Athletes often train without clear, objective feedback. Coaching can become overly dependent on personal interpretation. Teams routinely invest in prospects whose measurable tools don't consistently translate to game performance. Today, there is still no universal way to quantify true athletic capacity.
The Problem
- No unified metric for total athletic signal
- Existing tools are fragmented (motion capture, ball-flight data, stat lines, subjective evaluation)
- Development pathways can be inconsistent and influenced by subjective factors
- Non-traditional movers are routinely undervalued despite measurable upside
The Gap
- No standardized score that compares athletes across positions, sports, or eras
- No way to separate pure capacity (signal) from environmental factors (field)
- No clear mapping from mechanics → output ceiling, injury risk, or developmental upside
UBKT is the first attempt to formalize the physics of athletic signal into a single score.
Quantifying the Signal
UBKT measures rotational movement through five fundamental biomechanical metrics.
SEI
Sequencing Efficiency Index
"How efficiently does energy transfer through your kinetic chain?"
Measures how efficiently the body's kinetic chain transfers energy over time. In the proxy tier, SEI is computed from K/BB ratio with logistic normalization — capturing sequencing quality without outcome noise. For elite proxy seasons, SEI may saturate near 100 (proxy ceiling, not a claim of perfect sequencing). Logistic normalization ensures meaningful separation between good, elite, and exceptional athletes.
ASC
Axis Stability Correcting Coefficient
"How stable is your posture under rotation?"
Tracks deviation of spine angle and head position throughout delivery. In the proxy tier, ASC is computed as a weighted blend of BB/9 (60%) and WHIP (40%) with logistic normalization — capturing both pure command and contextual traffic management. Higher scores indicate better command potential and reduced injury risk.
RDR
Rotational Distribution Ratio
"Is your power rotational or linear?"
Computes the ratio of rotational power to linear momentum. Identifies mechanical style and predicts spin generation capability. Used as an assessment metric for style profiling only — not included in BSS. Linear athletes are not penalized for being linear.
PA
Perceptual Advantage
"How difficult are you to read?"
Measures the cognitive and strategic skills that allow a pitcher to maximize the effectiveness of their raw physical output. Scored using the official PA Rubric: five components (Game Plan, Hitter Reading, Deception/Tunneling, Count Leverage, Tempo/Composure), each scored 0.0–1.0. PA amplifies your engine — it never penalizes.
DM
Durability Multiplier
"How durable and reliable is your workload?"
A workload reliability multiplier relative to a 6 IP/start benchmark. In the proxy tier, DM is computed as innings-per-start durability adjusted by an HR/9 risk modifier — high HR/9 reduces the reliability score. Values above 1.0 indicate elite durability. DM labels: ≥1.00 = Elite Durability, 0.90–0.99 = Highly Repeatable, 0.75–0.89 = Consistent, <0.75 = High Variance.
BSS & SSQ
The five metrics combine into two headline scores: a universal mechanical rating and a sport-specific performance score.
Base Signal Score
BSS (0-100)Universal, cross-sport rating of core mechanical quality. Represents the pure quality of the movement engine, independent of sport context.
BSS = (SEI^0.5 × ASC^0.5) × 100Sport Signal Score
SSQ (0-100)Sport-specific performance score. Contextualizes BSS by incorporating Perceptual Advantage—for pitching, this is POW. PA is evaluated using the official PA Rubric.
SSQ = BSS × (1 + k × PA), where k = 0.25Understanding UBKT Scores: The Two-Tier Model
To ensure clarity and accuracy, UBKT scores are presented in two distinct tiers:
Generated using publicly available statistics (K/9, BB/9, WHIP, velocity, spin rate) as proxies for underlying biomechanical qualities. Reproducible and data-driven. For most elite MLB pitchers, Proxy BSS values typically fall within the 80–100 range. This score represents the floor of a pitcher's capabilities.
A high-fidelity assessment requiring motion capture analysis, high-speed video, and direct perceptual evaluation using the official PA Rubric. Captures the nuanced kinetic and strategic details that public stats cannot see. Represents the ceiling of a pitcher's capabilities.
View the PA Rubric →This two-tier model is not a limitation — it is a core feature. It honestly reflects the difference between what can be estimated from a distance and what can be known through direct, high-fidelity analysis.
Velocity Models (As Related to Pitching)
While the current framework focuses on the Signal, the long-term vision includes quantifying the Field—team context, opposition, and situational pressure.
Performance ≈ SSQ × FSQWhy Baseball as the Initial Testbed
Rotational & Measurable
Pitching isolates rotation in a controlled, repeatable motion — ideal for testing whether these metrics actually work.
Existing Data
Velocity, spin, command — the outputs already exist. UBKT asks whether we can trace them back to a unified input.
No Unified Framework
Lots of data streams exist. None of them talk to each other. UBKT is an attempt to connect them.
Real Stakes
Velocity, health, and development matter to athletes and teams. If UBKT works, it's useful. If not, we learn why.
Pitching is the testbed. If the framework holds here, it might generalize. That's the experiment.
How UBKT Differs
UBKT doesn't replace existing tools—it synthesizes their outputs into a single, interpretable framework.
| Parity | Traditional Coaching Focus on Technique and Output | Modern Data Labs Sports Tech | UBKT Signal & Field Mechanics |
|---|---|---|---|
| Approach | Primarily subjective evaluation | 40+ individual metrics | 5 core metrics → 1 unified score |
| Interpretation | Depends heavily on coach experience | Requires expert analysis | Designed to be interpretable by any athlete, coach, or scout |
| Consistency | Varies across evaluators | Highly accurate but often overwhelming | Standardized across athletes and contexts |
| Scalability | Limited by personnel and time | Data often siloed across systems | Built to integrate with existing tools and expand over time |
| Bias | Personality & logistics | Technology-based blind spots | Physics-based, neutral on physical traits |
Traditional
Primarily subjective, varies across evaluators, limited by personnel and time
Modern Labs
Highly accurate but often overwhelming, data siloed across systems
UBKT
Unified score, interpretable by anyone, neutral on physical traits
Current Status
UBKT V1.5 is a working hypothesis. Here's where it stands.
Framework
The mathematical model is defined. Five metrics, two scores, geometric weighting. V1.5 is complete and documented.
Calibrate
Currently benchmarking against known reference cases to verify scaling behavior and model coherence. Six elite MLB pitchers calibrated so far.
Iteration
Refining based on what the data shows. The framework will evolve as empirical testing and accurate data reveals what works and what needs adjustment.
This is an open research project. The goal is to test whether these principles hold up against real data.
The SFM Taxonomy
SFM operates on three levels: the mechanics of movement, the analysis layer, and the speculative frontier.
This taxonomy organizes the broader Signal & Field landscape behind UBKT. V1.5 focuses on the Primary Field—measurable biomechanics and kinetic outputs. Complete the sequence. 5 / 4 / 3.
Frequently Asked Questions
If You're Interested in Movement Measurement
This is an open project. If the framework interests you, feedback is welcome.
Get in Touch
Questions, feedback, or just curious about the project? Reach out.
Stay Informed
Follow the project
Get occasional updates on UBKT development, calibration progress, and any new findings.
Ask a Question
Feedback welcome
Have a question about the framework, the math, or how this might apply to your situation? Happy to discuss.
UBKT is a solo research project exploring whether athletic capacity can be measured through a unified framework. The work is ongoing.
Questions and feedback are welcome.