Part III - Racing Tactics and Strategy
Build 27 · 20 September 2026 · not yet reviewed
11Downwind Tactical Theory
Sections in this chapter
The 49er sails its downwind legs under a gennaker at an apparent-wind angle that stays close to 90 degrees S27, so each gybe is a broad reach relative to the true wind and the fleet spreads sideways far faster than it does on a beat. That geometry governs the whole chapter. Section 11.1 sets out the geometry itself: separation rates, closure, leverage and lanes. Section 11.2 applies it to routing, where wind speed rather than wind angle usually decides the leg. Section 11.3 treats the gybe as the unit of downwind decision, with its costs and its triggers. Section 11.4 describes high and low modes as bounded departures from the VMG angle that buy a tactical result. Sections 11.5 and 11.6 are the two faces of boat-on-boat work, defence and attack, and both rest on the wind shadow and the covering gybe. Section 11.7 closes the leg at the leeward gate, and section 11.8 orders all of it into a process by leg phase, with a review method that assigns every loss to its source.
11.1Downwind fleet geometry
Skiff geometry downwind differs from displacement-boat geometry because the boat's own speed is a large fraction of the wind speed. Apparent wind is the of the true wind and the boat's course S19, and as boat speed rises the apparent wind strengthens and moves forward S26. The crew builds that apparent wind over the gennaker from early in the wind range S29, which holds the apparent angle near 90 degrees S27 and makes each gybe a reach. Tracking of an Olympic-class regatta found VMG higher downwind than upwind S30. Lateral spread is therefore rapid, and geometry becomes the first tactical variable of the leg.
11.1.1Apparent-wind geometry and separation rates. Because each gybe is sailed as a reach relative to the apparent wind, the true-wind gybe angle is wide and a large share of boat speed is directed across the course rather than down it. Two boats on opposite gybes therefore separate laterally at the sum of their cross-course components, close to twice each boat's cross-course speed, while their progress down the wind axis differs only through mode. Two boats on the same gybe separate only through mode differences, a few degrees of heading sustained over minutes. The separation rate is the rate at which leverage accumulates, and a crew that gybes away from the fleet is buying leverage at that rate whether it wants it or not. The crew's logged gybe angles by wind band are crew value.
11.1.2When small course-angle differences create decisive leverage. A shift is priced by lateral separation: a 5-degree shift changes the relative position of two boats by 12 per cent of the distance between them across the wind S52. Downwind that distance grows quickly, so the same shift is worth more between two skiffs a minute apart than between two displacement boats sailed for the same minute. A boat sailing a few degrees lower on the same gybe is also building separation, slowly, and a later shift prices that separation in the same way. The decision rule is that separation is a wager whose stake is the lateral distance and whose odds are the crew's expectation of the shift or the pressure; a crew that cannot state the expectation does not take the stake.
11.1.3Managing overlap and closure rates at high speed. Boats on opposite gybes close at the sum of their cross-course speeds, so the interval between sighting a crossing and the crossing itself is short and the give-way decision is taken early. Rules 10 to 13 set the right of way S1: the port-gybe boat keeps clear, and her options are to cross with margin, to duck, or to gybe before the crossing becomes forced. The port boat ducks by heading up, which cuts her progress along the starboard boat's course, or by oversheeting the gennaker to slow, the control Bithell describes for extreme conditions S29; bearing away below her course lengthens that progress and is the cross, not the duck. Overlaps for mark-room are judged at the three-hull-length zone, and in reasonable doubt an overlap is presumed not obtained or broken S1. Closure rates by wind band are crew value.
11.1.4Awareness of boats above, below, ahead, and astern. Dirty air extends to leeward and aft along the line of the leading boat's apparent wind S54, and in lighter conditions it can reach 8 to 10 boat lengths to leeward S56. With the apparent wind forward of abeam, that shadow falls to leeward and slightly aft of the boat casting it, so the threat is a boat to windward and slightly ahead whose masthead pennant points at the target W2, and the opportunity is a boat to leeward and slightly astern. A boat approaching on starboard gybe holds rule 10 rights wherever she is, and a boat below on the same gybe is the right-of-way boat under rule 11 S1. The helm keeps a four-sector picture current: above and ahead for shadow threats and for opposite-gybe crossings, below and astern for shadow targets, and below and ahead for a same-gybe boat that may gybe across.
11.1.5Selecting lanes that preserve pressure access and tactical room. A lane is a corridor of clear air with room to gybe toward pressure without at once meeting a starboard-gybe boat. Dellenbaugh's passing lane is "an area to windward of another boat (or boats) where you can sail by unobstructed" W1. Bethwaite's practice of scanning at , at least 200 to 500 metres, for stronger wind on one side S50 tells the crew where the lane should lead. Bithell states that in conditions "going low is tactically generally the safer option" W7, because the low lane avoids exposure to boats gybing onto starboard. The rule is that a lane must pass both tests, access to pressure and room to move, and a lane that fails either is a trap.
11.1.6Gold-medal standard: know whether height, speed, pressure, separation, cover, or gate access matters most. The observable standard is that at every phase of the leg the helm can state, in one word, which of the six variables the boat is optimising, and that the on the log matches that word: a height call shows as a heading above the fleet's mean, a pressure call as a departure toward logged higher wind speed, a cover call as a gybe two to three lengths from a named boat S57, a gate call as a course laid to the chosen mark. An observer replaying the track with the audio can label every course change with one of the six words, and the crew's own labels in the debrief agree with the observer's.
11.2Pressure routing versus angle routing
Two quantities can be routed for on a downwind leg: wind speed, which sets how fast and how low the boat can sail, and wind direction, which sets which gybe is headed. Dellenbaugh states that the biggest downwind gains and losses often come from differences in wind speed across the course S51, and that with better velocity a boat sails faster and lower W5. SailZing places sailing in pressure as the key priority in light and medium air S63. Angle routing, gybing on lifts to stay on the headed gybe S51, is the secondary lever, and this section states when each governs and how the two are traded.
11.2.1Identifying whether pressure or angle has greater leg value. Walker's framework sorts a leg into an oscillating wind, one side obviously advantaged, and everything else, where the boat sails conservatively toward the middle S62. Downwind the first category is angle-driven and the second is usually pressure-driven. The test is the ratio of the pressure gradient to the shift . Bethwaite's measurements put gusts at 30 to 40 per cent above the mean and directional oscillations at about 5 degrees in light-medium air, rising to 40 degrees in heavier air S50; in light and medium air the pressure term dominates and the boat routes for wind speed S63, while in heavier air the larger oscillations, and the smaller VMG return on extra wind for a boat already planing, move the value toward angle. In steady wind the long gybe, which points closest to the mark, takes priority W6.
11.2.2Using water, cloud, fleet movement, and sail behaviour as pressure evidence. Pressure is read from the water first: the surface is roughly S50, and the crew , 200 to 500 metres, for the darker bands S50. Cloud is read next: the wind falls as a approaches, is null beneath it and rises once it has passed S44, and looking upwind a cumulus gives a lefty on its left side and a righty on its right S41. Fleet movement is the third source: boats astern that heel, plane or heat up show pressure arriving before it reaches the boat. Sail behaviour is the last and most local: sheet load felt by the trimmer W7, and a lift declared when the helm has to head up to hold speed W5.
11.2.3Gain from pressure versus increased-distance cost. Routing toward pressure means sailing above or below the VMG angle for a period, and either departure from the optimum costs VMG S28. Bithell's warning that the high road can mean "sailing extra distance for not much extra speed gain" S29 is the cost side. of the between the two pressures and the time the boat spends in the better one before the fleet reaches it or the band moves on. Gusts carry 30 to 40 per cent extra wind speed S50, so in light air the VMG difference is large and short excursions pay; in strong air the boat is already fast, the difference shrinks and the same excursion loses. The crew's logged VMG by wind band crew value gives the break-even.
11.2.4Avoiding angle-driven gybes into weak wind or poor waves. A lift is the signal to gybe S51, but a gybe is a manoeuvre with a speed cost, and if it lands the boat in a lull the crew pays twice, once for the gybe and once for the wind. Wave state adds a third check. raises wave height and disturbance above normal and wind with current lowers them S76, and the water is smoother where wind and current are aligned S51. Leftover waves running at an angle to a new breeze put one gybe into the waves and the other across them, so each gybe has its own mode S77. The rule is that a lift alone does not trigger a gybe: the destination lane must hold at least the pressure and wave quality of the present one, or the shift must be persistent.
11.2.5Updating routing as apparent wind and waves change. Routing is not set at the top mark and left. As the boat accelerates the apparent wind moves forward S26 and the achievable depth changes, so the target angle drifts with pressure: higher in lulls to hold VMG, lower in puffs once speed has been built W6, and slightly low before an expected lull so that the lull can be sailed high without losing the mark W6. Each change of pressure or wave direction reopens the routing question, and the answer is restated aloud so that trim and heading move together. Near the leeward mark any shift is played as a persistent shift S51, because no time remains for an oscillation to return.
11.2.6 Gold-medal standard: take distance only when exceeds its cost. The observable standard is that every departure from the fleet's mean course or from the boat's own VMG angle is preceded by a spoken estimate, however rough, of what it should win, and that the logged track shows the departure ending when the pressure arrives or the estimate fails. On the replay, heading excursions coincide with logged wind-speed differences and not with their absence. A crew that routes to pressure it can see and returns to the target angle on reaching it passes; a crew whose track wanders toward one side without a stated reason, or that holds a high mode after the puff has been reached, fails.
11.3Gybing strategy
A gybe is the unit of downwind decision. On an asymmetric boat the gybe turns the hull through a wide angle, the apparent wind swings aft and collapses as the boat passes dead downwind, and the gennaker must be rotated and refilled before the apparent wind can be rebuilt. Dellenbaugh's strategic rule is to gybe on lifts so as to stay on the headed gybe S51. Tracking analysis of the 49er and 49erFX concluded that manoeuvring efficiency matters more in strong wind S30, and the 49erFX data study associated extra manoeuvres on the first and last legs with shorter race times S37, so the number of gybes is not the variable; their quality and their reasons are.
11.3.1Gybing for pressure, geography, waves, and gate position. Four reasons justify a gybe. Pressure is the first, because wind-speed differences produce the largest downwind gains S51 and pressure is the priority in light and medium air S63. Geography is the second: a shoreline bend, a current line, or adverse current crossed early while it is weaker S68. Waves are the third, when one gybe runs with a leftover sea and the other across it S77. Gate position is the fourth, the gybe that arrives at the chosen gate on starboard with speed and room. In steady wind, with none of these present, the long gybe that points closest to the mark holds priority W6, and the boat does not gybe for the sake of activity.
11.3.2Cost of a gybe in speed, control margin, and exposure. The speed cost is the loss of apparent wind through the turn and the time to rebuild it, since speed must come before depth S27. The control cost rises with wind: well before bearing away to unload leech and rudder S12, and in extreme conditions the boat can pitchpole if driven into the back of a wave at full speed S29, so a strong-wind gybe carries a capsize probability that a straight course does not. The exposure cost is the change of rights: a boat that gybes from starboard onto port must keep clear under rule 10 S1 and has left the fleet's cover. Metres lost per gybe by wind band, from the crew's logs, are crew value.
11.3.3Gybing before the fleet to gain leverage. An early gybe buys lateral separation at the rate set in section 11.1.1, and separation is what a shift prices at 12 per cent per 5 degrees S52. It is justified when the crew holds a specific expectation, pressure or a persistent shift on the far side, or when the boat is behind and the same-side race is already lost. It is not justified by impatience, because the boat that gybes first sails toward a corner, and the rule is to "Stay away from the corners and laylines" S51. In strong wind the manoeuvre itself costs more S30, so the expectation must be correspondingly firmer. The early gybe is a bet sized to what the series can afford.
11.3.4Gybing with the fleet to reduce risk and retain cover. Gybing when the fleet gybes holds lateral separation constant, so a shift prices only the separation already standing and no more; it is the percentage move S51 for a boat ahead. Gladstone's instruction for a covering boat when the rival gybes is "Do not delay" W2, because every second of delay converts to separation at skiff rates. The leading boat also consolidates a gain by crossing ahead of the boats it has gained on S51, and a simultaneous gybe after crossing keeps the position without giving the next shift a chance to reopen it. The cost is the gybe itself, paid at the same moment the rival pays it, so the relative cost is nil and the risk retired is the whole leverage of the leg.
11.3.5Delaying a gybe for a fast lane or passing opportunity. A gybe is held when the present lane is fast, in pressure and clear air, and the alternative lane is shadowed or in worse waves. Dellenbaugh's passing lane to windward of other boats W1 is worth holding until the pass is complete. The limit of delay is the layline: at the layline the boat loses any chance of using a shift W5, and the advice to avoid the laylines until the very end W1 applies. SailZing's default, in the absence of other factors, is not to gybe before the layline to the leeward mark W6; the two statements agree, because the other factors are exactly the pressure, shadow and traffic reasons that end a delay early.
11.3.6Creating a gybe plan before the top mark. The plan is built on the last part of the beat, when the helm can see the downwind course from above it: which side carries pressure, which gybe is long, where the fleet ahead will be sailing, where the current sets and which gate follows from the next-beat plan. It is stated as an exit mode, a first-gybe trigger and a gate. The trigger is a pressure line, a shift or a traffic point, not a time. The helm makes the tactical decisions and steers while the crew handles the sails S9, and the helm initiates downwind decisions S81; no voice, radio or data transmission was permitted while racing at Paris 2024 S4, so the plan is made on board and nowhere else.
11.3.7Gold-medal standard: each gybe has a reason, target lane, and next decision. The observable standard is that every gybe on the audio is preceded by a call that names its reason (pressure, shift, wave, traffic or gate), the lane it aims for, and the decision that follows it, the next trigger or the gate. On the replay each gybe coincides with a logged wind-speed or direction change, a crossing, or the final approach; a gybe that coincides with none of these is a defect. No gybe is made inside the layline except the last one to the gate, and no gybe is made whose logged cost by wind band crew value exceeds the gain that was stated for it.
11.4High mode, low mode, and passing opportunities
On an asymmetric boat the apparent-wind angle basically remains at 90 degrees S27, and the sailing angle relative to the true wind is set by how much apparent wind the boat has built: speed comes before depth, because once the boat is going well the apparent wind goes forward and the boat can go lower S27. Sailing higher than the VMG angle gives more speed and less depth, sailing lower gives the opposite, and either costs VMG S28. Bithell frames the choice as the high road or the low road S29. A mode is a bounded departure from the VMG angle that buys a tactical result, and this section states what each mode buys and when it ends.
11.4.1High mode for cover, wave access, defence, and mark approach. Heading up moves the apparent wind forward, raises speed and moves the wind shadow with it. It is the defensive mode: "The best defense against a boat threatening from behind is to head up in front of him" W1, and after a gybe, coming out hot pushes the apparent wind forward and helps escape an attack W2. It is the wave mode, because the extra speed catches a wave that is then ridden down its face at a lower angle. It is the mark-approach mode when a high line lays the mark and saves a gybe, which is expensive near a downwind finish S60. Its cost is sailing extra distance for not much extra speed gain S29.
11.4.2Low mode for pressure, clear air, passing, and separation. Bearing away trades speed for depth. It is the passing mode when a boat ahead and to windward has let her shadow fall astern: soaking, matching the rival's speed while sailing slightly lower W2, takes the boat under the shadow and ahead. It is the pressure mode once speed has been built in a puff and the boat can sail lower W6. It is the tactically safer option in conditions W7, because a low boat is not exposed to boats gybing onto starboard across her bow. It is the separation mode on the same gybe, building lateral distance a few degrees at a time. It fails when sailed before the apparent wind has been built, since depth without speed costs VMG S27.
11.4.3Recognising a low mode that is faster rather than merely lower. The test is VMG, not heading. A low mode is faster when it follows built speed S27, when the boat is heeled to windward with the windward chine in the water and the gennaker is not undertrimmed, because under-trim lets the leech rotate up S12, and when the boat holds its logged speed target for the wind band. It is merely lower when the gennaker leech breathes, the hull comes off the plane, or the apparent wind drops aft and the mainsail goes soft. No published 49er polars were found, so the crew's own logged speed and angle by wind band crew value are the reference, and Njord builds polar targets and gain-and-loss comparisons from those logs S31.
11.4.4Converting speed advantage into lateral or vertical separation. Gladstone's premise is that "You won't pass anybody if you aren't faster than they are (or at least as fast and more clever)" W2. A speed edge is banked in one of two directions. Vertical separation, along the wind axis, comes from sailing the VMG angle with the edge and is security: a boat ahead on the same line cannot be shadowed by the boat behind. Lateral separation comes from spending the edge on a mode and is leverage against the next shift or pressure band. The rule is vertical when the boat is ahead and defending, lateral when it is behind and needs a shift to pass, and a gain once made is consolidated by crossing ahead S51.
11.4.5Returning to target routing after tactical gain. Every mode has a purpose and therefore an end. Once the pass is complete, the shadow is off, the pressure is reached or the mark is laid, the boat returns to its VMG angle, because any continued departure costs VMG S28 with nothing left to buy. The return is a call, not a drift: the helm names the trigger that ended the mode and the crew retrims for the target angle, easing the gennaker and letting the apparent wind settle aft as the boat bears away. A mode that persists past its trigger is the most common invisible loss on the leg, because the boat looks fast from the coach boat while its VMG falls.
11.4.6Gold-medal standard: leave aggressive mode once its tactical purpose expires. The observable standard is that on the logged each mode departure is a bounded pulse with a stated start trigger and a stated end trigger, and that the heading returns to the VMG band within a few boat lengths of the end trigger being called. An observer with the audio can match every departure to its purpose and every return to the moment the purpose was achieved. A boat that sails high after the covering shadow has been placed, or low after the passing lane has been cleared, fails; a boat whose modes begin and end on named events passes.
11.5Covering and defence
Cover downwind is a position between a rival and the leeward gate, held with or without a wind shadow. The shadow tools are strong: dirty air can reach 8 to 10 boat lengths to leeward in lighter conditions S56, and wind shadows extend farther and slow more in light air S55. The covering gybe is made two to three lengths away, close enough for the shadow to work and too far for a luff S57, and a gybe within a length leaves the coverer vulnerable to a hard luff W2. Gladstone's constraint is that "You can't cover on both jibes" W2, so cover is a choice of side rather than a shadow held for the whole leg.
11.5.1Determining when cover is necessary. Cover is necessary when a specific boat behind can change the outcome the series requires and the remaining leg offers little further gain from routing. It is unnecessary when the boats behind are many, because covering one uncovers the rest, and when the leg still holds a pressure difference large enough to outweigh any position taken against one boat, since wind-speed differences produce the largest downwind gains S51. The percentage principle, many little gains rather than one big gain S51, applies in reverse: cover protects a gain already made and must not risk it. The decision is made before the leg, from the scoring situation, and is named aloud.
11.5.2Loose cover versus tight cover in high-speed fleets. A tight cover places the covering boat directly upwind of the other boat's apparent wind and gives bad air; a loose cover stays between the other boat and the mark but leaves her clear air S51. In the tight cover the covering boat's masthead pennant points at the rival W2 and the covering gybe is placed two to three lengths off S57. At skiff speeds the tight cover demands a manoeuvre for every response, and manoeuvre cost is what rises most in strong wind S30, so the loose cover is the default in a fast fleet and the tight cover is reserved for the last part of the leg, for light air, or for a rival whose one gybe would otherwise cross ahead.
11.5.3Covering opponent route rather than copying every movement. The covered quantity is the rival's route, the side of the course and the gate she is sailing to, not each of her gybes. A gybe that keeps her on the same side is ignored; a gybe that changes her side is matched without delay W2. Dellenbaugh's rule to consolidate a gain by crossing ahead S51 sets the geometry: the covering boat crosses, gybes on the rival's side and sails the VMG angle toward the gate. Copying every movement doubles the manoeuvre cost for the boat ahead and gives the boat behind a free gybe each time, so it converts a position of strength into a duel that the boat behind chose.
11.5.4Avoiding defence that sacrifices pressure or wave quality. A cover that leaves the pressure band loses more than it protects, because the rival in pressure sails faster and lower W5 and the position is undone within the leg. The rule is asymmetric: when the rival heads for the pressure, the cover follows; when the rival heads away from it, the cover lets her go and holds the pressure, since the gain from wind speed S51 exceeds what a shift can return to her. The same applies to waves: a covering gybe onto the gybe that runs into a leftover sea S77 pays the wave cost on every crest. Cover is worth only what the lane it occupies is worth.
11.5.5Defending a score without creating a high-risk duel. A duel of repeated gybes or a luffing match slows both boats and lets the fleet through, and it multiplies rule exposure: every close crossing is a question under rules 10 to 13 S1, and the 49er's penalty at Paris 2024 and the 2025 World Championship was one turn S4 S5, cheap for an attacker with little to lose and expensive for a leader. Defence therefore uses the minimum manoeuvre: head up in front of a boat threatening from astern, and gybe only when the shadow cannot be escaped by heading up, the two options Dellenbaugh gives W1. The defended score is protected by a loose cover and the VMG angle, not by engagement.
11.5.6Gold-medal standard: cover only boats and outcomes material to the event objective. The observable standard is that before each downwind leg the helm names the boat or boats, if any, that the series position makes material, and that on the replay every covering gybe and every mode departure toward another boat involves a boat on that list. A course change in response to a boat not on the list is a defect, as is a leg with a material boat behind and no cover decision stated. An observer finds the cover loose by default, tightened only near the gate or in light air, and never held at the expense of a logged pressure difference.
11.6Attacking and creating separation
Attack is the mirror of cover: the boat behind chooses separation where the boat ahead chooses proximity. Gladstone's summary is "Chase to the sides; attack when they come back" W2, and the mechanism of a downwind attack is the wind shadow, placed on the rival's apparent wind from windward and astern. Dellenbaugh names the common failure, a shadow that falls astern of the boat ahead W1. Passing also requires speed, or equal speed and better decisions W2. This section states when to separate, when to close, and how to hold a pass once it is made.
11.6.1Situations where separation has greater value than cover. For the boat behind, separation is the only way to change the order, because two boats on the same line finish in the same order. Its value is the shift or pressure it exposes them to, priced at 12 per cent of the lateral distance per 5 degrees S52. Separation is preferred when a persistent shift or a pressure side is expected, when the boat ahead is covering loosely and will not follow, and when the series requires a place that the present order does not give. It is avoided when the leg is short, the wind steady, and the boat behind is itself defending a place against boats further back.
11.6.2Using early gybes to access pressure or wave advantage. The first puff after the exit W6 is often on one side, and the boat behind can reach it first by gybing early while the leader is still settling on the exit gybe. Better velocity lets a boat sail faster and lower W5, and this converts directly into vertical separation on the leader. Waves offer the same: when a leftover sea favours one gybe S77, the boat that takes it first surfs while the leader reaches across the crests. The early gybe is an attack only when the destination is defined; an early gybe toward nothing is the corner Dellenbaugh warns against S51, taken at the moment the leader is least able to be hurt by it.
11.6.3Attacking boats constrained by poor air, room, or mode. The easiest targets are boats that cannot respond. A boat already in a shadow, within 8 to 10 lengths to leeward of another in light air S56, is slow and cannot head up into the boat above her. A boat on the layline has lost the shifts W5 and can only sail on. A boat on the wrong mode, high in a lull or low without speed, is losing VMG S28. The attacker lines up so that her masthead pennant points at the target W2 and takes care that the shadow does not fall astern W1. The approach is from windward and astern, on the same gybe, with speed in hand for the moment the target heads up.
11.6.4Converting a passing opportunity into a protected position. Once the pass is made the geometry reverses and the former attacker consolidates by crossing ahead S51. Crossing over and gybing gains control and the starboard-gybe advantage W2, and a gybe two to three lengths off the passed boat's air S57 holds her. Where the gate is near, the protected position is the inside at the chosen mark rather than a shadow, because the rounding order is decided at the zone S1. The conversion is done at once; a pass left unconsolidated for the sake of a slightly better angle is a pass that the next shift can hand back.
11.6.5Avoiding attacks that compromise the gate plan. An attack that carries the boat to the far side of the course from the chosen gate, or onto the outside of a pack at the zone, buys one place and loses several. The pinwheel at the right-hand gate mark can pin a boat indefinitely S59, and Hunt calls being pinwheeled outside into bad air and unable to tack away the worst possible outcome W4. The rule is that no attack is launched in the last part of the leg unless its finishing position still leaves the chosen gate and the inside of the zone reachable on the approach gybe; if it does not, the gate plan is changed first and the attack is judged again.
11.6.6Gold-medal standard: attack for defined pressure, wave, lane, or gate gain, not impulse. The observable standard is that every attack on the audio is announced with its object, the pressure band, the wave gybe, the lane or the gate position it is meant to win, and that on the replay the boat's course change points at that object rather than at the rival. An attack launched because the rival is close, with no object stated, is a defect even when it succeeds. An observer sees the attacking gybe placed so that the shadow lands on the target's air and not astern of it W1, and sees the pass consolidated by a crossing within the same phase of the leg.
11.7Leeward gate management
The leeward gate ends the leg and begins the beat, and its two marks are rarely equal. Benjamin and Perry's criteria are that the mark farther upwind is closer, the least crowded mark is preferred and the down-current mark turns faster S58; Hunt adds that, with the marks square to the wind, the boat rounds the mark that sends her the way she wants to go on the beat W4. The zone is three hull lengths S1, the pinwheel at the right-hand gate mark can pin a boat indefinitely S59, and the choice therefore has to be made while the boat still has the freedom to execute it.
11.7.1Choosing the gate before final approach becomes constrained. The gate is chosen in the middle of the leg, when both marks can still be reached on either gybe with speed. Benjamin and Perry's test is that "The mark which you can round soonest is probably favored for you" W3, and Hunt's is that the marks are the same size, so the closer one appears larger W4. Each later gybe removes an option: after the last gybe onto the approach gybe, only one mark remains reachable without a further gybe. The decision rule is that the gate call is made before the last strategic gybe, and that a change of gate afterward requires a stated reason larger than the cost of the extra gybe and the lost inside position.
11.7.2Gate choice by next-beat strategy, current, pressure, and traffic. Four criteria are weighed in order. The next beat comes first: the boat rounds the mark that sends her toward the favoured side W4, the right-hand mark as seen on the approach for the left side and the left-hand mark for the right, because each rounding exits on the tack that points that way. Current is second: the down-current mark makes an easier and faster turn S58, and rounding it puts the first tack in the current the beat plan needs. Pressure at the exit is third, judged from the water above the gate. Traffic is last and overrides all: the least crowded mark S58, and the opposite gate as an escape route when the preferred one is crowded W4.
11.7.3Preserving preferred gate through final gybes. The final gybes are laid to the chosen mark, not to the middle of the gate. Benjamin and Perry advise approaching the three-boat-length zone with the most speed possible to obtain or break later overlaps W3, so the last gybe is complete and the boat at full speed well before the zone. Dellenbaugh's compromise is to avoid the laylines until close to the mark and then gybe onto starboard just far enough from it to complete a good gybe W1; SailZing's default, absent other factors, is not to gybe before the layline to the leeward mark W6. A boat that holds its gybe past the layline sails extra distance; one that gybes short of it arrives on port with no rights.
11.7.4Avoiding late manoeuvres that sacrifice speed or add rule risk. A gybe inside the zone costs speed where speed decides overlaps, and it adds rule exposure: mark-room under rule 18 turns on whether the boats were overlapped when the first of them reached the zone, and in reasonable doubt the overlap is presumed not obtained or broken S1. A port-gybe approach across starboard boats puts the boat under rule 10 S1 at the point of highest traffic, and the penalty at Paris 2024 and the 2025 World Championship was one turn S4 S5, a turn made while the fleet rounds. The drop is timed so the gennaker is down before the turn without the boat slowing early; the crew's drop distance by wind band is crew value.
11.7.5Rounding with an immediate upwind lane and acceleration plan. The rounding is wide then tight W4: the turn starts outside the mark and most of the rotation is complete before the bow reaches it, so the boat passes the mark close-hauled S51. The exit priority is maximum VMG and boat speed with the sails correctly trimmed W3, which requires the daggerboard down, the vang and cunningham reset and both sailors on the wire within the first lengths. The lane is planned before the rounding: which boats will be to leeward and ahead, whether the first tack is immediate or delayed, and where the shadows of the boats ahead fall. In a crowd the boat slows early and waits its turn to round wide and tight W4, rather than being carried outside into bad air.
11.7.6Gold-medal standard: integrate gate choice with the next beat’s first two minutes. The observable standard is that the gate call on the audio is made before the last strategic gybe and comes with the first upwind decision attached: the exit tack, the first-tack trigger and the lane expected. On the replay the boat rounds the called mark, exits close-hauled and reaches its logged upwind speed target crew value within a few lengths, and makes its first tack at the stated trigger rather than in response to a shadow. A gate changed inside the zone, a rounding that exits below close-hauled, or a first tack forced by the boats around are defects; a rounding that flows into the beat plan without a pause is the standard.
11.8Downwind tactical process
The material of this chapter is ordered here as a process, so that the crew applies the same sequence on every leg and can locate in the sequence the point where a leg was lost. The phases are the top-mark exit, the early leg, the mid leg, the final approach and the review, and each phase has its own dominant variable: clear air and the first pressure at the exit, pressure and separation early, fleet geometry in the middle, the gate at the end. The review closes the loop by assigning every loss to handling, mode, tactics or environment, from the log and the audio.
11.8.1Top-mark exit: route, risk posture, and first-gybe trigger. The exit executes the plan made on the beat in section 11.3.6. Handling comes first, because well before the bear-away to unload leech and rudder S12 and the gennaker set must be clean before any tactical call is possible. Clear air is the priority near marks W6, so the exit lane is chosen free of the shadows of boats ahead, which reach farther in light air S54. The risk posture is stated: conservative when ahead and the leg is a consolidation, open when behind and separation is required. The first-gybe trigger is named before the set is complete, the first puff after the exit W6, a shift, or a crossing point.
11.8.2Early-leg phase: access pressure, waves, and separation. The boat builds apparent wind over the gennaker as early as the wind range allows S29, because speed comes before depth S27, and the crew , 200 to 500 metres S50, for the pressure band that decides the leg in light and medium air S63. The gybe is matched to the wave direction where a leftover sea favours one gybe S77. Separation is taken only against a stated expectation, and only at the rate the series can afford; the rule against corners and laylines S51 applies from the first minute. The early phase ends when the boat is in the pressure it aimed for or has established that none exists.
11.8.3Mid-leg phase: fleet geometry, cover, and opportunities. In the middle of the leg lateral separations are at their largest and every shift prices them at 12 per cent per 5 degrees S52, so this is where cover and attack are decided. The helm reads the four sectors of section 11.1.4, names any material boat to cover, and applies the covering gybe at two to three lengths S57 or the loose cover between rival and gate S51. Attacks are made on constrained boats and consolidated by crossing ahead S51. Routing is updated with each pressure change as in section 11.2.5. The phase ends with the gate call, which precedes the last strategic gybe.
11.8.4Final-leg phase: gate selection and round-up positioning. The final phase executes section 11.7: the last gybe laid to the chosen mark, full speed into the three-length zone S1 W3, the drop timed to the crew's logged distance crew value, and the rounding wide then tight W4 to exit close-hauled S51. Near the mark any shift is treated as persistent S51 and used only to lay the mark, never to justify a further gybe. Traffic overrides the gate choice at this stage, with the least crowded mark preferred S58 and the opposite mark as the escape W4. The phase ends when both sailors are on the wire on the planned exit tack with the boat at upwind speed.
11.8.5Post-leg review: handling, mode, tactical, or environmental source of loss. Every loss on the leg is assigned to one of four sources from the log and the audio. Handling losses show as speed dips at the gybes, the set and the drop larger than the crew's logged cost by wind band crew value; manoeuvring efficiency matters most in strong wind S30. Mode losses show as VMG below the crew's logged targets in the same wind, with Njord's polar targets and gain-and-loss comparisons built from the logs S31. Tactical losses show as shadows entered, gybes without a stated reason, or a gate call made late. Environmental losses are pressure or shifts the other side had, read from S78 and from , the extra distance sailed from the course middle line S17.
11.8.6Gold-medal standard: shared vocabulary for route, pressure, mode, attack, cover, and gate. The observable standard is that the audio of a downwind leg contains one word for each of the six concepts and no synonyms, that each word is followed by a value or a boat, and that the crew's action follows the word without a query. World Sailing's division has the helm making tactical decisions and steering while the crew works the sails S9, and no transmission from outside the boat was permitted while racing at Paris 2024 S4, so the vocabulary is the only tactical channel the boat has. The crew's own protocol defines the words; the test is that an observer transcribing the leg can reconstruct the route, the modes, the covers, the attacks and the gate call from the transcript alone.