📚 Upside Studies: (1) NBA Study: Technical Fouls (2)MLB Study: In-Game Markerless & Lab Marker-Based Pitching Biomechanics (3) NHL Study: Determinants of Skating Speed In Ice Hockey Athletes
Upside Study #1: Strategic Impact of Technical Fouls in Basketball: A Two-Decade Analysis of Momentum Shifts in the NBA (2000–2021)
Introduction
In basketball, momentum is one of the most debated and least understood concepts. Often used to describe emotional surges, scoring runs, or shifts in body language and crowd energy, momentum is a psychological force that coaches try to create, maintain, or disrupt during high-stakes games. This study by Assaf Lev and colleagues offers a landmark empirical analysis of how one specific type of event—technical fouls (TFs) committed by coaches—impacts momentum and game outcomes in the NBA over a 21-year period (2000–2021).
Analyzing 4,196 technical fouls issued to NBA coaches, the researchers focused on how game context (home vs. away, score differential, game quarter) influenced the likelihood of a team shifting momentum and ultimately winning the game. The result is one of the most nuanced, data-rich explorations to date of how coaching behavior interacts with psychological momentum in professional sport.
Authors:
Assaf Lev1 ,
Yaniv Kanat Maymon2 ,
Tomer Ben Zion2 ,
Gershon Tenenbaum2,3
You can download the full PDF study by clicking on the button below:
Key Findings
1. Strategic Use of TFs Can Boost Win Probability—But Only in Certain Contexts
The study's most striking finding: context matters greatly. Coaches who were leading at the time of the TF saw their teams win 70% of the time, compared to only 22% when trailing. This confirms that TFs amplify existing momentum more effectively than they reverse it.
When committed at home, TFs led to a 44% win rate, but when committed on the road, the rate dropped to 22%. This sharp contrast reflects the power of home-court factors—familiarity, crowd support, and referee behavior—in amplifying emotional events.
2. Quarter-by-Quarter Effects Reveal Declining TF Effectiveness Over Time
The study found clear temporal patterns:
Home + trailing: win probability drops from 41% (Q1) to 10% (Q4).
Home + leading: win probability increases from 62% (Q1) to 84% (Q4).
Away + trailing: win probability drops from 33% (Q1) to 6% (Q4).
Away + leading: win probability increases from 44% (Q1) to 79% (Q4).
This suggests that early-game TFs are more likely to spark momentum changes, especially when a team is trailing. In contrast, late-game TFs are best used to consolidate a lead, particularly at home. In Q4, when time is limited and pressure is high, the psychological and tactical space for momentum shifts narrows dramatically.
3. Psychological Momentum and Collective Efficacy Play a Critical Role
The study highlights how TFs can trigger emotional contagion—players mirroring the energy and arousal of their coach. This effect is amplified in home games where the crowd acts as an emotional multiplier, motivating players to respond with increased aggression and focus.
For example, following a TF at home:
Field goal percentage improved.
Defensive rebounds increased.
Players showed heightened legal aggressiveness, suggesting a temporary increase in confidence and urgency.
These effects typically last 3–5 possessions, indicating a short-term spike rather than a sustained advantage.
4. TFs Can Serve Tactical and Symbolic Purposes
Beyond statistics, the study emphasizes the symbolic and strategic dimensions of technical fouls. Coaches may use them to:
Interrupt an opponent’s rhythm
Refocus their own team
Influence referee decisions
Signal dissatisfaction or urgency
A TF becomes a “performance cue”—not just a penalty, but a calculated risk meant to shift the emotional tone of the game. This reflects growing awareness among elite coaches of the psychosocial dynamics at play in elite performance.
5. Home Advantage Is Psychological, Not Just Physical
The research affirms that home teams maintain a statistically significant advantage, and that TFs are more effective when committed at home. This is tied to:
Crowd support and noise levels
Referee bias (conscious or subconscious)
Familiarity with the court and routines
Player motivation and social pressure
Interestingly, even when a team is less skilled, they perform better at home than equally skilled teams on the road, highlighting how psychological states—confidence, aggression, collective belief—are influenced by the environment.
Limitations and Future Research
While the study provides compelling quantitative insights, it acknowledges certain limitations:
Does not include playoff games, where pressure is higher and strategies may shift.
Focuses on male NBA players; further research is needed for the WNBA and international leagues.
Lacks qualitative data, such as player or coach interviews, that could provide insight into intent, team dynamics, or perceived effectiveness.
Future research could explore:
TFs during playoffs or rivalry games
Cross-cultural dynamics (e.g., EuroLeague, FIBA)
Team-specific cultures or coaching styles
The intersection between analytics and coaching intuition
Conclusion
This study provides a sophisticated, data-backed perspective on a previously under-analyzed aspect of basketball: how technical fouls, when used strategically, can influence game dynamics, team behavior, and win probability. Far from being impulsive lapses, TFs—when used wisely—can serve as powerful tools for shifting momentum, especially when deployed early in games, at home, and in advantageous situations.
The findings also highlight the nuanced interplay between emotional leadership, team psychology, and environmental context in determining performance outcomes. Coaches who understand these dynamics can better time their interventions, whether through substitutions, timeouts, or technical fouls.
Ultimately, the research confirms what seasoned coaches have long suspected: the most effective leaders know when to break the rules—strategically.
Upside Study #2: Comparing In-Game Markerless and Laboratory Marker-Based Pitching Biomechanics in Baseball
Introduction
Over the last decade, the rapid evolution of motion capture technologies has transformed how biomechanics are measured, interpreted, and applied in elite sports. In baseball—where the act of pitching is both biomechanically complex and performance-critical—understanding how different data capture methods influence the interpretation of mechanics is a vital question.
Traditionally, pitching mechanics have been measured in controlled laboratory settings using marker-based motion capture systems. These systems offer high spatial and temporal resolution, making them ideal for precise kinematic analysis. However, they are limited by artificiality: pitchers perform under constrained, often decontextualized conditions without the adrenaline, crowd, or opponent pressure of live competition.
To address this limitation, markerless motion capture systems like KinaTrax have emerged, offering the ability to analyze biomechanics in real-time, in-game settings. Yet these systems bring new challenges, especially in terms of data variability, accuracy, and standardization. Are they precise enough? Do mechanics change under game conditions? Can their data be trusted to inform return-to-throw protocols, pitch design, or performance interventions?
The present study by Lerch et al. (2025) is one of the first to directly compare within-pitcher variability of key kinematic variables between in-game markerless systems and in-lab marker-based systems, providing important guidance for both researchers and practitioners.
Authors:
Benjamin G. Lerch a ,
Glenn S. Fleisig b ,
Jonathan S. Slowik b ,
Gretchen D. Oliver a,
Accepted 20 May 2025 Journal of Biomechanics 188 (2025)
Available online 21 May 2025
You can download the full PDF study by clicking on the button below:
Study Design and Methodology
Participants
30 NCAA Division I pitchers analyzed using markerless KinaTrax system during live games
30 collegiate-level pitchers analyzed in a marker-based laboratory setting
Each pitcher threw 10 fastballs, allowing for intra-athlete variability comparisons. No crossover was performed (i.e., pitchers were not tested in both settings), which is a known limitation.
Measurement Tools
In-game: KinaTrax markerless motion capture system installed in stadiums
Lab: Multi-camera optical motion capture using reflective markers placed at anatomical landmarks
Variables Assessed (10 Kinematic Parameters)
Stride length (% of body height)
Foot placement offset
Knee flexion at foot contact
Trunk forward tilt
Trunk side tilt
Shoulder abduction
Shoulder external rotation (max)
Elbow flexion at max external rotation
Elbow extension velocity
Shoulder rotation velocity
Each variable was analyzed for both:
Mean values across pitches
Within-subject variability (standard deviation across 10 pitches)
Key Results and Interpretations
⚾ Pitch Velocity
In-Game: 91.2 ± 3.8 mph
In-Lab: 85.2 ± 1.5 mph
➤ Statistically significant (p < 0.001)
This ~6 mph difference is consistent with the literature: pitchers throw harder in competitive settings, influenced by adrenaline, crowd energy, and competitive intent. It validates the ecological relevance of in-game data but also reinforces the need to adjust context when comparing lab-derived mechanical baselines.
📏 Kinematic Differences (Mean Values)
Significant differences in 7 out of 10 variables:
Stride Length: Greater in games (88.9% vs. 82.5% of body height, p < 0.001)
Foot Placement Offset: Less deviation in games (5.8 cm vs. 18.8 cm, p < 0.001)
Shoulder External Rotation: Increased in-game (182.9° vs. 163.7°, p < 0.001)
Elbow Flexion: Lower in-game (78.1° vs. 101.1°, p < 0.001)
Shoulder Abduction, Trunk Side Tilt, Trunk Forward Tilt: All showed subtle but meaningful differences
Interpretation: These shifts likely reflect both performance intensity (e.g., higher intent = greater joint loading) and technological variation. Markerless systems might calculate joint centers or segment angles slightly differently than lab systems, which is important when aggregating multi-source biomechanical data.
📊 Within-Subject Variability
Only 2 of 10 variables showed statistically significant higher variability in the in-game (markerless) group:
Shoulder External Rotation (p = 0.001)
Elbow Flexion at Max External Rotation (p < 0.001)
All other variables—including trunk tilt, stride, shoulder abduction, and elbow velocity—showed no significant differences in variability between lab and game settings.
➤ This challenges the common belief that in-game biomechanics are inherently “noisier” or less consistent. Despite environmental stressors (game pressure, crowd, fatigue), elite pitchers largely maintain biomechanical consistency, reinforcing the robustness of learned motor patterns in highly trained athletes.
Broader Implications
🔬 Markerless Tech is More Reliable Than Previously Assumed
The fact that biomechanical variability was largely equivalent between lab and game settings supports the growing credibility of markerless motion capture. While not a perfect substitute for lab data, it provides actionable insights in ecologically valid conditions—something practitioners have long desired.
⚠️ But Don't Mix Data Sources
Although variability was similar, mean values were different, meaning you should not directly compare or combine data from markerless and marker-based systems. Differences in calibration, joint modeling, and reference frames can distort interpretation if mixed.
⚙️ Performance Is Context-Specific
Pitching mechanics are context-dependent. Throwing in a lab vs. in front of 30,000 fans is not the same. Coaches and analysts must consider intent and environment when interpreting movement data, and avoid overfitting training interventions to “lab-normalized” values.
Limitations of the Study
Cohort mismatch: The two groups (in-game vs. in-lab) were made up of different pitchers, limiting within-subject comparison.
Environmental confounds: Pitch count, opponent quality, mound height, and psychological state could all impact in-game results.
Technology disparity: Even small discrepancies in marker placement or camera calibration between systems may affect angular measurements.
Future work should use within-subject crossover designs, where the same pitcher is captured in both environments, ideally using synchronized markerless and marker-based systems.
Conclusion
This pioneering study sheds light on a critical question in sports biomechanics: how reliable is in-game motion capture data compared to lab-based gold standards?
The answer, based on 10 key kinematic metrics and 60 collegiate pitchers, is clear: biomechanical variability is largely consistent across both settings, with only two variables (shoulder ER and elbow flexion) showing meaningful differences. At the same time, mean kinematic values differ significantly—meaning technologies and environments matter, and their influence should be respected when designing training or rehabilitation programs.
For sports scientists, biomechanists, and high-performance practitioners, these findings offer confidence that in-game markerless systems are valid tools, especially when interpreted through a lens of context and intent. They also underscore the need for system-specific baselines, greater standardization, and ongoing research to refine our understanding of how and why athletes move the way they do under pressure.
Upside Study #3: Determinants of Skating Speed in Ice Hockey Athletes
Introduction
Ice hockey is a sport defined by its high physical demands, requiring athletes to perform frequent, intense bursts of skating and rapid changes in direction. The ability to achieve and sustain high skating speeds is widely regarded as a key performance differentiator among players. During a typical National Hockey League (NHL) game, players may execute up to 113 high-intensity bouts, including an average of 19 sprint-like efforts at speeds approaching 28.6 km/h. Given these demands, understanding which off-ice physical qualities best predict on-ice skating speed is of great interest to coaches, sport scientists, and performance staff seeking to optimize training and talent identification. This systematic review, conducted according to PRISMA guidelines, critically examines 19 studies published since 1990 that investigate the relationship between off-ice physical tests and on-ice skating speed in ice hockey athletes aged 16 and older1.
Methods and Study Selection
The review encompassed a rigorous search across PubMed, SPORTDiscus, and Google Scholar, using targeted keywords related to ice hockey and skating speed. Out of an initial pool of 172,755 articles, only 19 studies met the strict inclusion criteria: primary, peer-reviewed research focusing exclusively on ice hockey players aged 16 or above, with a main focus on skating speed. Studies were further screened for quality using the STROBE checklist, with only those scoring above 75% retained for analysis. The final selection included a total of 1,386 subjects (average age 19.5 years), spanning male, female, junior, collegiate, professional, and national team athletes. Most studies were correlational in design, with a few employing longitudinal or crossover approaches1.
Authors:
Michael A. Silvestri,1,2
Daniel J. Cleather,1
Samuel Callaghan,1
John Perri,
Hayley S. Legg3
Accepted 30 September 2020
Published Online First 3 November 2020
You can download the full PDF study by clicking on the button below:
Key Findings and Statistics
The review identified four primary categories of off-ice assessments linked to on-ice skating speed: sprinting, jumping, body composition, and anaerobic power. The relationships and predictive values of these tests are detailed below.
Off-Ice Sprinting
40-yard sprint: Demonstrated a significant correlation with maximum on-ice skating speed, explaining up to 25% of the variability in skating performance (p < 0.005, r = 0.51)1.
20-yard and 40-yard sprints: Significant predictors of maximum skating speed in male athletes (20-yard: p ≤ 0.05, r = 0.54; 40-yard: p ≤ 0.05, r = 0.62), but not in females1.
30-m sprint: Showed a strong relationship with 40-m on-ice skating sprint performance (p < 0.001, r = 0.80)1.
Pro-agility test: Correlated with on-ice change of direction speed (p < 0.05, r = 0.51)1.
Resisted sprints: Using 30 kg resistance for men and 15 kg for women, these were the best single predictors of on-ice sprint performance (men: p = 0.01, r = 0.74; women: p = 0.01, r = 0.70)1.
Jumping Ability
Vertical jump: Moderately to strongly correlated with various on-ice skating speed and acceleration measures (e.g., p < 0.05, r = 0.744; p = 0.025)1.
Standing long jump: Strongly correlated with on-ice speed, agility, and time trial performance for both forwards and defense (e.g., p < 0.05, r = -0.536; p = 0.016, r = -0.649; p < 0.001, r = -0.57)1.
Countermovement, depth drop, and squat jumps: All showed moderate to strong positive correlations with forward and backward skating speeds (e.g., backward skating: squat jump p < 0.01, r = 0.82)1.
Percentage drop jump: Moderately predictive of on-ice skating performance (p < 0.05, r = 0.375)1.
Jumping tests: Collectively, these reflect lower body power, a critical determinant of rapid acceleration and high skating velocity.
Body Composition
Lower body fat percentage: Weak to moderate association with faster skating times (e.g., 10-m sprint: p = 0.12, r = -0.36)1.
Lower body mass: Strongly correlated with better on-ice acceleration, especially in women (p = 0.03, r = 0.639)1.
Higher lean tissue mass: Associated with greater lower body explosiveness, which is linked to improved skating performance1.
Male vs. female differences: For women, lighter body weight is more relevant to acceleration over short distances, while for men, overall body composition (including fat-free mass) is more predictive of skating speed1.
Anaerobic Power
Wingate anaerobic power tests: Both peak and mean power outputs significantly correlated with on-ice acceleration, velocity, and sprint times (e.g., p < 0.01, r = -0.62 for sprint speed; p ≤ 0.05, r = -0.32 to -0.48 for acceleration and velocity)1.
Higher anaerobic power: Athletes with greater anaerobic power achieved faster acceleration and higher top skating speeds1.
Discussion and Practical Implications
The review highlights the multifactorial nature of skating speed in ice hockey, emphasizing that no single off-ice test can fully predict on-ice performance. Instead, a combination of tests—especially those assessing sprinting ability (including resisted sprints), jumping power (vertical and long jump), body composition, and anaerobic power—provides the most comprehensive evaluation of an athlete’s skating potential. The biomechanical similarities between sprinting and skating, particularly in terms of horizontal force development and muscle activation (notably the hamstrings and quadriceps), help explain the observed relationships1.
Change of direction ability, assessed through pro-agility tests, is also crucial given the unpredictable and multidirectional nature of ice hockey. Resisted sprinting tests offer a novel and more sport-specific assessment of acceleration, as the added resistance more closely mimics the postural and mechanical demands of the skating start1.
Jumping tests, particularly the standing long jump, are strongly linked to lower body power and are even predictive of draft status in elite contexts such as the NHL combine. Plyometric tests, such as depth drop and countermovement jumps, further capture the explosive qualities required for rapid changes in speed and direction on ice1.
Body composition remains an important, though not exclusive, predictor. Lower body fat and higher lean mass contribute to better skating outcomes, with some sex-specific nuances in predictive value1.
Anaerobic power, as measured by Wingate tests, is consistently associated with faster skating acceleration and higher top speeds, underscoring the importance of high-intensity energy system development in training1.
Limitations
The review notes significant methodological heterogeneity across studies, including differences in test protocols, distances, equipment, and subject populations. Most studies relied on correlational analysis, limiting the ability to draw causal inferences. There is a clear need for standardized, sport-specific testing protocols and more rigorous experimental designs to strengthen future research in this area1.
Conclusion
This systematic review confirms that skating speed in ice hockey is determined by a complex interplay of physical attributes, including off-ice sprinting ability, jumping power, body composition, and anaerobic capacity. While no single off-ice test can universally predict skating speed, a battery of assessments targeting these domains offers valuable insights for coaches and practitioners. The findings support the integration of resisted sprints, jumping tests, body composition analysis, and anaerobic power evaluations into athlete development and talent identification programs. However, the field would benefit from consensus on standardized testing methods and further research to refine predictive models for on-ice skating performance.
You may also like:
