Pensacola · 2026-11-06

• Upwind, both teams' angle and speed matched the best boats on the identical setup. At 15-17.5 km/h NZL (Valencia) sailed 50.7 deg at 44.8 km/h and AUS (Valencia) 50.2 deg at 45.7, for 28.9 and 29.3 km/h of VMG (72 and 57 periods). Race Team sailed higher and slower, 45.2 deg at 42.3 km/h, for 30.0 (25 periods). At 17.5-20 Race Team and Development both sailed 46 deg at 45-46 km/h for 31-32, against SWE (Valencia) at 46.1 deg and 42.8 km/h for 29.6 (22 periods).
• Development upwind at 15-17.5 km/h: sail about 2 deg higher. It sailed 52.6 deg at 44.6 km/h (38 periods). That is the speed of NZL (Valencia), but lower than its 50.7 deg and the 50.2 deg of AUS (Valencia), and the VMG that resulted was 27.3 km/h, about 1 km/h short.
• Race Team and Development downwind at 15-17.5 km/h: sail 2-3 deg deeper. Both sailed 131 deg at 53-54 km/h (18 and 19 periods). NZL (Valencia) sailed 133.5 deg at 52.9 km/h (38) and AUS (Valencia) 133.4 deg at 54.2 km/h (17). The speed is the same, so the angle is the whole gap: 35-36 km/h of VMG against 37.3-37.5.
• Development downwind at 17.5-20 km/h: sail deeper again. It sailed 128.8 deg at 53.2 km/h, against Race Team's 137.2 deg at 57.2 km/h (8 periods each) and 134.0 deg at 49.9 km/h for NZL (Valencia) (8). Going higher bought no speed, and VMG was 35.2 km/h against Race Team's 42.6.
• Development tacks: stay out of the water. At 12.5-20 km/h Development touched down in 6 of 20 tacks, against Race Team's 2 of 25, 5 of 45 for NZL (Valencia) and 4 of 38 for AUS (Valencia).
• Both teams' tacks: go in faster. At 15-17.5 km/h Race Team went in at 43.4 km/h and Development at 42.5, against 46.6-49.5 for the Valencia boats. Race Team made 18.7 km/h through the tack (6 tacks) and Development 19.1 (4), against 19.5 for NZL (Valencia) (10), 20.4 for AUS (Valencia) (6) and 20.7 for SWE (Valencia) (6).
• Both teams: test more twist upwind at 15-17.5 km/h. Race Team carried 13 deg and Development 17, against 18-22 for NZL (Valencia), SWE (Valencia) and AUS (Valencia). At 46-56 deg each team's fast quarter had about 20 deg and its slow quarter about 11: Race Team 47.8 km/h against 39.6, Development 48.2 against 37.8. It's a correlation, so test it on purpose.
• Race Team starts: on time and fast is now repeatable. Finish the last tack by 30 s to go. Guns 1, 2 and 5 were within 2 m of the line at 47-53 km/h, with pivots of 3.5-7.5 s, right on the 5-7 s aim. Guns 1 and 2 had the last tack done by -30 s and were on time. Of the three with it inside -17 s, two were 23-25 m late. Gun 3 came off the foils in an 11.5 km/h lull and was 76 m late.
• Both teams stayed on the foils more than NZL (Valencia). Upwind at 12.5-15 km/h Race Team foiled 64 % of the time and Development 78 %, against 45 %. Downwind it was 98 % and 100 %, against 45 %. Upwind off-foil spells over 3 s still cost each team about 5 minutes (Race Team 15 spells, Development 22), and most were not near a turn.

Read these before trusting a gap.

• Clear air: both teams sailed alone. Valencia had 13 boats practising together, with no races that day, on different water. So level numbers are encouraging, but they don't prove either team is quicker than a fleet boat in traffic.
• How the numbers are built: each boat's and each team's period figures are the average of its own better sailing, the periods above 95 % of its own median VMG in that band, with periods trading speed for angle left out. They compare better sailing with better sailing, and they don't show how often a team gets there.
• Wind: every boat is banded on its own sensor. If B15's reads 1.5 km/h differently from the Valencia fleet's, that is worth about 0.7 km/h of VMG upwind and 2 km/h downwind, as much as the downwind gaps above. The Deep dives tab has the detail, and it corrects the 5 Nov note that B15 reads high.
• Thin cells: a 5 or 8-period column or a 2-gybe column is a pointer, not a level. Ignore NZL (Valencia) upwind at 17.5-20: 5 periods, and it foiled only 38 % of its 5 minutes there.

• 12:04:38 (18:04:38 UTC) gun 2: the best start of the day, 1.0 m from the line at 52.8 km/h, 3.5 s pivot
• 13:40:43 (19:40:43 UTC) gun 5: on time at 47.3 km/h, then the fastest reach, 65.9 km/h steady and 65.3 km/h towards M1
• 13:25:24 (19:25:24 UTC) last tack before gun 3, into the lull: off the foils from 13:25:35 for 50 s, 76 m late at 16.5 km/h
• 13:34:14 (19:34:14 UTC) gun 4: last tack at -16.5 s, 28 km/h at -12 s, 23 m late at 35.9 km/h
• 13:49:05 (19:49:05 UTC) Race Team tack with a touchdown after gun 6: speed fell to 8 km/h and the lap stopped there
• 14:00:00 (20:00:00 UTC) Development's tack, the best of the day: 27.3 km/h through it, entry 49.4 km/h, lowest hull 860 mm
• 13:55:18 (19:55:18 UTC) Development's gybe, the best of the day: 44.6 km/h through it at about 20 km/h of wind
• 12:34:57 (18:34:57 UTC) Development off the foils upwind for 64 s at 14 km/h, 25 s after a touchdown tack
• 13:44:17 (19:44:17 UTC) Race Team off the foils upwind for 50 s at 14 km/h

Both teams' angle and speed were level with the best boats. On the identical wing, boards and rudders, their upwind speed is on a par with the best boats on the water that day.

Two ways of getting there at 15-17.5 km/h. Race Team sailed high and slower, 45.2 deg at 42.3 km/h, and that made the best VMG in the band. Development sailed 52.6 deg at 44.6 km/h, the speed of NZL (Valencia) but about 2 deg lower than NZL (Valencia) and AUS (Valencia), and that made the lowest. At 17.5-20 both teams settled on 46 deg at 45-46 km/h. SWE (Valencia) sailed the same angle 3 km/h slower, and AUS (Valencia) 54.7 deg at the same speed.

What differs at 15-17.5 km/h, on the same wing, boards and rudders (the settings table is below):

• Twist is the clearest gap, and each team's own fast periods point the same way (see Deep dives).
• CA6: NZL (Valencia) and AUS (Valencia) had the top of the wing inverted on average at 15-17.5, and neither team did. Race Team was +4.3 deg and Development +1.1, against about -2. At 17.5-20 both teams inverted it too, -2.8 and -4.6 deg.
• Jib lead sits 10-20 points further out for both teams than for all three Valencia boats. The jib is the same size, so that is now a real setup question for the crews.
• Camber is 1-2 deg lower for both teams than for the Valencia boats.
• Rudder rake can now be compared. Race Team carried about 0.4 deg more than NZL (Valencia) and AUS (Valencia), in line with its high mode.

The gap is angle, not speed. At 15-17.5 km/h Race Team and Development matched NZL (Valencia) and AUS (Valencia) for speed but sailed 2-3 deg higher, and the VMG that resulted was 1.5-2 km/h lower.

Race Team had about 1 km/h more wind than the others at 15-17.5, so its gap there is a little bigger than it looks.

Development is high at 17.5-20. It sailed 128.8 deg at 53.2 km/h, against Race Team's 137.2 deg at 57.2 km/h and 134.0 deg at 49.9 km/h for NZL (Valencia). It was the same at 12.5-15: 129.6 deg against Race Team's 137.8 (10 and 5 periods), though NZL (Valencia) and AUS (Valencia) sailed about 130 deg there too.

Deeper went with faster for both teams. Race Team's top quarter on VMG sailed 139.5 deg at 57.2 km/h and its bottom quarter 123.3 deg at 50.7 (53 periods). Development's sailed 133.0 deg at 54.2 km/h and 124.1 deg at 47.6 (65 periods). Periods bearing away on stored speed are already out of that.

What differs at 15-17.5 km/h (the settings table is below):

• Wing rotation is about 3.5 deg more for both teams than for NZL (Valencia) and AUS (Valencia), and it was the same at 12.5-15. Fast and slow periods don't separate on rotation, so it's a question rather than an answer.
• AUS (Valencia) ran a much tighter jib than both teams and NZL (Valencia): less sheet, further inboard, nearly twice the cunningham. The jib of NZL (Valencia) looks like both teams', so the jib alone doesn't explain the angle gap.

All Race Team, start line to M1 only: Development Team had no starts today. Gun 3 is left out, because the boat was off the foils for most of its reach.

The fastest reaches had the top of the wing inverted. Guns 1 and 5, at 97 and 91 deg in 18 km/h, made 65.6 and 66.1 km/h with 26-31 deg of twist and CA6 at -8 and -14. Gun 6 had a little more wind at 104 deg and made 61.7 km/h with 19 deg of twist and CA6 at 0. It was also 7-13 deg deeper, so angle may explain some of that. Twist and CA6 are the ones to test on purpose: cunningham and sheet load were highest on guns 1 and 5 too, but load follows speed.

The references point the same way, but only just. The fastest reach near this angle was AUS (Sydney), 56.9 km/h at 90 deg in 17.4 km/h, with rotation 9, twist 26 and CA6 at -7, close to guns 1 and 5. CAN (Sydney) and FRA (Rio), at 51.2 and 50.4 km/h, had rotation 19-21, twist 6-8 and CA6 at +15 to +17. ESP (Rio) doesn't fit on twist, with 24 deg at 49.8 km/h, though its CA6 was 0.

Don't read the speed gap as a result. Race Team's 61-66 km/h at 91-104 deg was on a clear course. Each reference is a single reach in a fleet start, with TWA wander of 4-13 deg against Race Team's 2-3, and their jib and crew aren't known. Sydney 28 Feb and Rio 10 Apr are the only race days on the identical wing, boards and rudders, and NZL sailed neither. Only 5 of Sydney's 40 reaches were at 90 deg or more.

Gun 4 is the outlier. It reached at 62 deg, 29-42 deg higher than the others, and at 49.1 km/h made 12-17 km/h less towards M1, reaching it at +34 s against +25-28 s. It started mid-line and 23 m back, and had 2-4 km/h less wind on the reach.

Through the filter, tacks have the most to compare. Filter Score 70 passed 16 of Race Team's 22 tacks and 13 of Development's 21, but only 5 of 13 and 3 of 11 gybes. All of these are medians, because the cells are small. SWE (Valencia) was the best tacker on the day and GBR (Valencia) the best gybing boat, on only 3 gybes, so AUS (Valencia) is shown beside them.

Tacks at 15-17.5 km/h:

• Entry speed is the gap for both teams. They went in 3-7 km/h slower than the Valencia boats. Development still came out as well as anyone, 37.7 km/h at +10 s against 37.3 for NZL (Valencia), and Race Team made 34.1. Build speed before the tack.
• Board overlap is fine. Race Team's 6.2 s matches the Valencia boats, and Development's 4.2 s is the shortest.
• At 12.5-15 km/h both teams held up: Race Team made 20.0 km/h through (2 tacks) and Development 19.7 (4), against 18.6 for NZL (Valencia) (7), 19.8 for SWE (Valencia) (4) and 18.9 for AUS (Valencia) (5). Half of Development's touched down.

Touchdowns (lowest hull under 100 mm). Development's were at 12:31:26, 12:34:32, 12:43:52, 13:14:07, 13:54:06 and 14:02:51, with one more at 13:11:54 in lighter wind. Race Team's were at 13:30:59 and 13:49:05. All local; add 6 h for UTC. Three of them, Development's 12:43:52 and 14:02:51 and Race Team's 13:49:05, fell to 10 km/h or less.

Gybes: neither team touched down in 31 gybes. At 15-17.5 km/h Race Team made 34.3 km/h through (2 gybes) against 30.3 for NZL (Valencia) (7), 33.7 for GBR (Valencia) (3) and 31.1 for AUS (Valencia) (5). That's good, but it rests on 2 gybes, and Development had too few through the filter to show.

The columns are all the turns that passed the filter; the Bear-away figures further down this tab use the report's usual filter, and they cover the boat across both teams.

Six guns, all Race Team, from the boat's own start channels (PC_DTL_m, BOAT_SPEED_START). The boat's XML ID confirms the course for each: 110606 for the first two and 110607 for the last four, on a new axis (33 deg against 348). Those positions also match the course files to within about 8 m.

What made the good ones good. Guns 1, 2 and 5 pivoted in the last 2 s and took 3.5-7.5 s to reach 90 % of reach speed. Gun 5 shows a late tack can work, but it had 33 km/h at -14 s where guns 4 and 6 had 28 and 32 km/h at -12 s, and both of those were 23-25 m back.

Gun 6 was quick but bore away at -12.7 s and was at 95 deg TWA on the gun, so it was already reaching before the line, 25 m back at 52.7 km/h.

Gun 3 is a wind and tack problem: 11.5 km/h at the gun, and the tack at 13:25:35 put the boat in the water for 50 s.

No comparison here. Valencia on 3 Sep had no races, so there are no starts from that day to compare with. Race Team's 47-53 km/h was on an empty line, so the timing and pivot look sound, but they haven't been tried in traffic.

All Race Team. Valencia on 3 Sep had no races, so there are no roundings from that day to compare with.

• M1 is a strength: Race Team carried about 60 km/h round it on guns 1, 2 and 5, on a clear mark with no traffic.
• Bottom gate: gun 5 came in fastest at 58 km/h and left slowest at 37, and gun 1 got the hull down to 411 mm.
• Gun 6's M1 fell to 40 km/h. It then went no further down the course, and the lap stopped with a touchdown tack at 13:49:05.

The day. Every event day in the calendar was searched, boat by boat and minute by minute, for the 24 m wing, LAB2 boards and LARW rudders. Valencia Thu 3 Sep is the only one with NZL on that setup in this breeze.

Rudders checked independently of the setup channel. Rudder rake at the same pitch separates the rudders. Foiling at 12.5-20 km/h, B15 sits inside the Valencia fleet on both points of sail, so both teams were on the same rudders as that fleet:

One rule, with each boat's own bar.

• Periods first: foiling for the tables, wing set (camber at least 15 deg), and not within 10 s of a turn.
• Speed for angle: periods where boat speed fell over 1 km/h inside the 6 s while turning up (upwind) or bearing away (downwind) are left out. VMG rewards that for a few seconds and a boat can't hold it.
• Each boat's bar: a period must be at or above 95 % of that boat's own median VMG in that band, and Race Team and Development each have their own bar from their own periods. About 55-80 % of periods clear it.
• The figures: averages of what clears the bar, so TWA, boat speed and VMG describe the same periods, and CA1 minus total twist equals CA6.
• Manoeuvres: app Filter Score of at least 70, away from marks, real turns only (at least 50 deg of heading change and the wing set). These stay medians, because most cells hold 2-10 turns.

A bar set from each boat's own median means the tables compare better sailing with better sailing. A boat that spends more time off the pace gets a lower bar, so these numbers don't show how often it reaches that level.

Best boat. It's the best average VMG with the wind taken out, weighted by how much of today's sailing was in each band. AUS (Valencia) is shown as well wherever it wasn't best.

SWE (Valencia) is only just ahead of AUS (Valencia) upwind. ESP (Valencia) wasn't near the front on any of them, so AUS (Valencia) is the better reference. Gybes were too thin to rank properly: GBR (Valencia) led on 3 gybes.

Reaches use Sydney 28 Feb and Rio 10 Apr instead, because Valencia had no races. They are the only race days on the identical wing, boards and rudders: every boat's channels say so, and rudder rake at the same pitch agrees (Sydney 3.6 deg upwind at -3.0, Rio 3.7 at -3.2). NZL sailed neither, and the jib and crew aren't known.

Every boat is banded on its own sensor, the report's usual rule. B15 sails alone, so its sensor can't be checked against another boat.

The marks don't settle it. Today B15 read 1.5 km/h above the mark boats. At Valencia on 3 Sep the boats read 6-8 km/h above their marks. So how far a boat reads above the marks changes a lot with the venue and the day, and it doesn't show that a sensor is wrong.

Correction to the 5 Nov work. That work treated B15 as reading about 2.4 km/h high, because Geneva boats read only 0-1 km/h above their marks. Reading above the marks doesn't show a sensor is high. B15's calibration is still unknown.

What an error would be worth. Across today's foiling periods, VMG rose about 0.45 km/h per km/h of wind upwind and 1.35 downwind. So 1.5 km/h of sensor error moves both teams' gaps by about 0.7 km/h upwind and 2 km/h downwind. The downwind gap to NZL (Valencia) and AUS (Valencia) at 15-17.5 is about that size, so treat it as a pointer. The 2-3 deg of angle is the part to act on.

Comparing a team with itself removes the sensor and the setup. These use each team's foiling periods with speed-for-angle periods out and no bar, because a bar would remove the slow ones. Same angle, wind taken out, fast quarter against slow quarter, as averages. At 15-17.5 km/h upwind NZL (Valencia) carried 20.6 deg of twist with CA6 at -1.8, and AUS (Valencia) 21.9 deg with -2.1.

Both teams tell the same story. Each team's fast quarter upwind has the twist of NZL (Valencia) and AUS (Valencia) and the same inverted CA6, about -2 deg, and its slow quarter has about half the twist and CA6 at +7 to +8. That is why more twist is worth testing. Two cautions. Load follows speed, so the jib load and cunningham load may be the result rather than the cause. And a slow period can simply be a boat recovering. Twist and sheet are the settings to test on purpose.

Downwind the two quarters sit in the same wind and about 8.5 km/h apart for both teams, and twist is again the biggest difference. Race Team's quarters are only 7 periods. Across both teams, rotation and jib lead did not separate fast from slow on either point of sail.

Each team's periods above its own bar, with the wind taken out so the teams' different breeze does not count.

Upwind the teams were close. Race Team was 0.6 km/h ahead, sailing 3 deg higher and a little slower. Downwind Race Team was 1.5 km/h ahead, faster and deeper. Its rudder was steadier downwind too, with 0.59 deg of movement in a period against 1.03.

A spell counts when the boat was sailing upwind with the wing set and either both hulls were under 400 mm or a hull was under 200 mm for over 3 s. Crew changes are left out. Most spells came from losing the flight on a straight line, not from a turn. Several of the longest began slow and low, at 25-29 km/h and 54-61 deg TWA, in 14-15 km/h of wind. The Upwind tab has the time in each sailing mode, and both teams foiled at least as much as NZL (Valencia) on the same setup.

• Wind bands: 2.5 km/h, on each boat's own wind reading corrected by its calibration offset for the day. Offsets (km/h, + = the sensor reads high), from moments boats sailed within 150 m of each other: B15 +0.0.
• Bands are never merged. A band is shown only with at least 5 of our periods and 3 other boats with 5 each; for manoeuvres and roundings, 2 of ours and 3 other boats with 2 each.
• Racing is split by leg (leg 1 = reaching, the final leg left out). Practice is split by wind angle: upwind at or below 75°, downwind at or above 120°.
• Samples: 6 s straight-line periods from the app (upwind, downwind), each the mean of its ~12 half-second readings; 2 s chunks of the steady part of leg 1 (reaching: 10 s after the start to 5 s before mark 1).
• Angle words: higher = closer to the wind (a smaller TWA) and lower = further from it (a bigger TWA), upwind and downwind. Upwind, 50 deg is higher than 52 deg. Downwind, deeper = a bigger TWA, so 140 deg is deeper than 130 deg, and 130 deg is higher.
• Periods: an upwind or downwind period is left out if it starts within 10 s of one of the boat's tacks, gybes, bear-aways or mark roundings (the boat is still accelerating), if the wing is not set (under 15 deg of camber, as on a tow), or if it is trading speed for angle: boat speed falling over 1 km/h inside the 6 s while turning up upwind or bearing away downwind. What is left must be at or above 95 % of that boat's own median VMG in that point of sail, sailing mode and wind band; with teams named, each team has its own bar. The table below shows what each step removed race by race.
• Period figures are averages of what clears each boat's bar, so VMG, speed and TWA describe the same periods. They compare better sailing with better sailing, not how often a boat gets there. Manoeuvres, roundings and starts stay medians, because they are few. No figure rests on a single sample. Fleet = the median of the other boats' figures, so each boat counts once. Best boat = the other boat with the best: average VMG (km/h) upwind and downwind, VMC to mark 1 on the reach, VMG through the manoeuvre on tacks and gybes, metres below target on bear-aways, exit speed on roundings, metres at mark 1 on starts. VMG % is against each team's own polar, so it is shown for us but never used to rank boats.
• Targets: SPEED v2 cheatsheet, LAB2 24m with the Big jib and LARW rudders: the sheet's row at the band's lower wind speed (the 30 km/h row for 30-32.5), as written, never interpolated. Clew targets are converted to mm with the sign turned round (- = to windward, as in the data); jib track is converted from sheet angle to jib lead % (0 % = fully inboard). Not compared: jib sheet load (the sheet's units are not kgf yet).
• Phases included: practice, racing. Races left out: none. Crew on board: 6. Notes: 24 m wing, LAB2, LARW. Race Team 11:45-12:18 and 13:18-13:52; Development Team 12:20-12:50, 12:55-13:15, and 13:52 to the end..
• Past event days (Valencia 2026-09-03): every boat follows one rule. Periods trading speed for angle are left out, then a period must be at or above 95 % of that boat's own median VMG in its band, and period figures are averages of what clears that bar. Tacks and gybes need a Manoeuvre Filter Score of at least 70 and stay medians. Bands are each boat's own sensor. The Deep dives tab has the detail.
• Manoeuvres are free manoeuvres only: anything within 10 s of the start, a mark or the finish belongs to the Mark roundings tab, and anything after the finish is left out. Practice tacks and gybes with an app quality score under 25 are left out (the boat was not trying); racing ones are all kept.
• VMG through the manoeuvre = ground made upwind (tacks) or downwind (gybes) over the app's 25 s window, in km/h - the measure that compares boats directly. Metres lost is the app's ghost-boat figure; each ghost holds that boat's OWN pre-manoeuvre VMG, so a boat that was quicker going in shows more loss for the same turn. For bear-aways it is distance sailed below the boat's target speed over 15 s. Touchdown = lowest hull under 100 mm.
• Reaches are start line to M1 only: the steady part, 10 s after crossing the line to 5 s before M1, straight and foiling, as 2 s samples, with medians of those. Practice sailing at reach angles is left out on purpose: on 6 Nov 2026 it was mostly turns and acceleration, with TWA moving over 3 deg inside 77 % of those periods. A reach is only compared inside one wind band and one TWA range (55-80, 80-100, 100-130 deg), both taken from the whole reach, because speed moves far more with angle than with wind.
• Tracks are GPS, and up the page is always where the wind is coming from: the average TWD of the 60 s either side of the moment, leaving out the 10 s around a turn. A turn's fixed point is the moment its heading crosses that wind, directly into it on a tack and directly with it on a gybe, and its line runs from 5 s before to 10 s after. That is not the TWA crossing in the tables, which lags the heading through a turn. Turns entered on port are mirrored. A start's line runs from 30 s before the gun to M1, and guns sharing a start line and M1 overlay on one chart. A bottom mark line runs from 7 s before the closest pass to 10 s after, the mark at the centre, on one chart for right turns and one for left, taken from the boat's heading.
• Turn time series share that fixed point as T = 0 and run from 10 s before to 10 s after, longer than the tracks so the board down button shows. Yaw rate and rudder read + into the turn; clew, heel and AWA-E1 are on the exit tack's sign, so + is to leeward coming out. The old hull and board are the ones to leeward going in, the new ones to windward going in. The AoA limiter is the boat's own limit, ANGLE_DB_RAKE_LIM_deg, a rake floor already allowing for pitch. They read the report's 0.5 s data: AWA-E1 is a 1 Hz channel and clew a 2 Hz one. The button marks are the new board's down press and the old board's up press, the moment the 20 Hz board command goes on, which is ahead of the stow states the app reports. The wing invert is the app's: camber 1 passing through zero. Each mark sits on the line at that moment. The dashed trend is a team's median at each moment, like every other manoeuvre figure, and needs two turns with a value there.

What the filters removed