If there's one thing I've noticed consistently over the years, it's how hard everyone finds it to work out the best speed to go upwind, especially on new gear or in a new class. Whether you're new to the sport or a seasoned sailor, getting upwind well is one of the hardest and most time-consuming things to nail. Which sail, which foil, how much power, how high to point, how fast to go? It's easy to spend whole sessions going up and down without knowing whether you actually made ground. On a relaxed sail, an efficient upwind leg means less effort, more fun, and getting there sooner. On the race course, it's the difference between rounding the top mark first and chasing the fleet all the way back to the bar.

One of the funniest races I've ever witnessed was a kitesurfing race with a whole field of novice riders. I'd just come in from my heat in second place when a friend borrowed my kite, and I gave him one piece of advice: "Whatever you do, don't go at top speed when you're heading upwind." Some job that advice did! The start was meant to be a close reach, sailing slightly upwind. Instead, the whole fleet launched and accelerated to top speed, with massive rooster tails, spray and board wakes everywhere. It seemed to start with one rider pulling away out front, and soon the entire fleet was racing at full tilt, legs burning and all. They flew past the reach turning mark, hundreds of metres too low, leaving those of us on shore wondering whether they'd been hit by a 30° wind shift. The real cause showed up when they tacked back to start their "upwind journey". Most of the fleet, rooster tails and all, barely made it back above the start line, all because their VMG at top speed was practically zero 😂😂. That ended up being a looooong race!

VMG stands for Velocity Made Good: your speed directly upwind or downwind, as opposed to the speed you're actually travelling. Sailing apps on watches, phones and tablets can now take the guesswork out of it. Many show live VMG, and some give you polar or radar charts or speed-versus-wind-angle graphs. You can check these while racing, or back on shore to see exactly what worked and what didn't.




Here's a polar from one of my recorded hydrofoil sessions. The wind comes from the top of the chart. Each dot is a moment in the session, plotted by its angle to the wind and its speed, and the yellow line traces the best speeds at each angle. VMG is the part of that speed going straight into the wind. Pinch too high and your speed collapses; sail too low and you go fast but don't gain much ground. The sweet spot here was about 40° off the wind at around 18 knots of speed over ground (SOG), giving about 13.5 knots of VMG on both tacks. So my upwind target on this gear in these conditions is simple: hold about 18 knots. If I'm slower than that, I bear away to build speed. If I'm well above it, I can point a bit higher.

The red lines show an 80° tack angle, 40° each side of the wind, and that's a little tight for this gear. Port tack holds its speed well, but on starboard I couldn't quite reach 18 knots that high. Sailing about 85° tack to tack would have kept my speed up and my VMG closer to its best, and less speed would have been lost in each tack.

You can also use the VMG number on the water. If your watch app lets you set the wind direction, point the wind axis accurately into the wind and it will read out a live VMG. There are a couple of ways to do this, but as a guide using the numbers from above: target speed 18 kts, VMG should be 13 - 14kts. When my VMG drops by more than 15% and it's a real header, if it looks like it will hold for more than 35 to 40 seconds and I'm not near the layline, it's worth a tack.






Every tack costs time and distance. The easiest way to measure it is in seconds: how many seconds of progress you give away. On my foil, a good 88° tack dropped me from 18 to about 10 knots and cost about 23 m upwind, which is around 3.3 seconds. A sloppier 74° tack in the same session cost 33 m, or about 4 seconds. Displacement boats lose fewer metres because they're slower, but about the same in seconds or more. A keelboat typically loses about one boatlength per tack in flat water and up to three in chop.
In one measured example, a 6 knot keelboat took 15 seconds to complete its tack but lost only about 5 seconds of upwind progress because it keeps moving towards the wind as it turns.

When should you tack on a header?
A header forces you further from the wind, so your VMG drops. Tacking onto the lift gets it back, but it only pays off if the shift lasts long enough to cover the cost of potentially 2 extra tacks.

Rough guide (when not close to the layline )
Multiply your tack loss in seconds by 11 for a 10 degree shift.
If you think the header will last at least that long in seconds, it's worth tacking into and out of again.

On my foil, a 10° header that costs me two extra tacks needs to last about 35 to 40 seconds before it pays off. Boats ahead and upwind can be a good indicator of what is to come. A keelboat in flat water needs about a minute, and one in chop needs 2 to 3 minutes. A dinghy with a slick roll tack in light air can profit from shifts of only 10 to 20 seconds. Smooth tacks and flat water let you use more shifts, while chop and stalled tacks mean only the big, long ones are worth chasing.

The best part is that you don't need to guess your numbers. A GPS debrief will show your target upwind speed, your VMG on each tack, and exactly how much each tack cost you.

*Full disclosure: I'm the developer of SailRacerPro, which does polar, VMG and tack-loss debriefs from any GPX or FIT file, plus Garmin watch apps for sailing, kite, windsurf and wing. Happy to run anyone's track through it if you'd like to see your own upwind numbers, or try it out for free yourself at www.sailracerpro.com

JonE said..
First up, you cant sail to vmg when heading upwind. If you steer the boat (up) to get a higher number, the number goes really high, then the boat stops.

Second you're only getting COG not heading from the GPS.

If meaningful data could be deduced by gps alone and occasionally going head to wind, we would have had it for 20 years.

On a lake with a sub 5 meter mast this might be useful. Everywhere else we have leeway, wind shear, tide, wave direction. So many reasons why you might go faster, or think you're going faster on one tack vs another.

Wanting to judge tacking performance is worthwhile but without having fantastic true wind angle information you cant make those claims about the cost of a tack based on shifts.

true wind is not possible without good STW, realtime wind, heading and calibration tables.

So 15 minutes a week spent on tacking practice with a stopwatch is going to be time better spent?

Thanks JonE, good points and I agree with a few of them.

You're right that you can't steer to VMG upwind. Chase the number and you pinch, and the boat stops. That's why I steer to a target speed (18 kts in that example) and only use the live VMG as a header alarm. If speed holds but VMG drops 15% or more, the breeze has probably shifted.

Also agreed that GPS gives COG, not heading, and I'm not claiming true wind angle, leeway or proper instrument-grade true wind.

Where I'd push back a little is that for racing, the ground is what counts. Leeway, tide and current are all baked into COG and SOG, and they're exactly what decides who gets to the top mark first. Comparing within one session (same boat, same tide state, similar breeze) still shows which tack angle and which speed got you upwind fastest over the ground. The wind axis comes from averaging your own tacks, so it's an estimate, and a tide change or big shift shows up as noise you can see in the track. But what if the shifts tend to be in the same spot, as you approach the shore or between island for example, you see it and take advantage of it.

Same with tack loss. Metres of ground made good lost per tack is a fair relative measure. It won't tell you why one tack was worse, but it tells you which ones were.

15 minutes a week of tacking practice with a stopwatch is great. Or just record it on the app which adds a GPS track, so you can see which of those tacks actually cost less ground rather than which one felt better

Happy to run one of your sessions through tracks through it if you'd like to see where it holds up and where it falls down.
The closing offer is the important part. If he takes you up on it, it turns a sceptic into a public test case.

I dont get why you cant sail to best VMG upwind. I always have in racing. If you get too close to the wind or pinching VMG drops and you need to turn slightly down. I spent time before racing measuring lat and long of each mark so VMG to the mark would be accurate. For example, 15 knots at 45° produces the best VMG to the mark, even though the boat can sail faster at 50° so sail at 45. I won trophies doing exactly that. Choosing when to tack can be tricky. What am I missing?

And it's not VMG to the mark that's CMG.

VMG doesnt tell you when to tack and CMG doesnt until you hit the layline,

VMG just quantifies the best compromise between pointing and boat speed but the GPS can't tell you VMG because it doesnt know where the wind is.

This is why you have to create polars by which you need to rigorously measure VMG at different angles whilst doing maximum boat speed and plot it. Someone cite me one reliable reference to how to create polars without accurately calibrated true wind, STW and heading. I have read everything I can get my hands on and it appears there is no shortcut. Hence my disagreement with this "advertorial" post.

JonE said..
And it's not VMG to the mark that's CMG.

VMG doesnt tell you when to tack and CMG doesnt until you hit the layline,

VMG just quantifies the best compromise between pointing and boat speed but the GPS can't tell you VMG because it doesnt know where the wind is.

This is why you have to create polars by which you need to rigorously measure VMG at different angles whilst doing maximum boat speed and plot it. Someone cite me one reliable reference to how to create polars without accurately calibrated true wind, STW and heading. I have read everything I can get my hands on and it appears there is no shortcut. Hence my disagreement with this "advertorial" post.

Predict wind data hub claims it will create AI polars for your boat over time based on actual performance of your boat.

S%$# in, S%$# out.
I loved poring over my GPS plots post race and always had VMG/VMC targets in racing. It was always worthwhile playing back and reviewing your overall race position and strategy and performance against polars.
But I also logged TWA,TWS, Heading, SoW, Heel, Pitch, Heave, Leeway, Barometric pressure and Density Altitude. And used all of that data religiously doing any review. None of those datasets you get with a GPS driven plot alone.

It's fun to use GPS to roughly estimate polars. But no way would I use GPS driven polars in an actual race, it's missing all of the other data that makes sailing to numbers worthwhile. Any decent boat has multiple polars for different sea states/environmentals, the different polars come from all of the data you don't get with just a gps device and a bit of software.
I get it when hydrofoilling at 20knots plus. You're not offshore, flatter water and less tidal influence. But as your speed drops, the higher the propensity in erroneous data.
For keel boats doing a passage race? I'd posit a post race function where you can review a GPS trace on a chart might be more beneficial.

Not trying to diss on you DaChopper. If peeps are doing keelboat racing without electronics, or play in OTB pickleforks or suchlike, then knock yourself out.
But polars derived from a GPS plot would be useless for me in keelboat racing (sans WAGs), especially using that data to affect decisions in the actual race itself. Way too many missing bits of info.
My two cents!

Trek said..
Quote from Sailing World:

America's Cup yachts race primarily to achieve the best VMG (Velocity Made Good), rather than maximum boat speed.
This applies both upwind and downwind.
1. Upwind sailing
The AC75 foiling yachts balance pointing angle against speed to maximise progress.
2. Downwind sailing
The same principle applies downwind.
AC75 yachts sail at angles to the true wind rather than directly downwind because their foiling speed generates strong apparent wind. They gybe between angles that maximise downwind VMG.

From David Dellenbaugh, tactician aboard the winning America³ team in the 1992 America's Cup.
“One combination of pointing and speed will maximize your VMG (your velocity-made-good) to windward.”

From Peter Isler — America's Cup winner - explaing his success “I experimented with sailing deeper angles downwind and finding the best downwind VMG.”

From Tom Slingsby's actual race communication, “Happy sailing VMG,”

But that's VMG to the wind direction, not the course to the mark.

Quite right. But for the technical discussion of VMG Im considering wind and speed to the mark only. Of course there are wind shifts, currents, dirty air, swell, chop and competitors boats in the way.

Shaggy,

I haven't got my **** together with true wind yet, despite having h3000 and good quality, if dated, gear that will do it.

for mine, and everyone else's edification- roughly how many minutes would you guess you spent tuning in order to get a decent true wind number? From what I've read, it looks like a solid few hours recording and tweaking plus a good 10-15 minutes before every race to tune it for conditions.