BTF · Downwind Performance

The Deep Downwind Gap

Pooled telemetry across every logged race says the boat is usually sailed shallower than its own best downwind angle. This isolates one lever, sailing 145–155° TWA as the default instead of the current ~132° habit — wind-axis VMG only, no course or finish-line tactics — and asks what that alone is worth.

+24%
more wind-axis distance made good, same hours sailed
10.9%
of downwind time actually spent in the 145–155° zone today
−82h
time to cover the same distance already sailed, at the deeper rate

VMG rises with angle, all the way to ~160°

Every downwind point in polar_coords_race at 8kt+ TWS, binned every 5° of true wind angle. VMG = Bsp × cos(TWA), averaged per point.

0 3 6 9 12kt 90° 105° 120° 135° 150° 165° true wind angle 145–155° TARGET current avg 132° 160° · 12.1kt (best trusted)
measured VMG (n ≥ 500 per bin) 165° bin thin (n=258) — suggestive only 145–155° target zone

What the gap is worth

421.6 hours of pooled downwind sailing (TWS ≥ 8kt, all logged races), actual vs. holding 145–155°.

Wind-axis distance made good
3,460nm→4,294nm
+834nm for the same time on the water
Time to cover that same 3,460nm
421.6h→339.7h
−81.9 hours (−19.4%)
Blended VMG rate
8.21kt→10.18kt
at the actual sailed TWS mix
Time already spent at 145–155°
45.9h of 421.6h
10.9% — the rest sailed shallower

Per wind-band breakdown

The gain holds across the whole range sailed, tapering only at the top end where the 24kt+ bin is a coarse catch-all.

TWS band Hours logged Actual VMG 145–155° VMG Gain
8kt39.84.75kt6.08kt+28%
10kt48.85.87kt7.30kt+24%
12kt56.96.59kt8.16kt+24%
14kt56.77.50kt9.28kt+24%
16kt65.38.58kt10.61kt+24%
18kt70.99.53kt12.01kt+26%
20kt49.810.61kt13.36kt+26%
22kt21.312.16kt14.59kt+20%
24kt+11.913.42kt15.30kt+14%

Case study: does it hold up race by race?

The pooled number is dominated by one race — TransPac ’15 alone supplies about 41% of the pooled downwind dataset, thanks to its long, steady trade-wind runs. Worth checking whether the same logic, run against individual races in isolation, gives the same answer.

Single-race check

TransPac ’15

106.4 hours of downwind sailing (TWS ≥ 8kt) at a weighted-average 133.7° TWA — almost identical to the pooled 132.1° habit, which is exactly what you'd expect given how much of the pooled dataset this one race supplies.

Wind-axis distance made good
880nm→1,068nm
+21.3% for the same time on the water
Time to cover that same 880nm
106.4h→87.8h
−18.7 hours (−17.5%)
Time already spent at 145–155°
7.6h of 106.4h
7.1% — even less than the pooled 10.9%
TWS band Hours logged Actual VMG 145–155° VMG Gain
8kt11.25.99kt6.89kt+15%
10kt18.66.72kt7.49kt+11%
12kt20.77.20kt8.49kt+18%
14kt19.08.11kt10.29kt+27%
16kt17.19.41kt12.09kt+29%
18kt13.410.67kt13.20kt+24%
20kt4.712.15kt14.52kt+20%
22kt1.413.78kt16.15kt+17%
24kt0.314.71kt16.94kt+15%

Same direction, slightly smaller effect on this one race: +21.3% here vs. +24.1% pooled — close enough to say the pooled number isn't an artifact of averaging across dissimilar races. The top two bands are thin for a single race (22kt n=200, 24kt n=33 — that last one is only ~2.75 minutes of steady-state 145–155° sailing) and shouldn't be leaned on individually, though they don't move the overall number much given how few hours they represent.

Single-race check

LA to Cabo San Lucas 2023

67.8 hours of downwind sailing (TWS ≥ 8kt) at a weighted-average 133.5° TWA — again close to the pooled 132.1° habit, but a shorter, more coastal point-to-point course than TransPac's sustained trade-wind runs.

Wind-axis distance made good
618.6nm→717.3nm
+16.0% for the same time on the water
Time to cover that same 618.6nm
67.8h→58.5h
−9.3 hours (−13.8%)
Time already spent at 145–155°
3.7h of 67.8h
5.5% — thinner than either race above
TWS band Hours logged Actual VMG 145–155° VMG Gain
8kt7.74.76kt6.07kt+27%
10kt8.55.92kt7.36kt+24%
12kt7.56.60kt8.28kt+25%
14kt5.27.49kt9.53kt+27%
16kt6.58.81kt10.44kt+18%
18kt8.910.04kt11.59kt+15%
20kt9.411.17kt12.56kt+12%
22kt7.512.59kt13.79kt+10%
24kt+6.814.44kt15.47kt+7%

Positive, but the smallest of the three checks: +16.0% here vs. +24.1% pooled and +21.3% on TransPac ’15 — still a real gap in the same direction, just shallower. Gain runs highest at the lower wind bands (+24–27% at 8–14kt) and tapers as breeze builds (+7% at 24kt+), the reverse of the pooled table's own top-end taper — and the low-wind bands here are thin for a single race (n=30–52 per band under 145–155°/8–14kt), so that low-end number is suggestive rather than solid. Consistent with a shorter coastal course spending less time in the kind of sustained, steady-state deep running that let TransPac and the pooled fleet push the angle harder.

Method

Pulled every downwind point (TWA 90–165°, TWS ≥ 8kt) from polar_coords_race, pooled across every race group currently logged. "Actual" is the real historical VMG per 2kt TWS band, weighted by real elapsed time and including maneuver time — it's what actually happened. "145–155°" is the steady-state VMG the boat has already demonstrated in that same wind band, with gybes/tacks excluded via the isTransition flag, so it's measured capability, not aspiration.

Each TWS band's actual hours were then re-priced at that band's 145–155° rate and summed, giving the total hypothetical distance for the identical time already logged. Every number on this page traces back to a live query against production telemetry, not a model.

What this doesn't say