A catalogue polar is not your boat

Almost every cruising or racing yacht comes with a theoretical polar diagram, calculated by a VPP (Velocity Prediction Program) from nominal hull shape, sail plan and displacement. It's a useful starting point, but it remains an abstraction: it doesn't account for the actual sails you carry, their wear, the real load on board, or the way your crew actually trims on a given day of true wind. Tack after tack, that polar drifts further and further from what the boat does on the water.

The natural question is: why not build a polar from your own boat's data instead of a theoretical model? With pressure sensors on board, that's now possible.

The three ingredients of an empirical polar

A polar diagram built on real data comes from three pieces of information collected at the same time while sailing: boat speed (SOG or STW), the true wind angle and speed, derived from the apparent wind measured at the masthead, and — the element most boats never log at all — how good the trim actually was at that moment.

Without that last piece, a polar built from GPS tracks alone risks capturing badly trimmed legs too, artificially lowering the recorded performance for every wind angle and speed.

📊 Why trim quality matters in the dataset

Two identical legs at the same true wind angle and speed can produce different boat speeds simply because in one the pressure differential on the sail was close to optimal, and in the other it wasn't. A polar that ignores this variable mixes good and mediocre performance into the same point on the chart, and the result isn't reliable.

The role of pressure sensors in filtering the data

This is where SailSensor pressure sensors change the approach. By recording the pressure differential between the two sides of the sail in real time, they let you isolate only the legs where trim was actually close to the best achievable for those conditions. Adding SailSensor BoomAngle for boom angle enriches the dataset with a second trim parameter, especially useful for separating performance on a beam reach and broad reach, where the theoretical polar tends to be least reliable.

The principle is simple: for every combination of true wind angle and speed, only the points where the pressure differential was stable and close to the best value already observed for that point of sail are kept in the dataset. Poorly trimmed legs are discarded, not averaged in with the good ones.

From log to data: building the model in practice

The simplest workflow is this: collect outings across a range of true wind conditions, sync the pressure logs with boat speed and onboard instruments through SailSensor Bridge, which merges pressure, boom angle and NMEA data into a single stream, then filter out the stable stretches of at least 60-90 seconds for each wind angle and speed. The average speeds from these stretches become the points of your new polar.

With a handful of outings well spread across close-hauled, beam reach and downwind, and across the range of true wind speeds you actually sail in, the polar starts to take shape. It won't be complete on the first try: gaps will remain in the less frequent wind ranges, and it's normal to fill them in over the following weeks, one outing at a time.

A model that keeps updating, not a fixed number

The advantage of a polar built on your own data is that it's never final. Every outing adds points, corrects the estimates in less frequented wind ranges, and lets you spot the effect of sail wear over time: a systematic drop in speed at the same angles and true wind speeds, with the same measured trim, is often the first objective sign that the sails are losing shape.

A polar that's regularly updated also becomes a genuine tactical tool: knowing precisely whether to point higher or foot off to gain VMG, based on your own boat's data rather than a generic table, is the difference between deciding and guessing.

From chart to racecourse

A polar diagram built on your own data doesn't replace crew experience, it makes it verifiable. Every leg becomes a comparison against an objective reference, and every small speed gain found in training stops being a vague memory and becomes a stable point in your own performance model — ready to be confirmed, or challenged, on the next outing.

There's no need to wait for perfect conditions to start: a handful of ordinary training outings is enough to lay the foundations of your own polar. From there the model keeps improving outing after outing, race after race, becoming a more reliable tool exactly in the moments when split-second decisions matter most.