Sage Journals: Discover world-class research

Abstract

Background:

Shoulder instability is a significant cause of disability in pitchers. The advent of advanced pitch-tracking technology now provides an opportunity to investigate the mechanical factors associated with these injuries on a large scale in Major League Baseball (MLB).

Purpose/Hypothesis:

The purpose of this study was to identify preinjury advanced analytic and pitch-tracking metrics associated with surgically treated shoulder instability in MLB. It was hypothesized that pitchers requiring surgery would demonstrate distinct preinjury profiles in advanced performance analytics and Statcast data compared with matched controls.

Study Design:

Case-control study; Level of evidence, 3.

Methods:

MLB pitchers who underwent primary capsulolabral repair from 2017 to 2024 (n = 35) were identified using public web sources used in previous studies. Each case was matched 2:1 with uninjured control pitchers (n = 70) based on age, role, handedness, season, and total pitch count. Advanced analytic and various pitch-tracking metrics were compiled from MLB-sponsored databases commonly used in player evaluation and prior research. The statistical analysis included unpaired t tests to compare groups and binary logistic regression to identify risk factors.

Results:

MLB pitchers who underwent capsulolabral surgery demonstrated superior preinjury performance metrics (including the FanGraphs version of wins above replacement [fWAR], fielding independent pitching [xFIP], expected earned run average [xERA], Stuff+, and walks plus hits per inning pitched) and pitch characteristics (including fastball velocity and spin rate), along with a more medial release point compared with controls. Binary logistic regression identified both Stuff+ and a medial release point as independent factors associated with surgery. Notably, the composite Stuff+ metric demonstrated a stronger association than velocity or spin rate alone in the final model (P < 05 for all significant factors).

Conclusion:

Several modern advanced analytic and pitch-tracking metrics were looked at as potential risk factors for shoulder instability in MLB pitchers for the first time. Pitchers who underwent capsulolabral surgery were more likely to throw harder with higher spin rates and a more medialized release point than matched controls. Overall, MLB pitchers who were more effective and had superior Stuff+ were at the highest risk for shoulder instability requiring surgical intervention. These findings point to a potential trade-off in modern pitching; the pursuit of elite performance is linked to a greater likelihood of injury.

Keywords

baseball shoulder labrum biomechanics pitching mechanics Statcast

The modern game of baseball is defined by data. From amateur showcases to major league front offices, advanced tracking technologies have transformed how pitchers are evaluated, developed, and managed.²⁹ The widespread adoption of PITCHf/x in 2007, followed by TrackMan radar in 2017, and the current Hawk-Eye optical system in 2020, has produced a continuous, high-resolution record of every pitch thrown in Major League Baseball (MLB) competition.²⁹ These systems now quantify kinematics including velocity, spin, movement, arm angle, release extension, and release position with precision, generating an unprecedented view of the pitching motion.⁴⁰ As a result, analytics once confined to biomechanic laboratories are now available on a league-wide scale, fueling a generation of pitchers who train and compete according to measurable performance metrics.

Amid this analytical revolution, injury rates have continued to rise despite pitch counts never being lower.³³ The resources available to help pitchers throw harder, spin the ball faster, and generate more movement are greater than at any previous point in the sport's history.^8,41 The economic and competitive impact of upper-extremity injuries has therefore intensified,^28,39 prompting years of investigation into risk factors and workload management strategies.^1,6,19,31,32 Yet despite these interventions, both shoulder and elbow pathologies remain leading causes of lost playing time.¹⁴ This disconnect between technological advancement and injury prevention underscores a critical gap: the same metrics and technology used to optimize performance may also hold the key to identifying early signs of breakdown.

Most large-scale pitch-tracking studies to date have focused on the elbow, particularly ulnar collateral ligament (UCL) injuries.^{6,9,23,24,26,38,48} Using public Statcast data, investigators have linked increased velocity, spin rate, command, overall pitching ability, and pitch-type specific differences to elevated UCL-surgery risk.^{6,9,23,24,26,38,48} These works have established an important analytic framework connecting in-game mechanics to medical outcomes. However, shoulder pathology, especially injuries of the labrum and capsule, has received far less attention within this realm, despite accounting for roughly 17% of all pitcher injuries.⁷

The glenoid labrum deepens the socket and stabilizes the humeral head, but repetitive overhead throwing subjects it to extreme rotational and shear forces, particularly during late cocking and deceleration phases of the throwing motion.^4,36 Over time, these stresses can lead to attenuation of the capsule and tearing of the labrum.^4,30,36 Although nonoperative management is initially preferred, professional pitchers often require arthroscopic repair, and surgical outcomes remain guarded; only 40% to 60% regain their prior performance,^15,42 and rehabilitation from surgery commonly exceeds a year.¹⁶ These clinical realities highlight the need for objective, data-driven indicators capable of detecting early mechanical deterioration or potential risk factors before irreversible injury that requires surgical intervention occurs.

The present study leverages MLB's Statcast infrastructure and other composite analytic metrics (Stuff+, Location+, Pitching+, FanGraphs wins above replacement [fWAR], expected fielding independent pitching [xFIP]) to explore pitch-tracking signatures preceding shoulder capsulolabral surgery. By extending this advanced-analytics methodology beyond the elbow to the shoulder, this work aims to define any potential advanced analytic or pitch-tracking risk factors for shoulder instability treated operatively and thus inform future efforts toward evidence-based injury prevention in professional baseball.

Methods

Study Design and Ethical Considerations

This investigation was designed as a case-control study to identify potential advanced analytical and performance predictors of shoulder labral or capsular injury among MLB pitchers. The study compared a cohort of injured athletes with a matched cohort of uninjured controls. As it did not qualify as human participants research per 45 Code of Federal Regulations Part 46.102, this investigation was classified as exempt from institutional review board oversight at our institution.

Case Identification and Cohort Assembly

The case group consisted of MLB pitchers who underwent primary shoulder capsulolabral repair between March 1, 2017, and December 1, 2024. Potential cases were initially identified by reviewing injured list designations for shoulder-related conditions during this time frame, cross-referencing FanGraphs and Spotrac injury databases.^11,44 Verification of both the diagnosis and the surgical management was performed through public news reports and official team announcements. To minimize bias, 3 independent reviewers (M.R.D., N.F., and M.A.M.) carried out this classification process, with disagreements resolved through consensus discussion, yielding an initial pool of 39 pitchers. To be included in the final case cohort for the primary analysis, pitchers were required to have ≥2 consecutive seasons of qualifying preoperative performance data at the MLB level, defined as throwing a minimum of 100 tracked pitches per season. The application of this criterion resulted in a final case cohort of 35 pitchers. A separate consideration was made for the analysis of 3 advanced metrics (Stuff+, Location+, and Pitching+) and arm angle, which have only been publicly recorded since 2020. To maximize statistical power for these variables, the 2-season data requirement was waived, resulting in a subanalysis cohort of 21 cases and 42 controls for the season preceding surgery (T1) and 22 cases with 47 controls for the index surgical season.

Control Identification and Cohort Assembly

A matched-control cohort was assembled at a 2:1 ratio (n = 70), employing an established methodology from prior case-control studies of UCL injuries in MLB pitchers.^23,24 To account for recognized covariates in pitching-related injury risk, each control was matched to a case by the season of surgery, age at the index year, handedness, pitcher role (starter vs reliever), and total pitch count. To be eligible, controls were required to have no history of shoulder surgery and to have thrown ≥100 tracked pitches in each of the 2 analyzed seasons, thereby ensuring comparable data availability with the case cohort.

Data Sources and Collection

Data were aggregated from several publicly available and well-validated baseball databases, primarily Baseball Savant, Brooks Baseball, and FanGraphs. These platforms compile and report statistics, advanced performance metrics, and Statcast tracking data. Statcast, which is installed in all MLB stadiums, uses high-resolution optical cameras and radar-based tracking technologies to capture every pitch.²⁹ Parameters measured by Statcast include pitch velocity, spin rate, release point coordinates, and horizontal and vertical movement.⁴⁰ An analytical data set was constructed in a structured Excel workbook with separate sheets for cases and controls at each time point: the index season (year of surgery) and the 3 preceding seasons (T1, T2, T3). Demographic and descriptive data included age, height, weight, body mass index (BMI), handedness, pitcher role (starter or reliever), and draft position. Average workload and performance data consisted of annual pitch counts and advanced sabermetric measures of overall value and effectiveness, including fWAR), xFIP, and skill-interactive ERA, as reported by FanGraphs.

Mean pitch-tracking data included fastball velocity, slider velocity, and curveball velocity. Pitch usage was measured as the percentage of total pitches that were fastballs, sliders, curveballs, or changeups. Spin rate was recorded as the rotational velocity of the baseball upon release, expressed in revolutions per minute (rpm). Release point variables included extension, defined as the distance in feet from the pitching rubber to the point of release toward home plate, as well as horizontal and vertical release positions, which represent the spatial coordinates of the ball at release relative to the center of the rubber. Arm angle was defined as the horizontal angle, measured in degrees, between a horizontal line parallel to the ground and a line parallel to the pitcher's throwing shoulder and the ball at release (Figure 1).

Figure 1.

Illustration of pitch release kinematic measurements. Pos X represents the horizontal release position, a measure of the horizontal distance of the ball at release from the center of the rubber. Pos Z represents the vertical release position, which is the height of the ball compared with the rubber. Arm angle is the horizontal angle between a horizontal line parallel to the ground and a line parallel to the pitcher's throwing shoulder and the ball at release.

Ball movement was also measured, with horizontal and vertical movement representing deviations in inches from a theoretical nonspinning trajectory. For analytical consistency, all horizontal movement metrics, along with horizontal release position, were normalized for pitcher handedness by using absolute values, while vertical movement, which is independent of handedness, retained positive and negative values. Approach angle was measured as both vertical approach angle and horizontal approach angle above average. These variables describe the trajectory of the pitch as it enters the strike zone and are normalized for release point and handedness, providing a standardized measure of pitch path. Finally, advanced composite metrics were included in the data set, specifically Stuff+, Location+, and Pitching+, which are proprietary statistics reported by FanGraphs beginning in 2020 and widely valued throughout the league (Table 1). These scores provide composite evaluations of pitch quality (Stuff+), command (Location+), and overall pitching effectiveness (Pitching+), with a value of 100 representing roughly the league average.

Table 1

Definitions of Commonly Used Advanced Analytics in MLB ^a

Advanced Analytics in MLB	Definition
FanGraphs wins above replacement (fWAR)	fWAR is a cumulative statistic that measures a player's value in all facets of the game by comparing how many more “wins” he's worth to a team than a replacement-level player, who is a readily available minor leaguer or a free agent; previous reports value 1 WAR to be equivalent to nearly $7 million in payroll value for pitchers, with an even greater value per WAR as a player exceeds over 2 WAR in a season.^7,27; affected by amount played; named after the developing organization
Expected fielding independent pitching (xFIP) and skill-interactive earned run average (SIERA)	Rate-based statistics that aim to measure what a pitcher's earned run average (ERA) would be after standardizing for the strength of defense, ballpark factors, and other possible confounding variables; not affected by amount played
Stuff+	Stuff+ evaluates the physical attributes of pitch movement such as velocity, vertical and horizontal movement, approach angle, and release extension; the most effective pitchers in MLB tend to grade out high in Stuff+¹²
Location+	Location+ is a count- and pitch type–adjusted judge of a pitcher's ability to put pitches in the right location, better known as their strike-zone command. Location+ looks at average locations of a specific pitch on a specific pitch count and assumes the pitcher's intent is generally the same across the league; command is critical for success and is often what distinguishes between the elite pitchers
Pitching+	Pitching+ uses the physical characteristics (Stuff+), command (Location+), batter handedness, stadium, and count of each pitch thrown to judge overall pitching ability; Pitching+ also captures the impact of pitch sequencing to set up and confuse hitters for future pitches, such as changing eye levels by throwing a high fastball followed by a low-and-outside slider; therefore, Pitching+ may also capture the mental process and execution of a pitcher that is otherwise difficult to evaluate, and Pitching+ has shown to be a strong predictor of future success³

MLB, Major League Baseball; UCL, ulnar collateral ligament. Stuff+, Pitching+, and Location+ are all scaled similarly to IQ: a score of 100 is graded out as approximately league average. Table reproduced with permission from Mastroianni MA, Kunes JA, El-Najjar DB, et al. Advanced analytic and pitch-tracking metrics associated with UCL surgery in Major League Baseball pitchers: a case-control study. Orthop J Sports Med. 2025;13(2):23259671241302432.

Statistical Analysis

Comparisons between cases and controls were performed at 2 primary time points: the season immediately prior to surgery (T1) and the index season. Continuous variables were reported as mean ± SD and compared using 2-sample Welch t tests. Categorical variables were assessed with Fisher exact test. These comparisons were conducted as unpaired analyses, meaning the matching procedure was not explicitly modeled in the hypothesis tests. While a 2:1 matched design was used to assemble the cohort and ensure baseline comparability, an unpaired analysis was retained for the primary comparisons to maximize statistical power and avoid the data loss associated with incomplete case-control sets in a conditional model.

To evaluate longitudinal preinjury trends, pitching metrics were analyzed across the 3 preoperative seasons (T3−T1). For each pitcher with ≥2 available preoperative data points, a simple least squares regression was performed to calculate an individual slope representing the average rate of change per season for each metric. These slopes were then compared between cases and controls using Welch t tests.

To identify independent predictors of injury, a multivariable logistic regression model was constructed using T1 data. Author judgment was used to exclude certain variables related to each other (eg, Pitching+ is influenced heavily by Stuff+ and Location+), to minimize collinearity while capturing clinically relevant metrics. Because fastball velocity, spin rate, and Stuff+ are closely interrelated measures of pitch quality, we first ran a preliminary model including only these variables to determine which demonstrated the strongest independent association with injury risk. The variable showing the most robust relationship was then carried forward into the final multivariable model with additional mechanical and usage metrics. Adjusted odds ratios (aORs) and their 95% CIs were calculated and reported for interpretable increments of each predictor. The logistic regression models, which included the post-2020 metrics Stuff+ and Location+, were necessarily run on this smaller subanalysis cohort. A complete case analysis was performed, resulting in final models derived from 22 cases and 47 controls.

Statistical Analysis and Software

All statistical analyses were performed using R (Version 4.3.1), with the dplyr, tidyr, broom, and flextable packages. A 2-sided P value <.05 was considered statistically significant for 2-sample t tests, whereas a threshold of P < .10 was used in the logistic regression model to identify potential predictors of injury.

Results

Demographics and Baseline Characteristics

A total of 35 cases and 70 matched controls were included in the primary analysis. There were no statistically significant differences between the case and control cohorts with respect to age, height, weight, BMI, draft position, or career MLB seasons. The groups were perfectly matched for handedness and pitcher role, as per the study design. Baseline workload metrics, including innings pitched and total pitch count in the season prior to the index year (T1), were also similar between groups (all P > .10), as detailed in Table 2.

Table 2

Demographics and Baseline Characteristics of MLB Pitchers Undergoing Capsulolabral Surgery Compared with Matched Controls^a

Variable	Cases (n = 35)	Controls (n = 70)	P
Age, y	28.7 ± 3.2	27.9 ± 2.9	.19
Height, cm	190.2 ± 6.5	188.4 ± 6.3	.19
Weight, kg	100.6 ± 9.7	97.5 ± 9.7	.15
BMI, kg/m²	27.9 ± 3.2	27.2 ± 2.7	.29
Draft position	238.4 ± 282.8	230.1 ± 269.2	.90
Career MLB seasons	6.8 ± 3.3	6.7 ± 3.4	.92
Right-handed	77.1%	77.1%	≥.99
Starting pitcher	45.7%	47.1%	≥.99
International free agent	14.3%	21.4%	.44
Innings pitched, T1	75.5 ± 54.2	61.8 ± 47.5	.13
Pitch count, T1	1250.3 ± 895.5	1042.1 ± 770.8	.15

Data are presented as mean ± SD or percentage. BMI, body mass index; MLB, Major League Baseball; T1, season prior to injury. P values were calculated using 2-sample Welch t tests for continuous variables and Fisher exact test for categorical variables. All between-group comparisons (case vs control) were not significant (P > .05).

Performance Metrics and Advanced Analytics

Analysis of the 3 plus metrics (Stuff+, Location+, Pitching+) included 21 cases and 42 controls for T1 and 22 cases and 47 controls for the index surgical season, secondary to our loosened inclusion criteria in a limited sample size secondary to these metrics only being available since 2020. Analysis of advanced performance metrics revealed significant differences between groups. In the season before surgery (T1), cases had a lower earned run average (ERA) than controls (3.90 ± 1.37 vs 4.98 ± 2.32; P < .01) and maintained an advantage in the index season (3.96 ± 1.87 vs 4.78 ± 1.78; P = .03). FIP was also superior among cases at T1 (3.92 ± 0.91 vs 4.50 ± 1.34; P = .01), though this difference did not persist in the index season. Similarly, cases demonstrated higher fWAR values at T1 (0.95 ± 1.11 vs 0.50 ± 0.95; P = .045), while expected ERA was significantly lower (3.91 ± 0.92 vs 4.57 ± 1.28; < .01). Of the advanced Statcast-based metrics, Stuff+ was markedly higher in cases at both T1 (105.85 ± 10.00 vs 97.33 ± 10.59; P < .01) and the index season (102.95 ± 9.57 vs 96.77 ± 10.72; P = .02). In contrast, Location+ and Pitching+ did not differ significantly at either time point (Table 3).

Table 3

Advanced Performance and Analytical Metrics in Capsulolabral Surgery Cases Versus Matched Controls ^a

	Season Before Surgery (T1)			Season of Surgery (Index)
Variable	T1: Case	T1: Control	P (T1)	Index: Case	Index: Control	P (Index)
ERA	3.90 ± 1.37	4.98 ± 2.32	<.01	3.96 ± 1.87	4.78 ± 1.78	.03
Innings pitched	78.19 ± 56.78	59.94 ± 46.79	.11	59.06 ± 44.89	70.72 ± 51.03	.24
WHIP	1.29 ± 0.21	1.43 ± 0.38	.02	1.30 ± 0.31	1.39 ± 0.31	.14
FIP	3.92 ± 0.91	4.50 ± 1.34	.01	4.12 ± 1.26	4.53 ± 1.07	.11
fWAR	0.95 ± 1.11	0.50 ± 0.95	.045	0.73 ± 1.06	0.69 ± 1.16	.87
xFIP	4.10 ± 0.86	4.40 ± 1.01	.11	4.22 ± 0.90	4.48 ± 0.89	.17
SIERA	3.97 ± 0.90	4.21 ± 0.86	.21	4.11 ± 0.98	4.29 ± 0.80	.35
xERA	3.91 ± 0.92	4.57 ± 1.28	<.01	4.11 ± 1.36	4.61 ± 1.28	.08
Location+, pt	98.25 ± 6.84	100.90 ± 5.63	.14	98.29 ± 7.15	99.88 ± 6.74	.39
Stuff+, pt	105.85 ± 10.00	97.33 ± 10.59	<.01	102.95 ± 9.57	96.77 ± 10.72	.02
Pitching+, pt	103.32 ± 11.49	98.86 ± 10.50	.15	100.86 ± 12.61	97.14 ± 9.94	.23

Data are presented as mean ± SD. Bold P values indicate statistically significant difference at P < .05. Comparison of ERA, FIP, fWAR, xERA, and advanced Statcast-based metrics (Stuff+, Location+, Pitching+) in cases and controls during the season before surgery (T1) and the index season. ERA, earned run average; FIP, fielding independent pitching; fWAR, wins above replacement (FanGraphs); SIERA, skill-interactive ERA; WHIP, walks plus hits per inning pitched; xERA, expected ERA; xFIP, expected FIP. Advanced analytic metrics (Stuff+, Location+, Pitching+) data have only been recorded since 2020. To maximize statistical power for these variables, the 2-season data requirement was waived, resulting in a subanalysis cohort of 21 cases and 42 controls for the season preceding surgery (T1) and 22 cases and 47 controls for the index surgical season.

Pitch-Tracking Characteristics

Pitch-tracking variables demonstrated key differences between cohorts. At T1, cases had significantly higher average fastball velocity compared with controls (94.33 ± 2.23 mph vs 93.35 ± 2.08 mph; P = .03), though this difference diminished by the index season. Spin rate was also greater in cases at T1 (2293.43 ± 190.34 rpm vs 2200.60 ± 198.40 rpm; P = .02), with a nonsignificant trend persisting into the index season (P = .06). Horizontal release position (release pos X) was more medial (closer to the body) in cases compared with controls at both T1 (1.69 ± 0.59 feet vs 1.97 ± 0.58 feet; P = .02) and the index season (1.63 ± 0.70 feet vs 1.91 ± 0.53 feet; P = .04). Other variables, including slider and curveball velocity, usage percentages, release extension, arm angle, and ball movement parameters, did not differ significantly between groups (Table 4).

Table 4

Statcast Pitch-Tracking Metrics in Capsulolabral Surgery Cases Versus Controls ^a

	Season Before Surgery (T1)			Season of Surgery (Index)
Variable	T1: Case	T1: Control	P (T1)	Index: Case	Index: Control	P (Index)
FB velocity, mph	94.33 ± 2.23	93.35 ± 2.08	.03	93.94 ± 2.22	93.31 ± 2.25	.18
SL velocity, mph	85.04 ± 2.94	84.59 ± 2.81	.49	84.10 ± 3.28	84.39 ± 2.77	.67
CB velocity, mph	80.23 ± 3.42	78.78 ± 2.89	.11	80.06 ± 3.43	79.07 ± 3.10	.27
FB usage, %	0.35 ± 0.09	0.36 ± 0.07	.59	0.37 ± 0.09	0.38 ± 0.08	.63
SL usage, %	0.26 ± 0.14	0.24 ± 0.14	.44	0.25 ± 0.14	0.25 ± 0.14	.82
CB usage, %	0.20 ± 0.13	0.14 ± 0.11	.09	0.18 ± 0.13	0.15 ± 0.11	.30
Changeup usage, %	0.11 ± 0.09	0.16 ± 0.13	.06	0.12 ± 0.10	0.16 ± 0.11	.06
Spin rate, rpm	2293.43 ± 190.34	2200.60 ± 198.40	.02	2280.80 ± 189.59	2201.99 ± 222.30	.06
Release extension, feet	6.15 ± 0.42	6.17 ± 0.49	.79	6.12 ± 0.45	6.20 ± 0.48	.39
Release pos Z, feet	5.87 ± 0.43	5.88 ± 0.37	.88	5.87 ± 0.41	5.83 ± 0.39	.65
Release pos X, feet	1.69 ± 0.59	1.97 ± 0.58	.02	1.63 ± 0.70	1.91 ± 0.53	.04
Arm angle, deg	41.83 ± 8.89	39.70 ± 11.28	.42	42.50 ± 10.48	38.55 ± 10.38	.15
Horizontal movement, inches	−1.48 ± 5.43	−2.12 ± 5.94	.61	−1.12 ± 5.39	−2.40 ± 5.92	.27
Vertical movement, inches	7.19 ± 3.47	8.11 ± 2.88	.22	7.25 ± 3.97	7.36 ± 3.12	.89
HAA, deg	1.58 ± 0.68	1.73 ± 0.72	.33	1.51 ± 0.88	1.63 ± 0.65	.48
HAA above average, deg	−0.19 ± 1.50	−0.01 ± 1.22	.56	−0.18 ± 1.43	0.07 ± 0.86	.368
VAA, deg	−6.58 ± 0.68	−6.56 ± 0.45	.88	−6.62 ± 0.64	−6.61 ± 0.58	.95
VAA above average, deg	−1.44 ± 2.64	−1.95 ± 3.15	.42	−1.83 ± 2.85	−1.58 ± 3.05	.67

Data are presented as mean ± SD. Bold values indicate statistically significant difference at P < .05. Comparison of pitch velocity, usage percentages, spin rate, release point variables, arm angle, and pitch movement in cases and controls during the season before surgery (T1) and the index season. CB, curveball; FB, fastball; HAA, horizontal approach angle; pos X, horizontal release position; pos z, vertical release position; rpm, revolutions per minute; SL, slider; VAA, vertical approach angle. Arm-angle data have only been recorded since 2020. To maximize statistical power for this variable, the 2-season data requirement was waived, resulting in a subanalysis cohort of 21 cases and 42 controls for the season preceding surgery and 22 cases and 47 controls for the index surgical season.

Longitudinal Preinjury Trends

Longitudinal slope analysis from T3 to T1 did not reveal significant differences in per-season changes for most variables. Cases trended toward greater increases in curveball usage compared with controls (Δ per season 0.03% ± 0.05% vs −0.00% ± 0.05%; P = .057) and toward less favorable vertical approach angle changes (−0.02°± 0.18° vs −0.14°± 0.26°; P = .07), although these did not meet the threshold for statistical significance. Similarly, controls demonstrated a nonsignificant trend toward greater improvement in Location+ (Δ 5.74 ± 7.58 points vs 0.89 ± 7.06 points; P = .07). No other workload, velocity, spin, or performance metrics differed between groups over this preinjury interval (Table 5).

Table 5

Longitudinal Slopes of Pitch-Tracking and Performance Metrics (T3–T1) ^a

Variable	Case Δ/Season	Ctrl Δ/Season	P (Welch)
Curveball usage, %	0.03 ± 0.05	−0.00 ± 0.05	.057
Changeup usage, %	−0.01 ± 0.05	0.01 ± 0.05	.14
Fastball usage, %	0.00 ± 0.04	0.00 ± 0.06	.74
Slider usage, %	0.02 ± 0.06	0.01 ± 0.07	.62
ERA	0.13 ± 1.18	0.09 ± 2.32	.92
FB velocity, mph	0.16 ± 0.77	0.03 ± 0.77	.52
FIP	−0.01 ± 0.67	0.05 ± 1.15	.78
Innings pitched	11.76 ± 34.17	0.74 ± 30.93	.18
Pitch count	223.12 ± 550.14	6.65 ± 492.53	.11
SIERA	0.01 ± 0.63	−0.09 ± 0.76	.57
fWAR	0.02 ± 0.76	−0.05 ± 0.78	.69
WHIP	0.05 ± 0.16	0.01 ± 0.36	.57
HAA, deg	−0.06 ± 0.26	0.00 ± 0.29	.48
HAA above average, deg	0.16 ± 0.60	0.17 ± 0.51	.93
Horizontal movement, inches	0.73 ± 1.07	0.53 ± 1.71	.61
VAA, deg	−0.02 ± 0.18	−0.14 ± 0.26	.07
VAA above average, deg	0.37 ± 1.11	0.66 ± 1.81	.48
Vertical movement, inches	−0.57 ± 0.95	−0.65 ± 1.33	.81
Location+, pt	0.89 ± 7.06	5.74 ± 7.58	.07
Pitching+, pt	3.68 ± 8.26	5.06 ± 9.36	.66
Stuff+, pt	2.83 ± 4.64	0.11 ± 4.96	.12
Arm angle, deg	0.17 ± 3.23	−0.87 ± 4.03	.42
Release extension, feet	0.05 ± 0.14	0.01 ± 0.15	.25
Release pos X, feet	−0.15 ± 0.30	−0.03 ± 0.26	.10
Release pos Z, feet	−0.01 ± 0.09	0.00 ± 0.11	.56
Spin rate, rpm	42.52 ± 98.11	12.17 ± 71.84	.18
Velocity, mph	−0.31 ± 0.78	−0.18 ± 0.96	.54
xERA	−0.01 ± 0.64	0.14 ± 1.26	.51
xFIP	0.01 ± 0.61	−0.08 ± 0.79	.60

Data are presented as mean ± SD. Longitudinal slopes of pitch-tracking and performance metrics from T3 to T1 were calculated for capsulolabral surgery cases and matched controls. ERA, earned run average; FB, fastball; FIP, fielding independent pitching; fWAR, wins above replacement (FanGraphs); HAA, horizontal approach angle; pos X, horizontal release position; pos z, vertical release position; pt, point; rpm, revolutions per minute; SIERA, skill-interactive ERA; Stuff+, Pitching+, and Location+, PitchingBot advanced run prevention metrics; VAA, vertical approach angle; WHIP, walks plus hits per inning pitched; xERA, expected ERA; xFIP, expected FIP.

Preliminary Logistic Regression

The preliminary regression compared fastball velocity, spin rate, and Stuff+ to evaluate which was the stronger predictor of capsulolabral surgery. Stuff+ showed the greatest effect size and approached statistical significance (P = .06), suggesting it better captures underlying risk than velocity or spin alone (Table 6). This variable was therefore included in the subsequent multivariable model alongside mechanical and usage metrics.

Table 6

Multivariable Logistic Regression Model for Predictors of Capsulolabral Surgery ^a

Variable	Coefficient	aOR (95% CI)	P
Fastball velocity	0.012	1.012 (0.721-1.421)	.95
Stuff+	0.068	1.972 (0.978-3.977)	.06
Spin rate	0.001	1.134 (0.836-1.539)	.42

The model analyzes predictor variables from the season prior to surgery (T1) to determine the adjusted odds ratio (aOR) of undergoing surgery. The aORs are reported for clinically relevant increments: a 1-mph change for fastball velocity; a 10-point change for Stuff+; and a 100–rate per minute change for spin rate. A P value <.10 was used to identify potential risk factors.

Multivariable Logistic Regression

In the full multivariable logistic regression model using T1 data, 2 variables demonstrated notable associations with capsulolabral surgery risk. A more medial horizontal release position was significantly associated with increased odds of surgery (aOR, 0.86 per 0.1 feet; 95% CI, [0.74-0.99]; P = .03). Higher Stuff+ was also a significant independent predictor, with greater values associated with increased odds of surgery (aOR, 2.07 per 10-point increase; 95% CI, [1.10-3.88]; P = .02). Full results of the multivariable model are presented in Table 7.

Table 7

Multivariable Logistic Regression Model for Predictors of Capsulolabral Surgery ^a

Variable	Coefficient	aOR (95% CI)	P
Release position X	−1.557	0.86 (0.74-0.99)	.03
Location+	−0.096	0.38 (0.12-1.23)	.11
Stuff+	0.073	2.07 (1.10-3.88)	.02
Fastball usage	−6.497	0.52 (0.24-1.14)	.10

The model analyzes predictor variables from the season prior to surgery (T1) to determine the adjusted odds ratio (aOR) of undergoing surgery. The aORs are reported for clinically relevant increments: a 0.1-foot change for horizontal release position; a 10-point change for Location+ and Stuff+; and a 10–percentage point change for fastball usage. Bold P values indicate statistically significant difference at P < .05. A P value <.10 was used to identify potential risk factors.

Discussion

Several modern advanced analytic and pitch-tracking metrics were looked at as potential risk factors for capsulolabral injuries requiring repair in MLB pitchers for the first time. Pitchers who underwent capsulolabral surgery were more likely to throw harder with higher spin rates and a more medialized release point than matched controls. Injured pitchers exhibited significantly lower ERAs, superior FIP metrics, and higher player value. Overall, MLB pitchers who were more effective and had superior Stuff+ were at the highest risk for capsulolabral injuries requiring surgical intervention. These findings suggest that the very analytical metrics now driving player development and evaluation throughout the MLB, and the physical capabilities required to achieve elite scores in performance metrics, may themselves represent risk factors for debilitating shoulder injury.

The modern era of baseball has been defined by an explosion of advanced analytics that have reshaped how pitchers are evaluated and trained. Statcast and similar systems now capture every measurable aspect of pitch behavior, fueling the rise of composite “plus” metrics that integrate velocity, spin, movement, release, and other beneficial physical attributes into a single run value–based score.²⁷ These models, developed through machine learning and validated on large sets of pitches, have rapidly become a benchmark for both player evaluation and development. Recently, Mastroianni et al,^25,27 expanded these analytics into injury research, demonstrating that pitchers who ultimately underwent UCL reconstruction possessed higher velocity, superior command (Location+), and higher overall performance (Pitching+) compared with uninjured controls, with velocity emerging as the dominant risk factor. That association reinforces decades of biomechanical work linking fastball velocity to greater elbow valgus stress,^2,5,12,13,31 yet it also highlights how performance optimization and injury risk can coexist within the same domain.

Our findings parallel this broader “performance paradox” but identify a potentially distinct risk signature for the shoulder. Whereas elbow injuries appear most closely tied to pure output magnitude, measured by how hard a pitcher throws, our regression analyses revealed that Stuff+, rather than velocity or spin alone, was the strongest predictor of capsulolabral surgery. Spin can be affected by a variety of factors including arm speed and angle, release point, and grip. Spin rate has been shown to be primarily driven by arm and shoulder rotational velocity,^40,48 while spin axis, direction, and efficiency is largely driven by forearm pronosupination.^4,20 In baseline comparisons, all 3 variables were elevated among injured pitchers, but once entered together, the composite measure outperformed its individual components. This pattern likely reflects the differing biomechanical demands of the 2 joints. The elbow functions primarily as a hinge transmitting valgus stress,⁴⁵ where torque scales directly with velocity.^21,41 The shoulder, by contrast, is a highly mobile ball-and-socket joint that relies on coordinated trunk and scapular motion, rotator cuff control, and precise sequencing to maintain stability under extreme rotational velocities.^46,47 Achieving elite pitch shape and movement, the components rewarded by Stuff+, may require complex mechanical adjustments that amplify cumulative stress across these stabilizing structures. These data also suggest that overlaps may exist in risk factors or maladaptive behaviors preceding shoulder and elbow injury, and targeted interventions may lead to concurrent injury risk reduction in both throwing extremity joints.

This distinction carries important implications. Stuff+ quantifies not only how fast a pitch is thrown but how unusually it moves, by rewarding outlier spin axes, seam-shifted wake effects, and deceptive release orientations.²⁷ The modern developmental focus on maximizing “stuff” thus represents a biomechanical challenge as much as a performance goal. While these adaptations enhance competitiveness, our findings suggest that pursuing increasingly complex pitch characteristics may synergistically impose higher mechanical and neuromuscular demands on the shoulder. A 2024 comprehensive MLB report on pitcher injuries—developed in collaboration with team physicians, athletic trainers, orthopaedists, and biomechanists—echoed these concerns, highlighting that the league-wide emphasis on developing greater so-called stuff has likely contributed to the persistent rise in pitcher injuries.³³ Collectively, these findings support the idea that the same pitch characteristics that enhance effectiveness may also introduce mechanical instability at the shoulder, underscoring the need for studies integrating performance analytics with biomechanical monitoring to establish safer limits of optimization.

The finding of a more medial release position in pitchers who later required capsulolabral surgery identifies a visible kinematic difference, but its underlying meaning remains uncertain. Because horizontal release can be affected by where a pitcher stands on the rubber, stride direction, and trunk or arm orientation,³⁸ this observation likely reflects an interplay of multiple mechanical factors rather than a single motion fault. Prior research has also produced conflicting results regarding its relationship to injury: some studies have linked a more medial release to increased UCL risk,^24,48 whereas others have shown either no association,²⁴ or an opposite trend toward greater risk with more lateral release points.³⁸ The lack of consistency across studies suggests that horizontal release position is a proxy capturing complex whole-body motion patterns, and that its contribution to injury risk likely depends on how it interacts with other mechanical variables.

One feature distinguishing this analysis from prior work was the inclusion of Statcast's arm-angle metric, a recently introduced anatomically grounded measure of throwing orientation.⁴⁰ While acute UCL injuries have been linked to abrupt drops in arm angle,⁹ and chronic injuries to greater pitch-to-pitch variability,²⁶ this study found no between-group differences in this measure. The combination of a more medial release point but similar arm angles between groups suggests that the difference arose not from changes in arm slot itself, but from altered trunk mechanics, particularly contralateral trunk tilt. When a pitcher leans the torso toward the glove side during delivery, the entire shoulder complex shifts toward the midline,^22,35 moving the release point medially without necessarily changing arm angle. This pattern has become common in velocity-focused training, as increased trunk lean has been associated with higher pitch speeds, but it has also been shown to elevate shoulder and elbow loading forces and association with injury risk.^{3,17,35,43,49,50} The interplay between trunk tilt and arm slot therefore provides a more comprehensive framework for interpreting these findings. Interestingly, Escamilla et al,¹¹ found that pitchers with moderate trunk tilt experienced the greatest shoulder and elbow torques, even exceeding those of extreme overhand or sidearm deliveries. This reinforces that isolated kinematic variables, such as release position or arm angle alone, cannot reliably explain injury risk without considering their combined mechanical context. The present results should thus be viewed as evidence that certain trunk-arm coordination patterns may manifest as measurable release differences, but the precise causal pathways remain unclear. Future studies integrating motion capture biomechanics with large-scale tracking data and injury histories are needed to clarify how these variables interact to influence shoulder stress and to determine whether observable changes in release position can serve as early indicators of pathologic adaptation.

Notably, several variables investigated showed no association with capsulolabral injury risk despite receiving substantial attention in injury prevention discourse. Pitch count showed no difference between groups, nor did we detect differences in release extension, approach angles, or pitch movement patterns. The absence of workload effects does not diminish the importance of load management; substantial evidence supports pitch count restrictions in reducing injury risk broadly, especially for young athletes,^{18,20,25,34,37} but suggests that how a pitcher throws may matter as much as volume. Similarly, our longitudinal analysis revealed no significant differences in season-to-season changes for most variables, with cases and controls showing similar trajectories in velocity, spin, and performance metrics leading up to surgery. This suggests that cross-sectional characteristics, what a pitcher throws, may be more predictive than rates of change; however, larger prospective studies with higher temporal resolution, particularly monitoring offseason training periods where dramatic improvements often occur, are needed to definitively address this question.

Practical applications emerge from these findings, as they arrive at a pivotal moment in baseball's evolution. While shoulder capsulolabral surgeries are still relatively uncommon compared with elbow UCL surgery, pitch tracking and other motion capture systems can play key roles in identifying trends indicative of or leading to injury. Modern training facilities employ high-speed cameras, force plates, and motion capture systems providing real-time feedback on how mechanical adjustments affect measurable metrics, enabling pitchers to experiment with different release points, grips, and body positions while observing immediate changes in pitch characteristics. Teams might implement enhanced monitoring for pitchers exhibiting elevated Stuff+ scores or excessive medialization of release position. Strength and conditioning programs might emphasize rotator cuff endurance, scapular stabilization, and posterior shoulder strengthening for high-risk populations. Most fundamentally, these findings suggest that injury prevention strategies must evolve alongside the analytical revolution transforming player development, moving beyond simple workload management toward comprehensive consideration of how biomechanical demands interact with individual tissue tolerance to determine injury risk. Whether targeted interventions can mitigate risk while preserving performance gains remains unknown and represents a critical avenue for future research.

Limitations

This study's findings must be interpreted within the context of its methodological constraints, which are inherent to its retrospective design and public data sources. A primary consideration is that the investigative framework, being retrospective in nature, does not permit the establishment of causal relationships between the identified mechanical factors and the subsequent incidence of injury. This limitation is compounded by our reliance on publicly available injury reports, which, while providing a valuable population-level data set, introduces several specific challenges. The timing of injury onset was necessarily approximated from these reports and may not precisely coincide with the actual symptomatic onset experienced by the athlete. Similarly, this also made it difficult to discern whether cases had more insidious and atraumatic onset of symptoms or if there was an acute traumatic injury. We also assume that our matched controls are asymptomatic and uninjured based off the lack of injury reports confirmed in the media. Furthermore, this data source did not afford access to granular surgical or clinical details, such as the specific location and severity of the labral tear, degree of bone loss, or the technical details of the procedures performed. It is also critical to recognize that the pitching metrics derived from Statcast data are exclusively representative of external kinematics. These data do not capture the internal joint loading, muscle activation patterns, or direct ligamentous strains that constitute the true pathomechanical pathway to injury. Consequently, the mechanical alterations we report should be understood as correlates of injury risk rather than as direct measurements of excessive tissue stress. Furthermore, our multivariable regression models were limited by a reduced sample size. The key composite metrics, Stuff+ and Location+, have only been publicly available since 2020, restricting the cohort for these model cases and controls with complete data. This smaller sample reduces the models' statistical power and widens their confidence intervals, and the findings should be interpreted with this constraint in mind. Given this study only included professional baseball players, the generalizability of our findings may be difficult to apply across all baseball levels where access to pitch tracking is not as obtainable. Finally, the absence of clinical information, such as reported levels of pain, fatigue, or functional limitations, precluded our ability to correlate these objective mechanical deviations with the subjective complaints that are foundational to clinical evaluation and management. While this gap currently restricts the direct translation of our findings into clinical practice, the identification of objective mechanical markers represents a critical preliminary step. Future research directions that successfully integrate these kinematic profiles with athlete-reported outcomes and data from wearable sensor technologies will be essential to bridge this divide and more directly link external mechanics with internal loading and symptomatic progression.

Conclusion

Pitchers who underwent capsulolabral surgery were more likely to throw harder with higher spin rates and a more medialized release point than matched controls. Overall, MLB pitchers who were more effective and had superior Stuff+ were at the highest risk for shoulder instability requiring surgical intervention. These findings suggest that the very analytical metrics now driving player development and evaluation throughout the MLB, and the physical capabilities required to achieve elite scores in performance metrics, may themselves represent risk factors for debilitating shoulder injury.

Footnotes

Final revision submitted December 28, 2025; accepted January 2, 2026.

One or more of the authors has declared the following potential conflict of interest or source of funding: C.S.A. is a paid consultant and receives royalties from Arthrex; he also serves as the head team physician of the New York Yankees. AOSSM checks author disclosures against the Open Payments Database (OPD). AOSSM has not conducted an independent investigation on the OPD and disclaims any liability or responsibility relating thereto.

Ethical approval for this study was waived by Columbia University Medical Center.

ORCID iDs

Michael A. Mastroianni

Morgan R. Dillon

Cole R. Morrissette

Nicholas Frappa

Andrew J. Luzzi

References

Agresta

Krieg

Freehill

MT.

Risk factors for baseball-related arm injuries: a systematic review. Orthop J Sports Med. 2019;7(2):2325967119825557. doi:10.1177/2325967119825557

Aguinaldo

Chambers

Correlation of throwing mechanics with elbow valgus load in adult baseball pitchers. Am J Sports Med. 2009;37(10):2043-2048. doi:10.1177/0363546509336721

Andrews

Why Is It Always the Year of the (Insert Pitch Here)? FanGraphs. Published on April 4, 2025. Accessed on December 28, 2025. https://blogs.fangraphs.com/why-is-it-always-the-year-of-the-insert-pitch-here/?utm_source=chatgpt.com

Andrews

Carson

Jr McLeod

WD.

Glenoid labrum tears related to the long head of the biceps. Am J Sports Med. 1985;13(5):337-341. doi:10.1177/036354658501300508

Bullock

Menon

Nicholson

Butler

Arden

Filbay

SR.

Baseball pitching biomechanics in relation to pain, injury, and surgery: a systematic review. J Sci Med Sport. 2021;24(1):13-20. doi:10.1016/j.jsams.2020.06.015

Camp

Tubbs

Fleisig

, et al. The relationship of throwing arm mechanics and elbow varus torque: within-subject variation for professional baseball pitchers across 82,000 throws. Am J Sports Med. 2017;45(13):3030-3035. doi:10.1177/0363546517719047

Cohen

Portney

Cohen

Bolia-Kavouklis

Weber

Saltzman

. Using pitch-tracking data to identify risk factors for medial ulnar collateral ligament reconstruction in Major League Baseball pitchers. Orthop J Sports Med. 2022;10(3):23259671211065756. doi:10.1177/23259671211065756

Conte

Requa

Garrick

JG.

Disability days in Major League Baseball. Am J Sports Med. 2001;29(4):431-436. doi:10.1177/03635465010290040801

Cooper

JJ.

Modern MLB Players Are Better Than Past Players—Here Are 5 Reasons Why. Baseball America. Published June 19, 2025. Accessed October 20, 2025. https://www.baseballamerica.com/stories/modern-mlb-players-are-better-than-past-players-here-are-5-reasons-why/#:~:text=Every%20year%2C%20pitchers%20throw%20harder,than%20two%20mph%20since%202008.

10.

Dillon

Mastroianni

Frappa

, et al. Using pitch-tracking metrics to identify warning signs immediately prior to acute ulnar collateral ligament injuries in Major League Baseball players. Orthop J Sports Med. 2025;13(12):23259671251389226.

11.

Escamilla

Slowik

Diffendaffer

Fleisig

GS.

Differences among overhand, 3-quarter, and sidearm pitching biomechanics in professional baseball players. J Appl Biomech. 2018;34(5):377-385. doi:10.1123/jab.2017-0211

12.

FanGraphs RosterResource. Injury Report. FanGraphs. Accessed August 10, 2025. https://www.fangraphs.com/roster-resource/injury-report

13.

Fleisig

Andrews

Dillman

Escamilla

RF.

Kinetics of baseball pitching with implications about injury mechanisms. Am J Sports Med. 1995;23(2):233-239. doi:10.1177/036354659502300218

14.

Fortenbaugh

Fleisig

Andrews

JR.

Baseball pitching biomechanics in relation to injury risk and performance. Sports Health. 2009;1(4):314-320. doi:10.1177/1941738109338546

15.

Frey

Kriz

Sciascia

Mathews

Bowman

Epidemiological trends of elbow and forearm injuries in high school baseball and softball players. JSES Rev Rep Tech. 2025;5(3):401-406. https://doi.org/10.1016/j.xrrt.2025.04.006

16.

Giberson-Chen

Shaw

Rudisill

, et al. Return to play after shoulder surgery in professional baseball players: a systematic review and meta-analysis. Orthop J Sports Med. 2023;11(1):23259671221140853. doi:10.1177/23259671221140853

17.

Harris

Frank

Jordan

, et al. Return to sport following shoulder surgery in the elite pitcher: a systematic review. Sports Health. 2013;5(4):367-376. doi:10.1177/1941738113482673

18.

Huang

Learman

Tsai

YS.

A comparison of throwing kinematics between youth baseball players with and without a history of medial elbow pain. Chin J Physiol. 2010;53(3):160-166. doi:10.4077/cjp.2010.amk026

19.

Hutchinson

Ireland

ML.

Overuse and throwing injuries in the skeletally immature athlete. Instr Course Lect. 2003;52:25-36.

20.

Keeley

Wicke

Alford

Oliver

GD.

Biomechanical analysis of forearm pronation and its relationship to ball movement for the two-seam and four-seam fastball pitches. J Strength Cond Res. 2010;24(9):2366-2371. doi:10.1519/JSC.0b013e3181b22aea

21.

LaPrade

Cinque

Safran

Freehill

Wulf

LaPrade

RF.

Using advanced data to analyze the impact of injury on performance of Major League Baseball pitchers: a narrative review. Orthop J Sports Med. 2022;10(7):23259671221111169. doi:10.1177/23259671221111169

22.

Lyman

Fleisig

Andrews

Osinski

. Effect of pitch type, pitch count, and pitching mechanics on risk of elbow and shoulder pain in youth baseball pitchers. Am J Sports Med. 2002;30(4):463-468. doi:10.1177/03635465020300040201

23.

Manzi

Estrada

Dowling

Ruzbarsky

Dines

JS.

Intra- versus inter-pitcher comparisons: associations of ball velocity with throwing-arm kinetics in professional baseball pitchers. J Shoulder Elbow Surg. 2021;30(11):2596-2603. doi:10.1016/j.jse.2021.04.017

24.

Manzi

Ruzbarsky

Krichevsky

, et al. Kinematic and kinetic comparisons of arm slot position between high school and professional pitchers. Orthop J Sports Med. 2023;11(10):23259671221147874. doi:10.1177/23259671221147874

25.

Mastroianni

Dillon

Frappa

, et al. Pitch-tracking risk factors and warning signs for ulnar collateral ligament injuries in Major League Baseball pitchers. Am J Sports Med. Published online November 6, 2025.

26.

Mastroianni

Kunes

El-Najjar

, et al. Advanced analytic and pitch-tracking metrics associated with UCL surgery in Major League Baseball pitchers: a case-control study. Orthop J Sports Med. 2025;13(2):23259671241302432. doi:10.1177/23259671241302432

27.

Mastroianni

Kunes

Mueller

, et al. Pitch-specific advanced analytic and pitch-tracking risk factors for ulnar collateral ligament injuries in Major League Baseball pitchers. Am J Sports Med. 2025;53(6):1440-1449. doi:10.1177/03635465251330564

28.

McGrattan

Stuff+, Location+, and Pitching+ Primer. Sabermetrics Library. Published 2023. Accessed October 19, 2025. https://library.fangraphs.com/pitching/stuff-location-and-pitching-primer/

29.

Mehta

Tang

Rajapakse

Juzwak

Dowling

Chronic workload, subjective arm health, and throwing injury in high school baseball players: 3-year retrospective pilot study. Sports Health. 2022;14(1):119-126. doi:10.1177/19417381211055142

30.

Meldau

Srivastava

Okoroha

Ahmad

Moutzouros

Makhni

EC.

Cost analysis of Tommy John surgery for Major League Baseball teams. J Shoulder Elbow Surg. 2020;29(1):121-125. doi:10.1016/j.jse.2019.07.019

31.

Mizels

Erickson

Chalmers

. Current state of data and analytics research in baseball. Curr Rev Musculoskelet Med. 2022;15(4):283-290. doi:10.1007/s12178-022-09763-6

32.

Nicholson

Hulburt

Beck

Waterman

Bullock

GS.

The relationship between pitch velocity and shoulder distraction force and elbow valgus torque in collegiate and high school pitchers. J Shoulder Elbow Surg. 2020;29(12):2661-2667. doi:10.1016/j.jse.2020.04.046

33.

Norton

Honstad

Joshi

Silvis

Chinchilli

Dhawan

Risk factors for elbow and shoulder injuries in adolescent baseball players: a systematic review. Am J Sports Med. 2019;47(4):982-990. doi:10.1177/0363546518760573

34.

Oeding

Boos

Kalk

, et al. Pitch-tracking metrics as a predictor of future shoulder and elbow injuries in Major League Baseball pitchers: a machine-learning and game-theory based analysis. Orthop J Sports Med. 2024;12(8):23259671241264260. doi:10.1177/23259671241264260

35.

Office of the Commissioner of Baseball. Memorandum: MLB Report on Pitcher Injuries. Major League Baseball. Published December 17, 2024. Accessed November 19, 2025. https://www.mlb.com/news/mlb-releases-report-on-pitcher-injuries-2024

36.

Olsen

2nd Fleisig

Dun

Loftice

Andrews

JR.

Risk factors for shoulder and elbow injuries in adolescent baseball pitchers. Am J Sports Med. 2006;34(6):905-912. doi:10.1177/0363546505284188

37.

Oyama

Blackburn

Padua

Myers

JB.

Effect of excessive contralateral trunk tilt on pitching biomechanics and performance in high school baseball pitchers. Am J Sports Med. 2013;41(10):2430-2438. doi:10.1177/0363546513496547

38.

Park

Loebenberg

Rokito

Zuckerman

JD.

The shoulder in baseball pitching: biomechanics and related injuries—part 1. Bull Hosp Jt Dis. 2002;61(1-2):68-79.

39.

Petty

Andrews

Fleisig

Cain

EL.

Ulnar collateral ligament reconstruction in high school baseball players: clinical results and injury risk factors. Am J Sports Med. 2004;32(5):1158-1164. doi:10.1177/0363546503262166

40.

Platt

Zacharias

Conley

, et al. Association between pitch break on the 4-seam fastball and slider and shoulder injury in Major League Baseball pitchers: a case-control study. Orthop J Sports Med. 2021;9(10):23259671211038961. doi:10.1177/23259671211038961

41.

Portney

Buchler

Lazaroff

Gryzlo

Saltzman

MD.

Influence of pitching release location on ulnar collateral ligament reconstruction risk among Major League Baseball pitchers. Orthop J Sports Med. 2019;7(2):2325967119826540. doi:10.1177/2325967119826540

42.

Sabesan

Chatha

Guo

Whaley

Amador

Lavin

Performance metrics and economics of superior labrum anterior-posterior tear repairs in Major League Baseball pitchers. JSES Int. 2022;6(4):569-572. doi:10.1016/j.jseint.2022.03.003

43.

Savant

(2025). Pitcher Arm Angle Leaderboard. Major League Baseball. Accessed September 13, 2025. https://baseballsavant.mlb.com/leaderboard/pitcher-arm-angles

44.

Slowik

Aune

Diffendaffer

Cain

Dugas

Fleisig

GS.

Fastball velocity and elbow-varus torque in professional baseball pitchers. J Athl Train. 2019;54(3):296-301. doi:10.4085/1062-6050-558-17

45.

Smith

Lombardo

Petersen-Fitts

, et al. Return to play and prior performance in Major League Baseball pitchers after repair of superior labral anterior-posterior tears. Orthop J Sports Med. 2016;4(12):2325967116675822. doi:10.1177/2325967116675822

46.

Solomito

Garibay

Nissen

CW.

Sagittal plane trunk tilt is associated with upper extremity joint moments and ball velocity in collegiate baseball pitchers. Orthop J Sports Med. 2018;6(10):2325967118800240. doi:10.1177/2325967118800240

47.

Spotrac. MLB Injured List Tracker: Summary. Injured List Tracker. Accessed August 10, 2025. https://www.spotrac.com/mlb/injured/_/year/2025/view/summary

48.

Stodden

Fleisig

McLean

Andrews

JR.

Relationship of biomechanical factors to baseball pitching velocity: within pitcher variation. J Appl Biomech. 2005;21(1):44-56. doi:10.1123/jab.21.1.44

49.

Werner

Gill

Murray

Cook

Hawkins

RJ.

Relationships between throwing mechanics and shoulder distraction in professional baseball pitchers. Am J Sports Med. 2001;29(3):354-358. doi:10.1177/03635465010290031701

50.

Werner

Guido

Jr Stewart

McNeice

VanDyke

Jones

DG.

Relationships between throwing mechanics and shoulder distraction in collegiate baseball pitchers. J Shoulder Elbow Surg. 2007;16(1):37-42. doi:10.1016/j.jse.2006.05.007

Advanced Analytic and Pitch-Tracking Risk Factors for Capsulolabral Injuries in Major League Baseball Pitchers

Abstract

Background:

Purpose/Hypothesis:

Study Design:

Methods:

Results:

Conclusion:

Keywords

Methods

Study Design and Ethical Considerations

Case Identification and Cohort Assembly

Control Identification and Cohort Assembly

Data Sources and Collection

Statistical Analysis

Statistical Analysis and Software

Results

Demographics and Baseline Characteristics

Performance Metrics and Advanced Analytics

Pitch-Tracking Characteristics

Longitudinal Preinjury Trends

Preliminary Logistic Regression

Multivariable Logistic Regression

Discussion

Limitations

Conclusion

Footnotes

ORCID iDs

References