Why one number was never enough
Distance to the line divided by boat speed is the first answer anybody writes down, and it is wrong in three separate ways at once. The boat is rarely pointed at the line. It is rarely doing the speed it could be doing. And it is rarely on the course it is going to start on.
Each of those is an assumption, and Race Control relaxes them one at a time. That gives three answers rather than one — and the three of them together say more than the sharpest of them alone.
One thing they share is the distance. Every tier divides the perpendicular distance to the infinite line through the two marks, because only that distance decides whether you have started — the same signed number, from the same two pings, that the start-line geometry computes. Its sign is load-bearing here too: once the boat is on the course side there is no countdown left to run, every tier goes quiet, and the card shows the crew that they are over rather than a number.
1 CURRENT — nothing changes
The first tier is dead reckoning: keep doing exactly what you are doing. Its only subtlety is that the line does not care how fast the boat is going, only how fast it is closing. A boat reaching along a line at eight knots is closing it at nothing at all.
So the boat’s velocity is projected onto the direction square to the line — the direction the perpendicular distance actually shrinks in — and that projection is the closing speed:
Read as a sentence: only the part of your speed aimed at the line gets you to the line.
Worked, on a boat 100 metres from the line: pointed straight at it at 5.0 knots, the whole speed closes and the answer is 38.9 seconds. Turn 60° off and only half the speed closes — 4.0 knots becomes 2.0 knots of closing speed, and the same 100 metres now takes 97.2 seconds.
This tier is the one with no dependencies. Position, speed, course, wind direction and two pings — no polar, no tuning sheet, no wind speed. It is what a boat with a bare GPS still gets.
A boat sailing exactly parallel to the line has no arrival time, and the card says so rather than dividing. The reason is worth naming: in floating-point arithmetic the cosine of 90° is not zero but a number around 6 × 10−17, so an unguarded division would produce a countdown of some sixteen billion years, rendered confidently in minutes and seconds. The floor sits at a hundredth of a knot of closing speed — five millimetres per second, at which a boat 100 metres out arrives in over five hours. That is not a countdown; it is an em dash.
2 TARGET — the moment you trim on
The second tier relaxes the assumption that the boat stays at the speed it is doing. Same course, same geometry, same projection — but the measured speed is replaced by the target boat speed from your polar at the wind angle that course implies.
Read as a sentence: how long this would take if the boat were going as well as it can on the angle it is already on.
Continue the worked example. The boat is 100 metres out, 60° off the line, doing 4.0 knots — 97.2 seconds at the current tier. Suppose the polar says this boat should be doing 8.0 knots at that wind angle in this breeze. Then the closing speed doubles and the answer halves, to 48.6 seconds.
That difference — 48.6 seconds — is not noise between two estimates. It is a measurement of one specific thing: the speed you have not got yet. Trim on and the first number walks down to meet the second.
This tier is indifferent to how fast the boat is actually going, including when the answer is “we don’t know”. A dead speed log takes the current tier with it and leaves this one standing. It needs the polar and the wind speed instead — and without either, it returns nothing rather than quietly falling back to the tier below wearing this tier’s name.
3 ROUTED — the fastest way there
The third tier relaxes the last assumption: that you hold this course. It asks what the fastest sailable path to the line is, and times that.
Sailable is the operative word. A boat can sail dead downwind; it cannot sail dead upwind. So for each destination the tier builds candidate paths and takes the cheapest:
| Candidate | When it exists, and what it costs |
|---|---|
| direct | One leg straight at the destination, when the wind angle that course needs is outside the no-go zone. Timed at polar speed for that angle, plus one maneuver if it is on the other tack. |
| two legs | Out to the layline and back, both legs at the same wind angle and therefore the same polar speed, plus the maneuver between them — and one more to get onto the first leg if it is not the tack you are on. |
The maneuver cost is your own, in seconds, set in the app for your boat — a keelboat loses more in a tack than in a gybe, and the card shows what it charged so a navigator can judge the number it produced. Whether a change of tack is a tack or a gybe is arithmetic rather than assumption: crossing the wind through the bow is shorter exactly when the two wind angles add to less than 180°, which correctly calls a bear-away from close-hauled on one tack to a broad reach on the other a gybe.
Three destinations are evaluated — the point on the line square to the boat, when that point falls between the two marks, and each end — and the soonest wins. Ties keep the simpler, earlier candidate, so the answer does not flicker.
Two constraints keep the routes valid. A two-leg path’s turning point has to stay on the pre-start side: a route that crosses the line to get to the line is not a plan, it is an OCS. And a destination outside the cone the two laylines span rejects itself — the geometry solves it with a negative leg length, which would mean sailing backwards.
4 The burn, and where the three meet
The burn is the number a crew actually sails to: how much time you have to kill, or how much you are behind.
Read as a sentence: positive means you will get there early and have seconds to burn; negative means you are late for the gun.
The preference order is deliberately not the display order. Which tier is shown is a separate decision, and the number carries a badge saying which one is talking; what the burn is measured against is always the most informed number the boat could produce.
The burn rounds pessimistically in both directions, which is one decision rather than two: spare time is rounded down and lateness is rounded up. Six and a half seconds early reads +0:06; six and a half seconds late reads −0:07. A number that flatters the boat is the one number a start instrument must not print.
What the gaps mean
Because the three tiers differ only in what they assume, the distance between them is a statement about the boat rather than about the arithmetic:
- CURRENT minus TARGET is the speed you have not found yet, expressed in seconds.
- TARGET minus ROUTED is the angle and the maneuvering you have not made yet — the cost of not yet being on the path you are going to sail.
So the three converge as you commit. On the final approach — settled on the course you are starting on, up to target speed, no maneuver left to make — the routed path is the course you are steering, the target speed is the speed you are doing, and the three numbers collapse onto one. The gap that remains is what you still owe.
Caveat: the three are not guaranteed to arrive in that order at any given instant. The routed tier aims at a specific point on the line and charges for maneuvers, so on some geometries it is the largest of the three rather than the smallest. Convergence is a property of the approach, not an ordering the platform enforces.
5 SPEED — the same geometry, read backwards
The tiers answer the navigator’s question. SPEED answers the trimmer’s: how fast may we go?
It is the current tier solved for the other unknown. That tier says the time is the distance over the closing speed; hold the time at the time left to the gun, and solve for the speed instead:
Read as a sentence: the speed that puts the bow on the line exactly as the gun goes, on the course you are steering.
Worked on the same 100 metres, with a minute to the gun: pointed straight at the line the requirement is 3.24 knots. At 45° to the line it is 4.58 knots. At 60° — where only half your speed closes — it is 6.48 knots, exactly double the square-on figure. The line does not care how fast you are going, only how fast you are closing, and this is that same sentence read from the other end.
The identity that pins it
Because SPEED and CURRENT share one projection rather than being derived twice, they are exact inverses of each other:
Take the boat 100 metres out at 5.0 knots square to the line: 38.9 seconds. Put the gun at 38.9 seconds and ask what speed makes it, and the answer comes back 5.00 knots. One piece of geometry, read from its two ends.
Read as the inequality a crew actually uses, that identity says: while the burn is being carried by the current tier, being over the ceiling and having time to burn are the same statement in two units. The boat 100 metres out at 5.0 knots with a minute to run needs 3.24 knots and has 21.1 seconds in hand; give it 30 seconds instead and it needs 6.48 knots and is 8.9 seconds late. The two readouts move together, and they cannot contradict each other. (When a sharper tier carries the burn, the two are answering adjacent questions rather than one question twice — the burn has taken account of a route the required speed knows nothing about.)
A ceiling, not a target
Sail above the required speed and the bow crosses early. So the move it asks for is to ease down onto it, and the card colours it amber only when the boat is over it by a margin worth acting on — three tenths of a knot. Underneath, it sits in the ordinary live colour with no celebration attached, because being under the ceiling is the desired state rather than an achievement.
It also needs no measured speed at all. The course sets the projection and the gun sets the deadline; whether the speed log is alive changes nothing. And it is deliberately not a fourth tier — it does not share the tiers’ unit, it is never badged, it never carries the burn.
What this does not know
- It needs a gun. With no sequence armed there is no deadline, so there is no burn and no required speed. The cell keeps its label and shows an em dash rather than vanishing — a column that only appears mid-sequence is one the crew meets for the first time at the worst possible moment.
- The required speed stops meaning anything at the end. It is a hyperbola in the time remaining: halve the time and you double the requirement, so inside the last few seconds it moves faster than anyone can trim to it. Below five seconds the platform declines to print it. By then the boat is committed, and the burn beside it already says by how much.
- The routed tier assumes clean execution. It charges your maneuver cost and then times every leg at polar speed from the first second. It does not model the acceleration out of a tack beyond that flat cost, and it knows nothing about waves, traffic, or the boat to leeward of you.
- Two tiers need documents you have to have uploaded. No polar or no wind speed means no target tier and no routed tier. The card badges which tier is answering rather than substituting one for another, so the number on screen always says how much it knows.
- None of it is advice about where to start. The routed tier finds the soonest-reachable point on the line, which on a biased line is the wrong end tactically. It is a clock, not a tactician.
The worked numbers on this page are arithmetic on a stated geometry — a boat a round 100 metres from a line, at speeds chosen so the sums can be checked by hand. They are what the formulas produce, not a claim about any particular start.
6 What the start report does with these numbers
Every tier on this page is computed on the boat, before the gun, by Race Control. The start report prints them afterwards, and it prints them exactly as they were recorded: the burn seconds, each tier’s answer and the speed behind them are read from the recording second by second and served unchanged.
That rule is the whole reason those columns are worth reading. The purpose of the page is to show a crew the start they sailed, and a number computed afterwards is a number they never saw — it would explain a decision nobody made. Where the tablet drew a dash, the report prints a dash.
Only what nobody recorded is computed: the progress to windward at the gun, the polar’s target angle there, the seconds to reach ninety- five per cent of target boat speed and of best progress, and the acceleration rate. Those come from the boat’s ordinary channels through the ordinary calibration, and the report names which columns are which.