Part III - Racing Tactics and Strategy
Build 27 · 20 September 2026 · not yet reviewed
10Upwind Tactical Theory
Sections in this chapter
The upwind leg is where the largest and most repeatable gains of a race are made or lost, and this chapter sets out the decision theory behind it. It begins with strategy, the plan for the course without other boats: how a preferred side is chosen, how much separation it justifies, and what evidence overturns it. Two sections then treat the wind itself, first as an oscillation to be phased and then as a persistent trend to be positioned for, with the geometry that prices every shift against lateral separation. The middle of the chapter turns to the fleet: lanes and dirty air, the modes the boat uses to change its relative speed, and the crosses and engagements in which strategy and tactics meet. Dirty-air recovery and layline discipline handle the two most common ways a good plan is lost. The closing section assembles all of it into a process by race phase, from pre-start plan to post-leg review, so that each decision has a stated input, a rule and a test.
10.1Strategic side selection
Strategy is the plan for the course without other boats; tactics execute it among them S51. Side selection is the first strategic decision of every beat and the one that sets the value of every tactical decision that follows, because a lane, a cross or a cover is worth more on the favoured side than off it. Walker's framework sorts legs into three categories, an oscillating wind, one side obviously advantaged, and everything else, where the boat sails conservatively toward the middle and consolidates S62. The section works through how that sorting is made and then defended.
10.1.1Separating strategic bias from short-term tactical opportunity. Strategic bias is the from the wind and water field over the whole leg: more pressure, a favourable persistent shift, better current or a smoother sea on one side. Tactical opportunity is a gain available now against particular boats: a cross that can be made, a lane that has opened, a boat that can be pinned. Both are measured in distance to the mark, but over different horizons, and a tactical move is taken only when its gain exceeds the strategic value it spends. Dellenbaugh's percentage principle, many little gains rather than one big gain S51, cuts both ways: a flyer is a large strategic bet, and a tack to cover one boat off the favoured side is a small tactical bet with a strategic cost.
10.1.2Assessing pressure, shift potential, current, waves, and geographic effects. Inputs are ranked before the start. Pressure is read at , at least 200 to 500 metres away, scanning for stronger wind on one side and for backs and clocks S50; 30 to 40 per cent extra speed S50, so a pressure difference a small angle difference. Shift potential comes from the wind's origin: friction turns the surface wind about 30 degrees over land and less over water S39, and the left turn over land in the Northern Hemisphere often S40. Current is read from lobster pots, moored boats and the choppiness of the water S68. raises wave height S76, so the smoother side is faster. are crew value.
10.1.3Deciding when the preferred side justifies separation. Separation is the lever through which a side pays. A 5-degree shift changes the relative position of two boats by 12 per cent of their lateral separation S52, so the expected gain from a side is the expected shift toward it multiplied by the separation built, and the loss if the assessment is wrong has the same size with the opposite sign. Separation is justified only when the probability of the side paying is well above even and the evidence is of a kind that persists, pressure visible on the water or a geographic shift, rather than . In Walker's terms only the leg with one side obviously advantaged justifies separation; the other two categories call for the middle S62.
10.1.4Protecting access to the favoured side from the start and first cross. Position on the line follows the first-beat plan, pin end for the left side and committee-boat end for the right S51, and the line bias is overridden when one side of the course is strongly favoured S51. The price of the override is known: for every 5 degrees of bias to one end, that end is ahead by about one-eighth of the line's length measured along a close-hauled course S65, which the crew compares against what the side is expected to give. From the start the boat holds the tack that leads to the side until a clean lane exists on the other tack, and at the first cross it consolidates any gain by crossing ahead S51 rather than by extending.
10.1.5Knowing when to abandon a pre-race assumption. A pre-race assumption is abandoned on evidence that contradicts its mechanism, not on a single bad reading. If the side was chosen for a sea breeze, the invalidating signals are the absence of the offshore fill that moves toward land as a S51 and cloud thickening beyond the thin or broken cover Houghton's rule requires S43. If it was chosen for pressure under a cloud, the rule that wind decreases as a approaches, is null overhead and increases once it has passed S44 sets a time limit. If it was chosen on a compass trend, a return through the pre-start mean invalidates it. Each assumption is written with its before the warning signal.
10.1.6 Gold-medal standard: commit early, but retain an for changing plan. The observable standard is that the side call is stated aboard before the warning signal, that the boat starts at the end and on the tack the plan requires, and that the first tack away from the plan is preceded by a spoken observation: a compass number outside the expected range, pressure seen on the water, or a cloud signal. On the track, the boat's signed lateral distance from the line between the marks, which is the crew's log version of rather than the published one S17, grows toward the chosen side in the first third of the beat rather than wandering. In the debrief every change of plan is matched to a signal named before the race; a change with no named signal fails the standard.
10.2Oscillating shift management
The wind is almost always shifting back and forth, covering a range from a few degrees to as many as 40 degrees S51. Oscillating shifts swing around a mean direction and come from vertical instability S51, and the strategy in them is to tack on headers so as to stay on the lifted tack S51. That rule is simple to state and hard to apply, because each header has to be recognised against noise, priced against the tack it costs, and weighed against pressure, traffic and the boat's position in the fleet. The section fixes the measurements and thresholds.
10.2.1Establishing expected shift range before racing. The expected range is measured, not assumed. Bethwaite's measurements give directional with periods from about 3 minutes in heavy air to 10 minutes or more in light-medium air, and amplitudes from 5 degrees in light-medium air to 40 degrees in heavier air S50, which sets the order of magnitude by wind band. Before the start the boat sails each tack in turn and logs the close-hauled heading, the extremes on each tack and the time between them; the mean of the extremes is the reference direction. permits up to two compasses with heading, heading memory and timing functions S3, and a logger with shift tracking S79 confirms the numbers afterwards. The logged pre-start range is crew value.
10.2.2 Distinguishing meaningful shifts from . Not every compass movement is a shift: speed rises and falls by about 7 per cent every 6 to 12 seconds, and the surface is roughly S50. A gust bends the apparent wind and the boat heads up without the true direction changing, so a reading is a shift only when it persists through more than one pulse and appears on the compass, not only on the apparent-wind angle. The shift is then priced: a 5-degree shift is worth 12 per cent of lateral separation S52, so what counts is the shift multiplied by the separation, a small header far from the fleet can outweigh a large one close to it, and a header worth less than a tack is ignored. Threshold and hold time are crew value.
10.2.3Prioritising lifted-tack time without poor pressure or traffic. The lifted tack is the priority only when it can be sailed in pressure and clean air. Sailing in pressure is the key priority in light and medium air S63, and gusts carry 30 to 40 per cent extra speed S50, worth more than a small lift. Upwind a boat tacks away from one on or to weather of her bow within 6 to 10 boat lengths, the greater distance in lighter conditions S56, so a lifted tack that leads into a shadow is not lifted for long. In light and medium air the order is pressure, then clean air, then lift: a tack onto the lifted tack waits until the lane on that tack is clean, and a tack off it is made when the lane closes.
10.2.4Using headers to regain leverage rather than reacting emotionally. A header is a repositioning opportunity as well as a signal. When the boat is on the wrong side of the fleet, the header is the moment at which tacking back costs least, because the boat gains on every boat it now faces at 12 per cent of separation per 5 degrees S52. When the boat is already well placed, the same header is used to tack toward the next pressure or toward the middle. The emotional reaction is tacking on every wobble, or refusing to tack because the previous tack was wrong; both ignore the measurement. A tack is made on a header that meets the threshold, in a clean lane, toward a better position, and not otherwise.
10.2.5Identifying phase changes in an oscillating pattern. An oscillation has a period and a mean, and either can change. The period shortens as the wind rises, from 10 minutes or more in light-medium air toward 3 minutes in heavy air S50, so a building breeze brings faster shifts and a dying one slower shifts. The mean drifts when a persistent process is superimposed, for instance a sea breeze moving roughly 5 degrees per hour S51. The test is whether successive extremes return through the pre-start mean: if both tacks' extreme headings walk the same way round the compass, the lift on one tack getting better each time and the header on the other getting deeper, the mean is moving and the pattern is no longer a pure oscillation. From that point the persistent rule of the next section applies.
10.2.6Gold-medal standard: combine compass, fleet geometry, pressure, and position before major change. The observable standard is that no tack away from the fleet or off the lifted tack is made without a four-part call: the compass number relative to the mean, where the fleet is and which boats the tack crosses or faces, where the next pressure is on the water, and where the boat sits in the race. On audio this is heard as a short fixed sequence before every major tack. On logged data, tacks coincide with headings that exceeded the threshold for the hold time, and the number of manoeuvres on the beat S78 is no higher than the number of qualifying headers. A tack that follows fails the standard.
10.3Persistent shift management
Persistent shifts move one way and come from fronts, or shoreline effects S51. In them the boat sails toward the side closer to the expected wind direction S51, accepting the headed tack early so that it is on the lifted tack later. The gain is the same geometry as in the oscillation, 12 per cent of separation for every 5 degrees S52, but the shift does not come back, so a boat on the wrong side has no header with which to repair its position. Near the mark any shift is played as a persistent shift S51.
10.3.1Recognising signs of a persistent shift rather than an oscillation. The distinguishing sign is a trend in the extremes. In an oscillation the compass swings through the pre-start mean and the highest heading on each tack is reached again; in a persistent shift both tacks' extremes walk the same way round the compass, so each successive lift on one tack goes further and each header on the other goes deeper, without a return. The fleet confirms it: boats on the side of the new wind gain on every cross, not on alternate ones, and the lifted tack toward that side lasts longer each time. Pressure often moves with the shift. A shift is treated as persistent once the mean has moved by more than the oscillation logged before the start crew value.
10.3.2Using geographic and meteorological cues to validate persistence. A trend is validated when a mechanism explains it. A sea breeze needs a land-sea temperature difference of 5 degrees Celsius or more and rotates clockwise through the day in the Northern Hemisphere S42; Houghton's rule expects one when an offshore is under 25 knots with thin or broken cloud S43, and it fills first offshore and moves toward land as a mini-front S51. A brings a wind increase and a shift that arrives before the rain S44. Near a shore, friction turns the wind about 30 degrees over land and less over water S39. When the compass trend matches a mechanism, the shift is treated as persistent for as long as the mechanism lasts.
10.3.3Preserving the ability to take the next expected shift. Sailing toward the new wind is right only while the boat can still use it. Because a shift moves the mark's layline, overstanding is to be avoided S51, and a boat that reaches the layline before the shift arrives has converted all its leverage into distance it cannot use. The rule is to sail toward the expected wind on the headed tack until the lifted tack will lay the mark after the expected shift, not before it, and then to tack. The size of the expected shift sets how far short of the current layline that tack is made; the larger the expected shift, the earlier the tack. The margin held is crew value degrees.
10.3.4Avoiding late crossings that surrender the advantaged side. A crossing is late when it is made after the shift has begun to arrive. Ladder rungs lie perpendicular to the wind S53, and once the rungs rotate the boat crossing back from the wrong side has already given away, measured as distance sailed on a close-hauled course, 12 per cent of separation for each 5 degrees already shifted S52. The decision to go to the advantaged side is therefore made once and early, on the trend evidence above, and the fleet is crossed before the shift rather than after it. A boat caught on the wrong side when the shift arrives tacks immediately to stop the loss growing, rather than continuing to a corner hoping for a return the shift will not give.
10.3.5Managing fleet pressure when the strategic route becomes obvious. When the fleet sees the shift, the route becomes crowded and the constraint moves from direction to lane. A boat upwind and ahead throws dirty air over a boat within 6 to 10 lengths S56, so the lifted tack toward the new wind fills with shadows. The early boat holds the side by accepting a slightly worse lane rather than tacking off it: a safe-leeward position gives a lift and more pressure as the wind bends around the front of the sailplan S54, so a lane bow-forward and to leeward of the pack is sustainable. Boats behind sail a loose cover, between the fleet and the shift, rather than a tight cover that puts one boat in bad air and loses the side to the rest S51.
10.3.6Gold-medal standard: position for the next phase before the fleet confirms it. The observable standard is that on the tracked race the boat's tack toward the expected new wind precedes the fleet's, and that the crew's call naming the mechanism, sea breeze, cloud, shoreline or front, is recorded before the compass shows the shift. A second test is that the boat never reaches the pre-shift layline on a beat where a persistent shift was called, and that the lift it reaches on the final approach exceeds every lift logged earlier on that tack. A boat that tacks toward the new wind with the fleet, or after it, has met the shift but not the standard.
10.4Lane control and fleet geometry
Dirty air extends to leeward and aft in the direction opposite the leading boat's apparent wind S54, and wind shadows extend farther and slow more in light air S55. In a skiff whose apparent wind is well forward the shadow is thrown almost straight aft and to leeward, so the positions that are safe and the positions that are trapped differ from those of a slow boat. A lane is a corridor of undisturbed air on one tack, and this section treats how a lane is created, valued and traded against leverage.
10.4.1Value of clean air in a high-performance skiff fleet. A skiff makes its speed from its own apparent wind, and as boat speed rises the apparent wind increases and moves forward S26; a loss of pressure therefore compounds, because the slower boat also loses apparent wind. Upwind a boat tacks away from one on or to weather of her bow within 6 to 10 boat lengths S56, and shadows are longest and most damaging in light air S55. The 49erFX data study found that VMG on the first beat had a almost ten times that of any other skill S37, and clean air is the condition for that VMG. In a fast fleet a lane lost is rarely regained on the same tack.
10.4.2Creating lanes through early positioning rather than emergency tacks. Lanes are made before they are needed. On the line the lane is the gap to leeward held until the gun; tracking at the 2015 Europeans measured a of 10 metres and a starboard-tack advantage of under one second in time to the line S16, so the first lane is decided by the hole, not by the tack. After the start the boat tacks into a gap on the other tack as soon as one is visible and the strategic side allows, before the boats to windward tack onto it. An emergency tack is one made after the shadow has arrived; it costs the tack, the speed already lost and usually the choice of landing.
10.4.3Knowing when to protect a lane and when to trade it for leverage. A lane is worth ; leverage is worth the expected shift multiplied by separation, at 12 per cent of separation per 5 degrees S52. Protecting the lane means sailing a mode that holds it, high to keep a boat to leeward and aft from climbing into position, or low to stay clear of one to windward. Trading it means tacking off a clean lane for a strategic reason and accepting a worse lane on the new tack. The boat trades when the expected strategic gain exceeds the cost of the tack plus the difference in lane quality, and protects when the shift or pressure that would justify the trade is not yet visible.
10.4.4Using bow-forward and bow-back positions to control opponents. Four positions exist on the same tack. Bow-forward and to windward is control: the boat can apply a tight cover, directly upwind of the other boat's apparent wind, giving her bad air, or a loose cover that stays between her and the mark while leaving her clear air S51. Bow-forward and to leeward is the safe-leeward position, with a lift and more pressure as the wind bends around the front of the sailplan S54, and it pins the windward boat's bow. Bow-back and to windward pins the leeward boat on the tack, since she cannot tack across the windward bow, but sits in bent air and is left first. Bow-back and to leeward is free and keeps the option to tack.
10.4.5Avoiding low-probability crosses and trapped-lane situations. A cross with a margin under a length is decided by the next puff, and the port boat that loses it must keep clear under rule 10 S1 by ducking or tacking. The rule is to duck early when the cross is not clear, because a duck of half a length preserves the lane and the tack, while a forced tack under the bow lands in the shadow. A trapped lane is one on which the boat can neither continue, because the air ahead is dirty, nor tack, because a boat to windward and aft will land on her. The trap is avoided by tacking before the windward boat arrives, or by footing low for separation before the shadow closes.
10.4.6Gold-medal standard: treat lane access as a strategic asset to preserve early. The observable standard is that in the first minutes of the beat every tack the boat makes lands in a visible gap, and none lands within the 6 to 10 lengths that Baird's America's Cup class testing gives as the tack-away distance S56, measured to leeward and aft of another boat's bow. On tracked data the boat's number of manoeuvres on the first beat S78 is at or below the number of qualifying shifts, and its speed after each tack returns to target within the crew's stated time crew value. A tack made to escape a shadow that was visible a minute earlier fails the standard; a lane held with a mode change passes it.
10.5Tactical modes and relative boat speed
A mode is a deliberate departure from target angle and trim, and every departure from the optimum costs VMG, the principle Ingham states for the gennaker leg S27 and which holds close-hauled. Upwind the trade is between height and speed: the foils generate sideforce only when the boat moves somewhat sideways S19, and most racing boats sail upwind in medium breeze with 3 to 5 degrees of leeway S23, so a higher mode adds leeway and a lower one adds speed. Modes are tools with a purpose and a cost, chosen for an outcome and dropped when it is achieved.
10.5.1Using fast mode to escape, regain lanes, or extend. Fast mode is a slightly lower heading with the sheets eased to match, so that speed rises and the apparent wind strengthens S26. It is used to escape a shadow, whose direction is aft and to leeward of the leading boat's apparent wind S55, because footing leaves that cone faster than pinching; to reach a gap on the other tack before it closes; and to extend into a band of pressure ahead. Its cost is height, paid for as long as the mode lasts, and it is worth paying only while the outcome it buys is worth more than the height lost. The exit condition is stated when the mode is called.
10.5.2Using high mode to protect a side or hold an opponent. High mode is a higher heading with the sails trimmed to match, at a cost in speed. It holds a lane against a boat to windward trying to roll over the top, and it protects a side by keeping the boat between an opponent and that side without tacking. Against a boat to leeward and slightly behind, height denies her the safe-leeward position from which she would receive the lift and pressure of the wind bending round the front of her sailplan S54. The mode is sustainable only while the boat keeps enough speed for the foils to work; below that the leeway rises and the height gained on the compass is lost through the water.
10.5.3Knowing when pointing mode is tactically expensive. Pointing costs most where speed is hardest to rebuild. In leftover waves running at an angle to a new breeze, one tack is into the waves and needs its own mode S77, and pinching on that tack stops the boat in each crest. In light air the apparent wind falls with boat speed S26, so height taken from speed is not recoverable. In a lull within a surface that is roughly S50, the pointing boat stalls first. Pointing is also expensive when the boat has no strategic reason to hold the side, because the speed given up buys nothing. Height is bought only when the position it buys is worth the distance.
10.5.4Using low mode for speed, clear air, and controlled separation. Low mode is fast mode held for a strategic purpose rather than an escape. It builds lateral separation to leeward of a covering boat, and because the shadow points aft and to leeward of the covering boat's apparent wind S55, a boat far enough forward and to leeward is clear of it. It is the mode for sailing toward pressure seen to leeward and for the headed tack in a persistent shift, where distance toward the new wind matters more than height. It also serves in adverse current on one side, where the boat takes its medicine early and tacks away in the weaker current S68. Its limit is the separation the crew is willing to be wrong by.
10.5.5 Matching mode to race position and . Mode choice follows what the position needs. A boat ahead sails the mode that keeps it between the fleet and the mark and covers loosely, staying between the other boats and the mark while leaving them clear air S51; it does not use modes to make gains it does not need. A boat behind uses modes to make separation, because separation is the only way to gain on a fleet of equal speed, and a series that needs a result accepts more of it. Fletcher's rule for the start, that the plan weighs for the result the series needs S64, applies to modes: the riskier mode is chosen only when the series needs its reward.
10.5.6Gold-medal standard: use a mode for a defined outcome, then return to target mode. The observable standard is that every mode is called with a name and an exit condition, low until clear of the shadow, or high until the boat to windward tacks, and that the boat returns to target trim and angle when the condition is met. On logged data, periods in a mode are bounded, speed and heading return to the target within the crew's stated time crew value, and the S78 shows no long tail of below-target sailing after a mode. A mode that continues after its purpose has ended, or one called without a purpose, fails the standard.
10.6Cross management and engagements
A cross is where strategy and tactics meet: the right-of-way cases of rules 10 to 13 S1 decide who gives way, but the crew decides whether the cross is worth having. Dellenbaugh's rule is to consolidate a gain on other boats by crossing ahead of them S51. Every engagement costs distance to both boats, and the part of the fleet that is not engaged gains on both. The section sets out how a cross is priced, when the lee-bow is taken, when a boat is pinned, and when an engagement is refused.
10.6.1Evaluating crosses by margin, speed trend, and post-cross consequence. A cross is judged on three things. Margin is the distance by which the bow clears, measured in lengths at the moment of decision and reduced by the leeway of 3 to 5 degrees S23 on the crossing boat. Speed trend is whether the boat is entering pressure or leaving it, because a margin that is growing is a cross and one that is shrinking is a duck. Post-cross consequence is which tack and side the boat is on afterwards, and whether the crossed boat can tack onto her hip. A cross that lands the boat on the wrong tack for the strategic side is refused even when clear. Ducking early costs less than ducking late.
10.6.2Lee-bow positioning: control versus lost leverage. The tactical lee-bow is a tack close to leeward and ahead of a starboard-tack boat. Its effect comes from the wind: the leeward boat receives a lift and more pressure as the wind bends around the front of her sailplan S54, while the windward boat sits in the air deflected around the front of the leader's sailplan rather than in the disturbed air that extends aft and to leeward of the leader's apparent wind S55 and must tack away. The position gives control of one boat at the cost of leverage, since both boats are now committed to the same tack. Current plays no part: Perry shows there is no lee-bow effect from current and that a boat cannot gain by pinching to put the current on its lee bow S67. The two are kept distinct.
10.6.3Pinning an opponent versus preserving strategic freedom. A boat is pinned when she cannot tack without breaking rule 13 while tacking or rule 10 once on port S1 against a boat to windward and aft, or without landing in that boat's shadow. Pinning is worth having when the pinned boat wants to go the other way, toward a shift or a mark, and the pinning boat does not. Its cost is that the pinning boat is committed to the same tack for as long as the pin lasts; if the shift comes from the other side both lose to the fleet. The rule is to pin only toward the strategic side and to release the pin, by tacking, as soon as the pinned boat's escape would take her away from where the pinning boat wants to go.
10.6.4Bow-forward and bow-back choices in unstable wind. In an oscillating wind the winning position between two boats changes with each shift. Ladder rungs lie perpendicular to the wind, and the boat that has climbed higher is ahead S53; a lift favours the boat bow-forward and to windward, a header favours the boat bow-back and to leeward, and the exchange is 12 per cent of separation per 5 degrees S52. The choice at a cross is therefore made on the expected next shift: cross and take the windward position if a lift is due, duck and take the leeward position if a header is due. The compass trend of section 10.2 supplies that expectation.
10.6.5Avoiding engagements that force both boats from the preferred route. A tacking duel takes both boats off the lifted tack and away from pressure, and the third boat gains on both. An engagement is entered only when the tack it requires is one the boat would make anyway, on a header, into a lane or toward the side. Walker's conservative category, where the boat sails toward the middle and consolidates S62, is the default whenever an opponent tries to force an engagement in a wind with no clear side. A boat ahead breaks off a cover that has drifted off the strategic route and returns to a loose cover S51; a boat behind refuses the duel by splitting once, decisively, rather than repeatedly.
10.6.6Gold-medal standard: engage only when tactical gain exceeds strategic cost. The observable standard is that in the first two-thirds of the beat the boat makes no tack whose only purpose is to affect a single opponent unless that tack also goes toward the strategic side or onto the lifted tack, and that in the final third each cover or pin is on a boat that matters to the series score. On tracked data, every tack made within a few lengths of one opponent is matched in the debrief to a stated gain and a stated cost. A duel that leaves both boats further from the mark than the boats that stayed out of it fails the standard.
10.7Dirty-air recovery
Wind shadows extend farther and slow more in light air S55, and upwind a boat is affected on or to weather of another's bow within 6 to 10 boat lengths, the greater distance in lighter conditions S56. Recovery is measured in time spent in bad air, not in whether a tack was made. The section covers the earliest signs, the choice of escape, the discipline that prevents one escape from becoming several, the special case of a poor start, and the calls and roles that make the escape fast.
10.7.1Recognising earliest indicators of compromised air. The first indicator is geometric: dirty air extends to leeward and aft in the direction opposite the leading boat's apparent wind S54, so a boat whose bow is entering that cone will lose air before any instrument shows it. The next indicators are on the boat: the jib luff telltales lift at an unchanged sheet, trapeze load softens, the apparent-wind angle opens and speed drops before height does. Because a skiff's S26, the loss accelerates once it has started. The crew calls the shadow when the geometry is seen, before the speed loss, and the decision to escape is made at that call.
10.7.2Selecting immediate escape direction and mode. The escape is chosen by the shape of the shadow and by the strategy. If the other tack is clean and consistent with the plan, the boat tacks at once. If the other tack is worse, the boat foots low in fast mode, because the shadow points aft and to leeward of the leader's apparent wind S55 and a lower, faster course leaves the cone sooner than a higher one. Pinching to hold height inside the cone is the worst choice, since it holds the boat in disturbed air longest with the least speed. The direction is chosen in one call, and the mode is held until the boat is clear and back at target speed.
10.7.3Avoiding repeated tacks that compound speed loss. Each tack costs distance and the speed needed to rebuild apparent wind, and concluded that training should enhance manoeuvring efficiency, most prominently in strong wind S30. A boat that tacks out of one shadow into another has paid twice for nothing. The escape tack is made only into a lane that can be seen to be clean for the time it takes to regain speed; if none exists the boat foots instead. A second tack is made only on a shift or for a strategic reason, never as a second guess. Repeated tacks in dirty air are the most common way a bad start becomes a bad beat.
10.7.4Recovering from a poor start without overreaction. The 49erFX data study found that distance behind the line at the start lengthened the race and that first-beat VMG carried a almost ten times that of any other skill S37; a poor start is therefore recovered through VMG on the beat, not through a manoeuvre. The boat takes the first clean lane, often the port tack after a short duck of the boats to windward, and sails it in target mode toward the strategic side. Tacking immediately onto port into the fleet lands in successive shadows. The recovery is complete when the boat is in clean air on a tack consistent with the plan, not when it has regained places.
10.7.5Rebuilding a lane before returning to the strategic plan. Clean air comes before the plan, because a plan sailed in dirty air is not sailed at all. The sequence is escape, settle, reconnect: the boat leaves the shadow, sails long enough in clean air to return to target speed, and only then looks for the next shift or lane that carries it back toward the strategic side. A boat that tacks back toward the side while still slow arrives slow and last into the lane it wanted. If the lane it holds is on the wrong side, the boat waits for a header, which repays the return at 12 per cent of separation per 5 degrees S52.
10.7.6Gold-medal standard: use pre-agreed escape calls and recovery roles. The observable standard is that a shadow is named aboard with a fixed word and answered with a fixed action, tack or foot, with no discussion, and that the roles are fixed: the helm makes the tactical decisions and steers S9, while the crew, who is trained to initiate tactical decisions upwind S81, calls the shadow and the lane. The vocabulary is in the crew's protocol crew value. On logged data, the time from the speed drop to the recovery of target speed is short and repeatable, and the boat never tacks twice within a short interval in dirty air. A shadow argued about aboard fails the standard.
10.8Layline discipline
Laylines "are definitely a dead end when the wind is shifting" S51, and the rule is to stay away from the corners and laylines S51. A skiff closes on a layline quickly, so the period in which options remain is short and the discipline has to be applied early. Near the mark any shift is played as a persistent shift S51, because there is no time for it to return. The section treats the cost of early laylines, the three kinds of layline, traffic on the starboard approach and the rounding itself.
10.8.1Hidden cost of early laylines in a fast skiff fleet. A boat on the layline gains nothing from a lift, having already overstood, and loses on a header, having then to tack twice more or sail below the mark; because a shift moves the mark's layline, overstanding is to be avoided S51. The layline is also the tack every boat behind will make, so a boat that arrives on it early sails the rest in the shadows of boats tacking on her bow within 6 to 10 lengths S56. Current adds its own cost: when the current sets toward the windward mark the true layline lies short of the wind-only one S68, and a boat that judged it on wind alone overstands. A skiff reaches the layline in minutes, and an early layline is a long time without options.
10.8.2Approaching the windward mark with options. while the mark lies inside the cone between the two laylines. The approach is made from the side of the expected next shift, so that a header arrives as a lift on the final tack and a lift can be tacked on before the layline. The boat aims to reach the starboard layline late and with a small margin, on a heading verified on the previous tack, rather than to be on it early. The tacks left are counted: the boat keeps the option of one more tack for as long as it can, and gives it up only when the mark is laid on starboard with the margin the crew has fixed crew value.
10.8.3Conservative, leverage, and tactical laylines. Three laylines are distinguished. The conservative layline is the line from which the mark is laid in the current wind with a margin for a header, for leeway of 3 to 5 degrees S23 and for the boats that will tack on the approach. The leverage layline positions the boat inside a group on the expected shift, to the side of the new wind, so that the group has to cross her. The tactical layline arrives at the mark on starboard with rights under rule 10 S1 and with mark-room at the zone of three hull lengths under rule 18 S1. The crew names which layline it is sailing to; the three coincide only in a steady wind with no fleet.
10.8.4Managing traffic, disturbed air, and starboard-tack congestion. The starboard layline carries a queue of boats, each in the disturbed air of the one ahead. A boat approaching on port can tack into the queue under a boat's bow and take her shadow, tack above the queue and overstand, or duck the queue and tack behind it in clear air. The choice depends on distance to the mark: close to it the shadow lasts a short time and tacking under is accepted; far from it the duck is cheaper. A tack inside the zone in front of boats already on starboard is constrained by rule 18 S1, and in reasonable doubt an overlap is presumed not to exist S1, so a boat wanting the inside establishes the overlap clearly before the zone.
10.8.5Protecting the required final approach for a clean rounding. The rounding is sailed as a whole from the final tack. The classic rounding swings wide on the near side and cuts close on the far side, and an inside position approaching the mark is worth fighting for S51. considerably before bearing away to reduce leech and rudder loads S12, so the crew needs a short, clean stretch of close-hauled sailing before the mark for the ease and the set preparation. The final approach therefore avoids a tack in the last lengths, a duck at the mark, and any position from which the boat must luff for a boat inside. The required approach is fixed on entering the final third and defended from then on.
10.8.6Gold-medal standard: enter final third with a mark-rounding plan, not only position. The observable standard is that before the boat enters the final third the crew has stated the approach tack, the layline being sailed to, the expected traffic and the set that follows the rounding, and that this plan is heard aboard rather than improvised at the zone. On tracked data the boat's last tack is made below the layline by the crew's fixed margin crew value, the boat never sails above the mark on the final approach, and the number of tacks in the last part of the beat is no more than two. A boat that reaches the mark well placed without the rounding prepared has met the position but not the standard.
10.9Upwind tactical process
The sections above are executed as a process by race phase. The sequence matters because the inputs change: before the start the crew has time and no fleet, on the first beat it has a fleet and little time, in the middle it has evidence, and at the end it has geometry. The 49erFX study found that having a good start and an efficient first beat is the most important combination to sailing a shorter race S37, and Fletcher's method of working the approach back from the outcome wanted S64 applies to the beat as well as to the line.
10.9.1 Pre-start strategy: preferred side, , and risk limits. The pre-start plan states three things: the preferred side and the mechanism behind it, the observations that would invalidate it, and the risk the series position allows. Fletcher states that some starting approaches are riskier than others and that the plan weighs for the result the series needs S64; the same holds for the side. With a discard in hand the plan can accept committed separation on a strongly favoured side; without one, Walker's conservative middle S62 and Dellenbaugh's many little gains S51 set the limit. The compass log of the pre-start, with its mean and range crew value, is the first-beat reference.
10.9.2First-beat priorities: clean air, pressure, phase, and lane access. The first beat is sailed in a default : clean air, pressure, phase, lane. Clean air comes first because VMG on the first beat had a almost ten times that of any other skill S37 and none of it is available in a shadow. Pressure comes second because 30 to 40 per cent extra speed S50, and phase third because the lifted tack is worth 12 per cent of separation per 5 degrees S52, a distance, and it outranks a pressure band only where both have been priced in metres and it comes out larger. Lane access enables all three, and the first tack of the beat is made into a lane that serves them. The strategic side is pursued through these priorities rather than against them.
10.9.3Mid-beat priorities: leverage control and changing evidence. By the middle of the beat the boat has a position in the fleet and the wind has shown some of its pattern. Leverage is now controlled deliberately: the boat's , the unsigned extra distance sailed upwind from the course middle line on either side S17, is held within the value the plan allows and increased only on evidence of the kind set out in section 10.3. The evidence is re-read against the pre-start: whether the shifts return through the mean or trend, whether the pressure is where it was expected, and whether the boats that went the other way are gaining. Crosses follow section 10.6, consolidating gains by crossing ahead S51.
10.9.4Final-third priorities: layline discipline and rounding preparation. In the final third the horizon shortens and the rules of section 10.8 apply. Any shift is played as a persistent shift S51, so the boat tacks toward the side of a header immediately rather than waiting for a return. The layline is approached late from the side of the expected shift, the type of layline is named, and the traffic on the starboard approach is read before the boat commits. The rounding plan, approach tack, inside position and set, is stated before the zone of three hull lengths S1. Gains are no longer sought on the beat; the priority is to convert the position held into a clean rounding.
10.9.5Post-leg review: correct, incomplete, and misinterpreted information. Every decision on the beat is reviewed against what the information was and what was done with it. Three classes are distinguished: the information was correct and the decision followed it; the information was incomplete, in which case the review asks what observation was missing and whether it was available; and the information was misinterpreted, a header read as a persistent shift or a pulse read as a header. Tracked metrics, gap to leader, , number of manoeuvres and cross-track error S78, and the gain-and-loss comparisons built from log files S31, separate speed losses from decision losses. The review produces one change to the process, not a list.
10.9.6Gold-medal standard: concise, forward-looking communication focused on the next decision. The observable standard is that communication aboard on the beat consists of short calls that name an input and the next decision: a compass number and the tack it implies, a pressure band and the mode to reach it, a shadow and the escape. Nothing is transmitted from outside, because the Paris 2024 Notice of Race prohibited voice, radio or data transmissions from the first warning signal to the end of the last race of the day, except in an emergency or with equipment provided by the organisers S4, so every input is generated aboard within the division of labour in which the helm makes the tactical decisions and steers S9. On audio, no call refers to a decision already made except in the post-leg review. The protocol vocabulary is crew value.