Match speed to bait type first, then correct for current and sea state. A lure needs 6-9 knots through water, dead bait 4-6 and live bait 1.5-3.5.
One extra knot changes the bait more than the boat
One knot is 0.51 metres per second, which sounds small until it is applied to a bait already working near the edge of its stable range. A skirted lure at seven knots travels 3.6 metres every second. At eight knots it travels 4.1 metres and the same head that tracked straight begins to yaw, then blows out and skips. The boat has gained half a knot of progress; the lure has lost its action entirely.
Every speed table in the sources is organised by bait rather than by species. Tackle Room's offshore speed guide puts ballyhoo at 6-8 knots, spoons at 4-6 and high-speed wahoo lures at 12-16. For yellowfin tuna on skirted lures, its separate spread article gives 6-9 knots. TrollTrue's table runs from live bait at the slowest end through cigar minnows at 4-5 knots and lipped hardbaits at 5-8. The bands overlap because the baits overlap, and the authors agree more than the numbers suggest.
The apparent conflict in king mackerel numbers resolves the same way. Tackle Room lists 4-6 knots for king mackerel trolling, while Gomdori's slow-troll guide gives 1.8-3.5 knots. One is written around lures and spoons, the other around live blue runners and cigar minnows on a stinger rig. A king will eat either presentation, but a live bait dragged at five knots is a dead bait that has not noticed yet. Speed follows the bait in the water, and the species only sets the upper limit of what it will chase.
The practical consequence is that speed is a per-rod decision before it is a boat decision. Four lines carrying four bait types cannot all be correct at one throttle setting. Either group the spread by speed band, or accept that two of the four are compromises and check them first when the bite stops.
A speed band for each bait type
Live bait occupies the bottom of the range. Gomdori puts king mackerel slow trolling at 1.8-3.5 knots depending on bait size, current and the mood of the fish, and Tackle Room's live-bait guide gives 1.5-3. When a blue runner is dragged faster than about 3.5 knots it spends its energy trying to stay upright, and a bait that is fighting the boat is not swimming in a way that invites a strike. Above four knots most baitfish either skip at the surface or roll, and a rolled bait is a wasted line.
Dead bait sits in the middle. A chin-weighted ballyhoo needs enough speed to hold a straight track and make the skirt breathe, which is the 6-8 knot band in Tackle Room's table. A naked cigar minnow or a small pilchard on a planer can run slower, around 4-5 knots, because there is less drag and less hardware to destabilise it. Spoons flutter at 4-6 knots and lose their wobble above that. The bait's own geometry sets the floor and ceiling, and both are narrower than most anglers assume.
Artificial lures cover the top. A speed-trolling skirt has no lip and is built to run straight, and 10-12 knots is normal for that class rather than extreme. Bullet-heads and jet-heads for wahoo run at 12-16 knots, which is a different style of trolling with heavier heads and shorter leaders. A lipped hardbait at those speeds will either spin or dive to the surface and trip. Buying a lure without knowing its design speed is how a good spread ends up with three baits that cannot be fished together.
Adjust the band by half a knot for conditions rather than rewriting it. Cold water in the 60s Fahrenheit usually wants the slower half of each band, because baitfish metabolism and tuna aggression both drop. A flat sea with no current allows the upper half. A confused sea forces the lower half because the boat cannot hold a steady speed. Depth gear reopens the same question, and downrigger selection by depth sets out the hardware side of that trade. Keep those moves to half a knot rather than a whole one, and log them next to the catch.
Current moves your real speed more than the throttle does
The bait feels speed through water, while the GPS reports speed over ground. In still water the two are the same. In a 1.5-knot current they diverge by 1.5 knots, which is larger than most of the adjustments anglers make all day. Trolling with the current at 7 knots over ground means the lure sees 5.5 knots. Trolling into it at the same GPS reading means the lure sees 8.5. One direction is inside a ballyhoo's band and the other is outside it, with no change on the throttle.
Worked example: a boat trolling north along a Gulf Stream edge reads 6.8 knots over ground into a north-setting current of 1.2 knots. The lures are actually running at 8.0 knots, and the ballyhoo that worked on the calm morning now spins. Turning south, the same throttle gives 8.0 over ground and 6.8 through the water, and the baits settle. If the crew only records the GPS number, the southbound pass looks faster and the northbound pass looks slower, when the water speed is identical.
The correction is simple arithmetic: water speed equals speed over ground plus the component of current pushing against the boat. Measure the current rather than guessing it. A quick drift with the engine in neutral for 60 seconds gives a speed and direction; a chartplotter's current overlay gives a second estimate. Around rips and inlet mouths the surface current can be one to two knots even when the offshore forecast says calm, and it can reverse between the ebb and the flood.
Run the spread in both directions when the current is strong enough to matter. If the baits only work on one heading, the speed is wrong on the other, and the fix is a throttle change of 0.5-1 knot rather than a new lure. Recording speed over ground on both passes also builds a local correction table for the next trip, which is more useful than a single number copied from a magazine.
Waves and side wind set the real limit
A head sea costs speed before it costs comfort. In a two-foot chop with 12-15 knots of wind, a 21-foot centre console typically loses 1-2 knots of achievable speed at the same throttle, because the hull climbs and lands instead of tracking. Trolling a 6-knot lure band into that sea may require an indicated 7.5 knots just to keep the baits tracking. The engine is working harder and the lures are still correct, which is why throttle position is a poor speed reference.
A following sea does the opposite and is more dangerous for lure action. The boat surfs, speed over ground swings by 1-2 knots in each wave, and the lures alternately slack and snap. An autopilot correcting heading in a quartering sea can produce the same surging without the helmsman noticing. Watch the rod tips during a following-sea pass: if they unload and reload once every wave, the spread is being cycled through the top and bottom of its speed band, and a small throttle reduction will stabilise it.
Side wind matters most for slow trolling. Live bait at 2.5 knots in 15 knots of beam wind gives the hull enough leeway that the baits trail at an angle, and the lines closest to the wind climb while the leeward lines sink. The speed is still right through the water, but the spread geometry is not. Either take the wind on the bow or shorten the leeward lines; adding throttle to straighten the spread only drowns the bait.
Set a personal upper limit before the trip. For a 21-23 foot boat, 15 knots of sustained wind with two-foot seas is a common working ceiling for offshore trolling, and 20 knots with three-foot seas is a different day. The numbers are not universal, but the principle is: the sea state sets the fastest speed the boat can hold steadily, and the bait sets the slowest. When those two cross, the spread cannot be fished correctly and the trip should move inshore or wait.
Wake, rod tip and bait as speed instruments
The wake is the cheapest speed instrument on the boat, and the same reading drives line lengths in building a trolling spread. At 6-8 knots a displacement hull leaves a clean V with a modest white crest and the transom settles evenly. Above 10 knots the crest widens and the wake starts to curl; above 14 knots a small hull throws a steep white wall and the prop note rises. None of this gives a number, but it gives a repeatable picture. When the wake looks like it did on the pass that produced fish, the throttle is close.
The rod tip reports the bait. A ballyhoo running correctly produces a steady load with a small, fast pulse, roughly one to two cycles per second. A lure that has blown out makes the tip go slack and then snap as the head catches again. A bait dragging a ball of weed shows a heavy, dead load with no pulse at all. These are different signatures, and they are visible from the helm without picking up a rod.
The bait itself is the final check. At the correct speed a live bait swims with its head into the current and its tail beating; too fast and it skips or porpoises; too slow and it tangles the leader or wraps the line. A dead bait should track straight with the chin weight leading and the skirt breathing. A skirted lure should leave a clean bubble trail with no periodic eruption at the surface. Pull the spread every 20-30 minutes and look at the baits, because a lure that has collected a single strand of weed can look perfect on the rod tip.
One failure mode deserves its own line. A lure that blows out at eight knots may be perfectly stable at six, and anglers often retire it as defective. Before discarding it, run it beside the boat at 5, 6 and 7 knots and watch the head. Many soft-head lures have a narrow window, and finding it turns an unused lure into a productive one. The same test applies to a ballyhoo that spins: check the chin weight and the hook alignment before blaming the speed.
Calibrating two speed readings on one boat
Most boats carry two speed sources that disagree: GPS speed over ground and a paddle wheel or transducer speed through water. The paddle wheel is the one that drifts. Fouling, a bent paddle or a bad calibration can make it read 10-20 percent low, which at 7 knots is 0.7-1.4 knots of error. Clean it at the start of the season, then calibrate it against a measured distance rather than against the GPS on a current day.
Pick a one-nautical-mile stretch with known marks, run it in both directions at a steady 7 knots, and average the times. Running both directions cancels most of the current. If the average run takes 9 minutes 30 seconds, the true speed is 6.3 knots, and the paddle wheel can be corrected to match. Do the same at 4 and 10 knots if the unit allows a multi-point calibration; the error is rarely linear.
Keep a log by bait rather than by throttle. A page with columns for bait type, GPS speed, water speed, heading and result will outperform any generic table after three trips. The useful entry is not '7 knots caught fish' but 'ballyhoo, 6.4 knots through water, northbound into 1.2 knots of current, two strikes'. That line tells you what to repeat. The same log exposes a spread that only works on one heading.
The common failure is calibrating on a calm day with no current and then trusting the number in a rip. Calibration removes instrument error; it does not remove current. Treat the paddle wheel as a check on the GPS rather than a replacement for thinking about set and drift. When the two disagree by more than about 0.5 knots in water you believe is slack, one of them is wrong, and the bait action will tell you which.
Short version
- Set speed by bait type: live bait 1.5-3.5 knots, dead bait 4-6, skirted lures 6-9 and high-speed lures 12-16 through the water.
- Correct speed over ground for current; a 1.5-knot current changes the bait's speed by more than most daily tuning does.
- Read the wake, rod tip and bait track, and calibrate the paddle wheel over a measured mile in both directions.
- Log speed by bait and heading rather than by throttle, because the same setting produces different water speed on each tack.
Sources
- Why Am I Not Getting Bites Offshore? Spread, Speed, and Depth Fixes — Tackle Room · T3
- Trolling Lures for Saltwater Fishing: Types, Rigging & Spreads — Tackle Room · T3
- Best Offshore Trolling Speed by Bait and Species — TrollTrue · T3
- King Mackerel Slow Trolling: Live Bait Spreads and Stinger Rigs — Gomdori Fish Camp · T3
- King Mackerel Fishery Management Plan Update — North Carolina Department of Environmental Quality · T1