the late cocking phase to the acceleration phase⁷⁾. The late cocking phase represents the transition from translational to rotational motion¹⁴⁾, during which the shoulder joint reaches maximum external rotation (MER)^{15, 16)}. This phase involves complex scapular movements, including internal rotation, upward rotation, and posterior tilt¹⁷⁾.

Comparing shooting angles (Fig. 2), the front view made the transition from translational to rotational motion clearly visible, and the front view also made elbow height easy to assess. The alignment of the left shoulder, right shoulder, and right elbow was clearly distinguishable in the dorsal view, but rotational motion was harder to perceive since the dorsal aspect of the body was constantly in frame. The rear view captured both translational and rotational movements but was less effective than the front and dorsal views in terms of assessing elbow height. Finally, although the ventral view was effective for capturing the entire throwing motion, it was not ideal for evaluating elbow drop. This may be because in the ventral view the left shoulder is obscured by trunk rotation, making it difficult to assess right shoulder position and trunk tilt angle. These findings suggest that filming from the front or dorsal angle is the most appropriate approach for evaluating the “elbow dropped” position in a limited clinical setting. Additionally, converting videos into stick figure images using SPLYZA Motion and providing immediate feedback may enhance players’ understanding.

This study has several limitations. First, the target population was restricted to elementary and junior high school players, and the findings may not be generalizable to high school, collegiate, or elite-level players. Second, the study focused solely on the recognition of elbow drop and did not assess the comprehensibility of the entire throwing form, including the movement of the lower limbs and trunk. Third, the impact of SPLYZA Motion-based instruction on remedying the “elbow dropped” position was not evaluated. Future research should aim to broaden the target age range, examine the effectiveness of using SPLYZA Motion for instructional purposes, and further explore the optimal shooting angles for evaluating the entirety of the throwing form.

Conflict of interest

The author declares no conflict of interest.