ESTIMATOR METHODOLOGY.

The unified BTE engine predicts an anticipated WOO-equivalent jump height for Andy in a specified riding opportunity. It separates what the conditions and setup make available from how efficiently the take-off mechanics convert that opportunity.

Wind, gust spread, kite area and rider mass establish the aerodynamic opportunity layer. R28.3 uses a conservative effective-wind prior of mean wind plus 32% of the gust delta, then calculates dynamic pressure from a fixed standard air density of 1.225 kg/m³.

The same wind and kite do not create the same rider-relative opportunity at every body mass. BTE therefore expresses aerodynamic opportunity as a load-to-weight potential: dynamic pressure × kite area divided by rider weight × gravity. This is a modelled opportunity index, not measured line tension.

Personal take-off mechanics provide the conversion layer. Edge, Pop and Pop Velocity are always used when available; Approach/Take-off Speed joins the model when the evidence set is large enough. The engine finds Andy’s nearest historical mechanics neighbours rather than applying a generic skill-level multiplier.

R28.3 uses a rider-specific opportunity factor rather than the old generic Power Index. A 92 kg rider on a 9 m kite in 28 kt mean / 34 kt gust provides the internal reference load. The opportunity adjustment is intentionally conservative and capped until richer wind-linked sessions support stronger empirical calibration.

Water state remains contextual evidence rather than a hard-coded height coefficient. The current unified model uses standard air density, so temperature, pressure, humidity and altitude are not yet allowed to move the headline prediction. Those variables can be added later only if they improve leave-one-session-out accuracy.

The physics layer changes the opportunity; personal historical mechanics determine how that opportunity normally converts into height. Expected Andy is the normal personal outcome for the mechanics and opportunity. Best Andy is the conditional high-performance benchmark from Andy’s own strongest comparable evidence. The 20 m line remains a goal, not an expert assumption.

The result is a personal performance estimate with an evidence range and confidence score, not a guarantee. Sparse context, extrapolation beyond the rider’s demonstrated mechanics, rapidly changing wind and missing GPS reduce confidence.

Raw knots are not enough. The engine uses mean wind, gust, wind direction, spot geometry and — when rider heading and speed exist — a vector Rider Apparent Wind Proxy. Barassie uses W as the onshore reference; Troon South uses WSW. BeachStats observations remain the preferred local wind evidence.

Edge, Pop, Pop Velocity and Approach/Take-off Speed are first-class conversion variables. GPS-derived Speed Retention, Heading/Carve Change and jump distance are designed as the next high-value features as BoostIQ phone/watch GPS coverage grows. Airtime, distance and actual height are outputs and are not fed back into the same jump’s height prediction.

The current R28.3 engine uses fixed standard air density (1.225 kg/m³). Temperature, pressure, humidity and altitude are deliberately excluded from the live headline prediction until there is enough matched evidence to prove they materially improve accuracy.

The model is personal and conditional. Comparable sessions are weighted by spot, wind, kite size/model, direction, board, water state, tack, line length and harness context. Kite model, board, line length and harness only receive a numerical personal factor when there is enough contrasting session evidence; otherwise they remain context with reduced confidence rather than invented coefficients.

Prediction confidence falls when wind, kite, vector apparent-wind evidence or equipment contrast is missing. The selected session is excluded from its own replay/training comparison so the engine cannot learn its answer from the result it is trying to predict.

The predictor is a performance-intelligence model, not a safety or kite-sizing tool. Manufacturer wind ranges, local knowledge, rider competence, equipment condition, gusts, water state and rescue options take precedence over any BTE height estimate.