Markerless motion capture
Phone-camera approaches to sprint and landing analysis, under review.
Watched
The technology stack
Instruments first.Then intelligence.
Testing, your Performance Twin and daily programming, connected. I explain the inputs behind every decision.
The stack
Four layers inone loop.
Measure, model, deliver, verify: choose a layer to see what runs in it.
Laboratory-standard instruments, one day
- Force plates
- Force and rate of force development, reactive strength, left–right asymmetry.
- Velocity devices
- Bar sensors: velocity on every set, load–velocity profiles, velocity-loss thresholds.
- Timing gates and video
- 10, 20 and 30 m splits, flying 30 m, ground-contact time; sprints and landings filmed.
- Metabolic cart and lactate
- Metabolic testing for VO₂max, thresholds, economy and resting metabolic rate; lactate for the maximal lactate steady state.
- DXA and bioimpedance
- Lean mass, fat and bone density by region; phase angle between scans.
- Blood chemistry
- Clinical panels at partner laboratories, including ApoB and Lp(a), glucose and insulin, hormones, iron, thyroid, micronutrients, inflammation.
- Wearables
- HRV, resting heart rate and sleep from validated devices; GPS load in the field; trends against baselines.
- Genomics and epigenetics
- DNA and nutrigenomic profiling, where chosen and clinically appropriate, discussed with a qualified professional; an epigenetic clock only where chosen.
The Performance Twin
- Inputs
- Assessment, workload, recovery and your feedback; never pooled.
- Features
- Baselines and changes over time.
- Models
- A daily readiness summary and possible training plans for review.
- Tooling
- Built by Sainath; not borrowed software.
Daily, in your pocket
- Portal
- Today’s readiness, the session written from it, the protocol cards, your results.
- Direct access
- Sainath, not a support queue; weekly review, monthly recalibration.
- Welcome kit
- The EVOX Passport, the protocol cards and the first block, hand-packed.
- Handover pack
- What your coach, physiotherapist and physician receive, in a form they can use.
The re-test and its discipline
- Same conditions
- Same protocol, instruments, time of day and preparation at week twelve, so comparisons are more meaningful.
- Calibration
- Every testing day: plates zeroed against a known load, gates squared, carts calibrated with reference gas, the check recorded.
- Measurement uncertainty
- Every measurement has uncertainty; a small change may not be distinguishable from measurement error.
- Findings report
- Week three and week twelve: written, explained in a working session, the record the next block is written from.
The Performance Twin
A model of your bodythat explains itself.
I build the Performance Twin from your numbers and explain how each output is produced.
- One number each morning. The readiness score, from sleep, HRV, load and how you feel.
- A change to review. Changes in workload, asymmetry and recovery prompt review of the training plan. These signals do not diagnose or predict an individual injury.
- A plan compared before the commitment. A camp, a trip or a weight change is laid out against your own numbers first, to support a decision; it does not forecast the result.
- A record a clinician can read. Every alert carries its thresholds and weights, for a physician to overrule with evidence.
- Yours. Never pooled, never sold; deleted or handed over at your written instruction within thirty days.
Wearables
Trends over weeks.Never one night.
I use wearable trends alongside measured thresholds and your assessment.
| Device | What it gives | How it is read |
|---|---|---|
| Oura | Sleep staging, HRV, resting heart rate, temperature trend | Strong for sleep timing and HRV trends; staging is an estimate, read over weeks. |
| Whoop | HRV, strain, sleep, recovery score | Useful daily trend; the strain number is proprietary, so the model uses session RPE instead. |
| Garmin | Heart rate, HRV status, training load, GPS | The field athlete’s companion; heart-rate zones are re-anchored to measured thresholds. |
| Apple Watch and Apple Health | Heart rate, HRV, sleep, activity | Good enough for trends; exported to the model through Health where the client agrees. |
| Polar and chest straps | Beat-to-beat heart rate | The reference for HRV and for sessions where wrist optical sensors lag. |
| Continuous glucose monitors | Glucose every few minutes for a fortnight | Where it would change the plan; read against meals, sessions, travel and sleep. |
Device names identify products the practice reads from, not endorsements; data is exported only with your consent and stays in your model.
Accuracy
Every instrumenthas a noise floor.
Tests and wearables have measurement limits. I compare changes with that error before reporting progress.
- Same protocol, same time of day, same preparation at every re-test.
- Calibration before every test, the check recorded with the results.
- Measurement uncertainty stated for every marker in the report.
Next: adopted when validated, not before
Next
Adopted when validated.Not before.
Technologies under review, checked against reference methods before adoption:
Continuous lactate and ketone sensing
Continuous lactate and ketone sensors remain under review; reference testing stays in place.
Watched
Digital-twin simulation
Season-planning simulations remain under review; they are not forecasts of individual results.
Watched
Epigenetic clocks
Offered only where chosen, read as a trend, never as a diagnosis; the science is still maturing.
Where chosen
For founders, clubs and performance directors
Shaping a performance department’s data or AI strategy? Explore the advisory consultation.
About the stack: five questions, straight answers
About the stack
Straight answers,no jargon.
No. Testing is coordinated near you, at partner laboratories and clinics, billed at cost; a wearable you own is usually enough.
Never. Each model is one person’s numbers, never sold or aggregated, and deleted or handed over at your instruction within thirty days when you leave (Policies page).
Accuracy varies by device and measure. Sleep-stage estimates are interpreted cautiously, alongside trends and how you feel.
The model highlights changes in your inputs for coaching and medical review. It cannot reliably forecast when or where an individual injury will occur.
Once validated against a reference method and once it would change a decision; markerless motion capture and continuous lactate sensors are watched most closely.
Measurement tools support assessment; they do not replace clinical care.
Technology overview
Part II · Tools and planners
Explore the coaching tools.
Use your inputs or labelled examples. Calculations stay in your browser; they do not diagnose health or predict injury.
Explore the readiness-score demonstration
Tool 01 · readiness
How a morningscore is composed.
Explore how four inputs affect an illustrative readiness score.
- Sleep, against your usual window, not eight hours.
- HRV, against your own rolling baseline; one day is noise.
- Yesterday’s load, against the chronic average.
- Soreness and mood: what the wearable misses.
—Today’s score
—
Representative model · not a client
Illustrative only. Coaching decisions also consider history, symptoms and context.
Explore your wearable data
Tool 02 · your own data
Read your wearable data.
Drop a thirty-day CSV export here. It stays in your browser.
- A date column and at least a sleep or HRV column; resting heart rate and load are used where present.
- Columns are matched by header name.
- Nothing is uploaded, stored or sent.
The demonstration uses a simplified calculation, not a clinical model.
Or drop the file anywhere on this card.
Your thirty days will be read here.
Example — not a client
| Day | Sleep | HRV | Resting HR | Load | Readiness | Read |
|---|
Illustrative only: not a verdict on your health, and not a substitute for the assessment.
Explore estimated training zones
Tool 03 · training zones
Two numbers.Five zones.
Enter measured maximum and resting heart rates for a heart-rate-reserve calculation. A tested threshold is optional.
Use measured values. Calculated zones are an orientation, not a fitness test. Measure your maximum heart rate in a supervised test; do not attempt a maximal effort alone to find it.
| Zone | Purpose | Heart rate |
|---|
The assessment can refine intensity targets using your ventilatory and lactate thresholds.
Tool 04 · live pulse
Your pulse, against the zones.
Reads the standard Bluetooth heart-rate service from a chest strap or armband. Works in Chrome and Edge on Android, Windows and macOS over HTTPS; not in Safari or Firefox. Your pulse never leaves this page. It is not a medical device and cannot tell you anything about heart health.
—bpm
—
Last sixty seconds · the zone is read from the table above
Two planners and a demonstrator
Work backwardsfrom the date.
Plan an event, a time-zone change or a twelve-month marker comparison.
Illustrative tools, with the calculation shown.
Open the season planner
Planner 01 · the date
The season,laid out backwards.
Choose your event and date to see a timeline, including the week-twelve re-test where two blocks fit.
- The taper reduces volume before the event.
- Physique preparation is shown without dehydration protocols.
Enter the date and the timeline is laid out here, backwards from it: assessment, build, specific preparation, taper or peak week, the date, then recovery and re-test.
The full battery, bloods and a structured history; findings and the model at the end of it.
The essentials of the battery in a week, so what follows is written from numbers rather than guesses.
The general block: capacity, strength, the engine; volume progressed from the lowest effective dose.
The quality the date rewards, at the speed and range it is asked for; volume gives way to intensity.
Volume cut sharply, intensity kept, freshness allowed to arrive.
Composition and glycogen, not water and salt; the final session written from the readiness score.
A controlled deficit with protein kept high; strength tracked alongside composition and recovery.
Peak week: composition and glycogen, not water and salt; nothing dehydrated for the stage.
Load rebuilt by design and the same battery again, under the same conditions, so the next block is written from what moved.
Energy intake restored by design, volume rebuilt, and a post-date blood panel reviewed with your physician.
The same battery under the same conditions; the second block is written from what actually moved.
An illustrative timeline, not a promise; the real plan is written from your assessment. Proportions shown: the assessment two weeks; the remainder split about 45 : 35 between the build and the specific preparation (reversed for a physique contest, where the reveal takes the larger share); a taper of one to two weeks, or a seven-day peak week; one to two weeks of recovery and re-test after the date. Under six weeks the planner shows what is realistic rather than a full block; over forty weeks it plans two blocks with the week-twelve re-test between them. It cannot decide whether your goal is achievable by then; the assessment does.
Open the travel-planning demonstration
Planner 02 · travel
Plan the first three daysin a new time zone.
Enter your time zones, departure date and sleep window for an illustrative three-day adjustment plan. Medication and melatonin are for your physician to advise on.
- Use the illustrative seek/avoid windows below; direction and timing matter.
- Review the schedule with a coach if travel is frequent or competition is close.
Model assumptions. The estimated low point is two to three hours before habitual waking. The demonstration shifts it by about one hour a day eastward or one and a half westward. Individual responses vary; these are illustrative timings.
Choose the two zones and the plan for the first three days is written here.
| Day | Low point | Seek light | Avoid light | Sleep window |
|---|---|---|---|---|
| — | ||||
- Before departure, from the third day before the flight: shift sleep and meal times towards the destination by 30–60 minutes a day.
- The seek and avoid windows illustrate the model’s assumptions; individual responses vary.
- Caffeine only before the destination’s midday, and in any case finished eight to ten hours before the sleep window.
- Eat on destination time from the first meal on board; the last large meal three hours before the sleep window.
- Travel routines and recovery needs are individual.
- Training after arrival depends on sleep, symptoms and travel demands.
- Melatonin, sleep medication and anything prescribed are a conversation with your physician.
Typical figures from a simple model, not a prescription; the written travel protocol is adjusted to you. Listed offsets are standard time; the planner uses the offset in force on the departure date. All times are destination local time, days counted from arrival; the low point is the body clock’s core-temperature minimum on that day.
Explore the long-term trend demonstration
Demonstrator · the pace of change
Not an age. A direction.
Compare nine markers over twelve months. Each contribution is shown; changes below the noise threshold count as none.
This is a transparent trend demonstration, not a biological-age estimate.
Example values — not a client
| Marker | 12 months ago | Now | Weight | Full scale | Change | Normalised | Contribution | Direction |
|---|---|---|---|---|---|---|---|---|
| VO₂maxml/kg/min | 3 | +3.5 | — | — | — | — | ||
| Resting heart ratebpm | 1 | −6 | — | — | — | — | ||
| 7-day HRV averagems | 1 | +10 | — | — | — | — | ||
| Sleep regularity% of nights within ±30 min | 1 | +20 | — | — | — | — | ||
| Grip strengthkg | 2 | +5 | — | — | — | — | ||
| Lean mass (DXA)kg | 2 | +2 | — | — | — | — | ||
| Waistcm | 2 | −5 | — | — | — | — | ||
| ApoBmg/dL | 3 | −20 | — | — | — | — | ||
| Five-repetition sit-to-stands | 1 | −2 | — | — | — | — |
—
Direction of travel, twelve months
−100+100
Enter two readings twelve months apart for each marker.
The formula. Index = 100 × Σ(wi × ci) ÷ Σwi, where ci = (now − twelve months ago) ÷ full-scale change, clipped to ±1, with the sign reversed for the markers where lower is better (resting heart rate, waist, ApoB, chair-rise time) and set to zero when the change is smaller than the noise floor, taken here as a tenth of the full-scale change. Weights and full-scale changes are illustrative; in the practice each marker is read against its own measurement error and the trend is read with your physician where a laboratory value is involved.
The sources
The sources behind the demonstrations.
The training calendar, light-timing model and measurement limits behind these tools.
Sources & further reading (8)
- Eastman & Burgess, Sleep Medicine Clinics 2009 — how to travel the world without jet lag (the light-timing model)
- Waterhouse et al., The Lancet 2007 — jet lag: trends and coping strategies
- Khalsa et al., Journal of Physiology 2003 — a phase response curve to bright light in humans
- Bosquet et al., MSSE 2007 — effects of tapering on performance, a meta-analysis
- Hopkins, Sports Medicine 2000 — measures of reliability, the noise floor behind ‘no change’
- Science for Sport — the acute:chronic workload ratio, explained and criticised
- WHOOP — everything launched in 2025, including the Healthspan feature
- Oura — the new Oura app experience
These demonstrations support learning. Discuss clinical concerns with your physician.
Understand your model
See your numbersin the model.
One testing day, one findings session, a model that explains itself.