1 What you need first
Three things, in this order:
- Telemetry for the day. Import it, or let Race Control send it live — see Importing telemetry.
- A polar. Every number in the report is a distance from a target, and the polar is where the target comes from. Upload one under Polars and targets. A boat with no polar still gets a report; every target-relative layer of it says there is no polar.
- Tuning tables, if you keep them. Heel, rudder, forestay load and trim targets are drawn behind their panels when the boat’s tables carry a column for them.
Calibration entries and upwash corrections apply at read time, so entering one after the regatta changes the report. That is intended: see The calibration lens.
2 Opening a report
Open Reports. Pick the boat, then the regatta; the regatta’s sailing days are listed under it with the name and the type the crew gave each one, and the number of races in it.
Each day offers its two XY reports, upwind and downwind. They are separate documents on purpose: the two halves of the polar have different reference lines and different target curves, and putting a downwind day under an upwind heading would be wrong rather than merely confusing.
Show what was logged on a day reads that day’s coverage and lists the channels that are missing or thin, before you open anything. A day with no forestay-load sensor says so here.
A sail’s own report — its speed against target over the sail’s whole life — is not a day report and is not listed here. It lives with the sail, on the boat’s sails page; see The sails log.
3 Naming the day, and saying what it was
A report header prints the day’s date, the crew’s name for the day, and the day’s type: practice, testing or race. Both fields are set in the same place — open the regatta under Debrief and use Name day on the day heading.
They are two different fields and neither replaces the other. The name is your own sentence about the day. The type is the classification a report can read: Practice day, Prac, Training and Övning are four spellings of one fact, and a report that prints the day’s character needs the fact rather than the spelling. Leave the type unset and the report simply does not print one.
4 Mode and grouping
Two controls change what the report argues about. Both are in the address, so a link you send opens the same document.
| Control | Choices | What it changes |
|---|---|---|
| Mode | Upwind, Downwind | Which half of the polar the document covers. Reaching is in neither and is not measured. |
| Grouping | By race leg, by tack, by sail, by companion set | How the segments are split into the coloured series. |
The four groupings answer four different questions:
- By race leg separates the racing from the warm-up, and one race from another.
- By tack is the first place to look for an instrument problem or a crew habit: a boat that is consistently better on one board at the same wind speed is telling you something.
- By sail counts a segment for a sail when the whole segment sits inside a closed interval of that sail in the log. Keep the log and this grouping fills itself.
- By companion set groups by the whole set of sails flying together. When two sails are always up at once, this is the only attribution the data supports — it reports the combinations that flew, and models no per-sail role.
Season reference adds a dashed curve behind each series: the same trend over the boat’s last 365 days. It is the expensive half of the report, so leave it off for a first look and switch it on when you want to know whether a day was normal for the boat.
5 Reading a panel
The first three panels are the day against its targets: Δ boat speed, Δ true wind angle and VMG as a percentage of target, each against true wind speed. Then come the covariate panels — what the boat was set to while it produced those numbers.
Every panel carries the same layers:
| Layer | What it is |
|---|---|
| One dot | One steady segment — at least 30 seconds of settled sailing, reduced to its median. |
| Thick line | The smoothed trend for that series. It fades where the evidence behind it thins, and it stops where there was not enough. |
| Shaded strip around it | The approximate 95 % uncertainty band for that trend. |
| Dashed line | The same trend over the boat’s last 365 days. |
| Solid horizontal line | The reference: zero on a delta panel, 100 on a percentage panel. A raw covariate has none. |
| Target curve | The boat’s own target for the quantity, from its polar or its tuning table. The legend says which sheet. |
| Two shaded regions | The covariate’s spread over the fastest and the slowest quarter of the day by VMG. |
The two shaded regions are the part worth spending time on. If the fast quarter carried a degree less rudder than the slow quarter at the same wind speed, the regions separate and the page has told you something to try. If they overlap, that covariate did not distinguish the fast sailing from the slow, and the overlap is the answer.
The words at the top and bottom of each y axis say which direction is better — Faster and Slower, Wider and Narrower. A signed delta with no words on the axis is read backwards about half the time.
6 Where a chart is missing
A panel that could not be drawn still prints its heading, with the reason under it. The reasons and what to do about each:
| Reason | What to do |
|---|---|
| No polar in force | Upload one, or set an existing source to the performance role. |
| Not logged on this day | The channel was never recorded. Check the instrument, or the channel mapping on the import. |
| No steady segment in this mode | A day of manoeuvres has nothing steady to measure. Try the other mode. |
| Too few segments | The data exists and there is not enough of it. A coarser grouping usually fills the panel. |
| No target for this quantity | Add the column to the boat’s tuning table if you have a target for it. |
| Still being computed | Nothing. The page asks again on its own and the layer fills in. |
7 Reading a panel as a table
Every chart has a Show as table control under it, carrying exactly the numbers the chart drew: the series, the wind speed, the value and the weight behind each point. It is there for readers who cannot use the chart, and it is the fastest way to copy a figure into a crew note.
8 Printing
Print opens the report as a printable document: white ground, A4 landscape, 10 mm margins, at most six panels a page. Print it from the browser, to paper or to PDF.
Every page carries the boat, the sail number, the date, the day name and type, the regatta, the report and the page number. The last page carries:
- The yardstick — which polar, in which role, uploaded when; how many calibration entries and upwash rows were in force; the version of the computation; a fingerprint of the telemetry.
- The limits that apply to this document, in plain sentences.
- The standing caveat about what a performance number is and is not.
The printed sheet draws the same document as the screen. There is no second computation and no second set of numbers, so a page you print cannot disagree with the page you read.
9 The manoeuvres report
Every sailing day also offers a report of its manoeuvres, in two halves: tacks, and gybes. They are separate reports because a gybe turns through nearly twice the angle of a tack and costs several times the metres, and putting both on one scale flattens every tack against the bottom of it.
It carries four figures:
- The day’s best three, aligned. One strip per quantity — progress to windward, boat speed against target, wind angle, heel, rotation rate, rudder — with all three manoeuvres drawn on each. Every strip shares one time axis so you can read a column straight down the page, and the dashed vertical line is the instant they are aligned on. The caption says which instant that is.
- The path over the ground. North up, with a scale bar in metres, coloured by boat speed as a percentage of target. Where a second had no reading the track still runs through it, unpainted.
- Every manoeuvre of the day, ranked. The metres each one cost, what your own best in that wind speed cost, and the difference. Fewer metres is the better number and the colouring reads that way round.
- The shape a manoeuvre makes, by wind band. Six measurements on one chart, drawn so that further from the centre is the better number on every spoke. A shape that stops short on two spokes says so under the chart: it did not score badly there, it had no reading.
The reference throughout is your own best, not a manufacturer’s figure: the twentieth percentile of this boat’s recorded history of that manoeuvre in that wind speed. It needs twelve manoeuvres across three wind bands before it is drawn at all, and below that the target column says so instead of guessing. The arithmetic is in Manoeuvre loss.
10 The start report
Every race with a gun offers a report of its start. Open it from the race in the report library, or from the button on the debrief player while you are watching that race.
- This start against the day’s others. Boat speed and progress at the gun, how far from the line you were, how long the crossing took, and how long it took to reach ninety-five per cent of target speed and of best progress.
- The last three minutes. The approach over the ground with the line on it, coloured by boat speed against target.
- Second by second, from 1:20 before the gun. Distance behind the line, wind angle against target, boat speed, progress, wind direction, the recorded burn and the rudder. The second you crossed is emphasised. Long enough to run to several printed sheets, and each one says which rows it holds.
Only the numbers nobody recorded are computed afterwards, and the report says which those are. The reasoning is in Time to the line, and the line geometry in The start line.
11 The race legs report
The same race, leg by leg: how long each leg took, the wind it was sailed in, boat speed and angle against target on it, progress to windward, and how many manoeuvres it held. A leg the wind instruments did not cover carries its times and nothing else.
Legs come from the fleet-tracking provider for that race. A race with no provider data is reported as one leg, and the report says so rather than inventing turns out of a heading trace.
12 Exporting a PDF
Every report page has an Export as PDF button. Press it and the report is rendered on the server, which takes a few seconds; the button then becomes a download link.
The PDF is the same document as the page, laid out for A4 landscape, with a footer on every sheet carrying the day it was generated and the page number, and the standing caveat in full on the last sheet.
- Asking twice costs one render. The file is keyed on what the report is made of — the boat, the day or race, the polar, the calibration in force and the version of the computation. A second request for a report nothing has changed under hands back the file that already exists.
- Change an input and you get a new file. Enter a calibration, upload a polar, import a missing hour, and the next export is genuinely a different document, because those are the things that move the numbers.
- The link lasts fifteen minutes. Press the button again for a fresh one. That costs one request and no rendering.
Printing from the browser is still there and produces the same sheets: the Print button opens the report on a light theme laid out for the same paper.
13 What a report is not
It is evidence, not a verdict. Sea state, crew, rig tune, sail age and a team getting better at a manoeuvre are all real and none of them is modelled. A report says what the boat did against its own targets, with the weight behind every statement printed beside it.
It is also not a player. There is no scrubber, no hover read-out and no zoom: exploring a race second by second is what the debrief player is for. A report is the paged, reproducible argument you take to the next debrief.
For every constant, gate and formula behind these pages, see Performance trends.