Sesh 4.0: the biomechanics, sprint reserve, and continuous multi-session energetics of competitive swimming

The Evolution: Sesh 1.0 → 2.0 → 3.0 → 4.0

Swim training tracking historically oscillated between two extremes: simplistic metre-counting clipboards and cumbersome manual spreadsheets. Sesh was created to turn natural coaching shorthand into immediate, actionable physiological intelligence. The progression across four generations marks a generational transformation from static text parsing to continuous, multi-session sports science:

1. The Fundamental Problem with Counting Metres

Consider three distinct workouts swum by the same national-standard swimmer, each totalling exactly 1,000 metres of freestyle:

Traditional training logs record all three sessions identically as "1,000m Freestyle." This is a critical failure. Volume is merely the vehicle; intensity, interval density, and biochemical fatigue dictate the adaptation. Sesh 4.0 eliminates this blind spot by estimating the internal physiological and biomechanical demands of every single repeat.

2. The Sesh 4.0 Pipeline: From Whiteboard to Longitudinal Biological Load

A coach writes a session as they always have: on a poolside whiteboard, in a notes app, or using shorthand. From there, seven unified stages transform raw text or imagery into longitudinal biological intelligence:

3. The Decoupling Revolution: Speed vs. Cost

The core intellectual breakthrough of Sesh is the absolute separation of kinematic velocity from metabolic cost:

• Velocity is Hydrodynamic: Dictated by propulsive surface area, body alignment, turning mechanics, wave drag, and locomotion mode.
• Cost is Biochemical: Dictated by active muscle mass, mechanical efficiency, substrate phosphorylation (aerobic vs. glycolytic vs. phosphagen), and cardiovascular drift.

Systems that calculate workout intensity purely from swimming speed produce absurd conclusions. For example, kicking with a kickboard is substantially slower than full-stroke swimming because legs produce roughly 15% of total propulsion while creating large frontal drag. A speed-based tool assumes the swimmer is loafing. In reality, kicking isolates the largest muscle groups in the human body (quadriceps and gluteals) under sustained isometric core tension, producing immense localized metabolic demand.

In Sesh 4.0, locomotion mode velocities are calibrated directly against empirical competitive swimming kinematics: freestyle kick velocity ratio is anchored at $0.60$, backstroke kick at $0.68$, butterfly kick at $0.71$, breaststroke kick at $0.82$, and pulling at $0.89$ of whole-stroke speed. By modeling velocity projection and bioenergetic cost independently, Sesh 4.0 resolves the classic "kicking paradox" cleanly.

4. The Dual-Anchor Breakthrough: Anaerobic Speed Reserve (ASR)

A longstanding blind spot in endurance sports science is the over-extension of threshold models into sprint training. Classical swimming analytics rely on a single aerobic benchmark: Critical Swim Speed ($CSS$). While CSS accurately predicts aerobic capacity across 400m to 1,500m efforts, CSS possesses virtually zero correlation with pure sprint capabilities.

Two swimmers can possess an identical 400m CSS of 1:10.0/100m ($1.43\text{ m/s}$). However:

• Swimmer A (Pure Sprinter): Possesses a 50m Freestyle PB of 23.50s ($v_{\max} = 2.13\text{ m/s}$).
• Swimmer B (Distance Specialist): Possesses a 50m Freestyle PB of 27.50s ($v_{\max} = 1.82\text{ m/s}$).

If a coach writes a high-intensity set: 8 × 50m Max Effort @ 2:00, a single-anchor CSS model prescribes the exact same pace and calculates the exact same metabolic score for both swimmers. For Swimmer B, holding that pace requires 100% of their maximum neuromuscular capacity. For Swimmer A, it operates well below their sprint ceiling.

The Push-Start Block Invariant

A 50m competition PB is initiated with a dive start from the starting blocks, generating an entry breakout velocity of $3.0\text{ to }3.8\text{ m/s}$. During training repeats, however, swimmers push off the wall from a stationary start. Treating competition dive velocity as attainable in training push-offs creates unrealistic velocity targets.

Sesh 4.0 incorporates a calibrated push-start offset ($\sim 0.8\text{s}$ over 50m) for non-dive sprint sets. The sprint velocity ceiling is mathematically transformed to reflect wall push-off physics unless the set explicitly specifies dive starts (e.g., dive or off_block).

Supra-Critical Distance: Quantifying Anaerobic Exposure

To give coaches precise visibility into pure anaerobic work, Sesh 4.0 introduces Supra-Critical Distance:

$$\text{Supra-Critical Meters} = \sum_{\text{reps}} \max\left(0, d_{\text{rep}} - (v_{\text{CSS}} \cdot t_{\text{rep}})\right)$$

This isolates the exact distance swum at speeds faster than aerobic threshold, providing a pure, uncorrupted metric of anaerobic volume per workout, micro-cycle, and season.

5. Continuous Multi-Session Dynamics: The 2-Channel Carryover Engine

Swimmers live in an extreme training environment: competitive athletes regularly swim 6 to 10 pool sessions per week, frequently completing double-day workouts where an afternoon session begins only 4 to 6 hours after the morning session ended.

Legacy sports tracking models evaluate workouts in isolation or rely on discrete rolling daily sums (e.g., traditional daily TSS). This approach fails in swimming:

• Fatigue does not reset when the morning session ends.
• An afternoon session swum 4 hours after a threshold morning set demands a dramatically higher biological and psychological cost than the same session performed after 24 hours of rest.
• Conversely, returning to full 5,000m volume after a two-week layoff imposes an acute unadapted shock on muscle tissue and energy systems.

Compounding Double-Day Turnarounds

When a swimmer begins Session 2 on the same day, the engine computes the exact elapsed resting duration ($\Delta t$). If the interval between sessions is brief ($\Delta t < 8\text{ hours}$), acute residual fatigue remains significantly elevated. This residual strain compounds the biological difficulty of the afternoon session, reflecting the true systemic cost of back-to-back training days.

Chronic Work Capacity & Layoff Shocks ($b_i$)

To model training consistency without fragile rolling windows, Sesh 4.0 maintains a continuous 35-day Exponential Moving Average (EMA) of chronic work capacity ($b_i$). When a swimmer maintains consistent training density, $b_i$ remains robustly adapted.

However, if an athlete takes an uncharacteristic layoff ($\ge 7\text{ days}$ of inactivity), Sesh 4.0 detects the detraining gap and calculates a layoff shock penalty ($D_i$). When the swimmer returns to the pool, the initial workouts are evaluated with contextual awareness of their detrained state, protecting coaches against prescribing full acute training spikes to an unconditioned athlete.

Dual-Channel Reporting: Immutable sTSS vs. Contextual sTSS

A vital architectural breakthrough of Sesh 4.0 is the preservation of scientific truth:

• Raw Baseline sTSS (`load_stss`): An immutable, pure representation of the physical and metabolic work performed in the water. It is derived strictly from physics ($P_D \propto v^3$) and never modified by yesterday's fatigue or equipment.
• Contextual Load sTSS (`contextual_load_stss`): Reflects the athlete's internal cost of completing that session given their current readiness, double-day turnaround, and chronic conditioning base.

To prevent mathematical runaway, the contextual difficulty multiplier ($M$) is strictly bounded via a hyperbolic tangent envelope: $M \in [1.00, 1.30]$. Swimmers and coaches always see both numbers: the absolute physical load and the contextual internal cost.

6. Articular Loading & The "Paddle Paradox"

The most common medical issue in competitive swimming is "swimmer's shoulder": subacromial impingement, rotator cuff tendinopathy, and biceps tendinitis affecting between 40% and 80% of competitive athletes over their careers.

Historically, training logs either completely ignored equipment or treated hand paddles as a simple cardiovascular load multiplier. Recent biomechanical and energetic research exposes why this is flawed:

If an engine naively multiplies training load for paddles, it misrepresents metabolic conditioning. If it lowers training load, a 4,000m paddle pulling set registers as "easy recovery" right as the swimmer's shoulder tendons are being overloaded.

Sesh 4.0 solves this via Decoupled Dual-Book Accounting:

• Metabolic Track: Paddles on steady aerobic pull sets are calibrated with an efficiency-informed metabolic factor ($1.02$ paddles alone, $0.94$ with buoy), avoiding artificial cardiovascular load inflation.
• Articular Exposure Track: Local glenohumeral joint torque is tracked in an independent musculoskeletal channel ($1.35\times\text{--}1.50\times$ moment multiplier), generating an automated High Shoulder Strain Advisory whenever cumulative articular exposure exceeds 140 units.

7. Individual Stroke Profiling: The Individual Stroke Index (ISI)

In legacy training models, every swimmer was assumed to possess identical stroke proficiencies: Breaststroke was modelled as a fixed fraction of Freestyle velocity, with Backstroke and Butterfly assigned rigid population percentages.

In competitive swimming, this assumption is fundamentally flawed. A finalist in the 200m Breaststroke possesses vastly different stroke velocity ratios compared to a pure distance freestyler. Applying fixed population constants severely over-prescribes fatigue for the distance swimmer and severely under-trains the breaststroke specialist.

Where sufficient competition data is available, athlete profiles account for comparable event distances and course conditions, reducing distortions caused by differences in race format, turns, and starting conditions.

8. Resolving Data Sparsity: Cross-Event Machine Learning (Project M11)

A longstanding hurdle in swimming science is the Missing Distance Problem. Over 40% of competitive club athletes rarely or never contest the 400m Freestyle, focusing primarily on sprint programs (50m and 100m events). Without a verified distance anchor, classical two-point Critical Swim Speed models cannot be computed without arbitrary guesswork.

Sesh integrates MyAthlete's proprietary cross-event performance prediction technology (Project M11) to estimate missing race anchors. This allows athletes with incomplete competition histories to receive individualised training calculations without requiring laboratory testing or additional competitive events:

• Cross-Event Performance Transfer: Models non-linear performance relationships across different race distances and strokes.
• Dual-Anchor Imputation: Automatically estimates both the 400m aerobic threshold anchor and the 50m sprint velocity anchor from existing marks.
• Confidence-Aware Predictions: Adjusts calibration certainty based on the depth and breadth of each athlete's verified competition history.

9. Mathematical Safeguards & The Evidence Hierarchy

To ensure scientific transparency, all mathematical parameters and coefficients within Sesh 4.0 are classified into three distinct evidence tiers:

• Tier 1 — Fundamental Physics & Hydrodynamics: Cubic velocity drag scaling ($P_{\text{drag}} \propto v^3$), CSS velocity benchmarking ($v / v_{\text{CSS}}$), turning wall push-off impulses, and course conversions (SCM vs. LCM vs. SCY).
• Tier 2 — Peer-Reviewed Empirical Physiology: Empirical kick-to-swim velocity ratios across strokes (Bartolomeu 2018), pull velocity ratios (Morris 2016), net metabolic stroke costs (di Prampero, Zamparo 2020), 15-second sprint duration breakpoints (Soares 2014), and Banister two-compartment multi-session kinetics.
• Tier 3 — Principled Engineering Heuristics: Decoupled shoulder joint exposure indices, bounded contextual load multipliers via $\tanh$, and drag-resistance gear modifiers. These are explicitly documented as coaching decision-support proxies, not invasive clinical diagnostics.

10. The Unified Platform Advantage

MyAthlete's integrated platform connects competition performance, training activity, and athlete development within a unified environment. This allows information from different stages of an athlete's journey to contribute to more informed performance analysis and individualised training intelligence:

• Longitudinal Athlete Histories: Competition results directly inform daily training calibration, replacing generic squad assumptions with verified individual capabilities.
• Integrated Performance Intelligence: Training loads, workout density, and competitive performances sit within a single continuous framework rather than fragmented spreadsheets and disconnected tools.
• Coaching Continuity: Whiteboard sessions instantly bridge the gap between workout design on deck and long-term athlete development.