Suppose you are running a 5k race on an out-and-back course, and there’s a strong headwind on the way out—should you aim to run at the same effort the whole way, allowing the wind to slow you down on the way out and speed you up on the way back? Or should you maintain the same speed, working harder on the way out and relaxing more on the way back?
For most runners, it feels intuitive that constant-effort running should be better—in the sense of leading to better performance—than constant-speed running: constant-effort pacing is better than speeding up and slowing down, after all. But here’s a potential counterargument: if you run faster when going into the wind, you spend less time exposed to the wind, so you’ll save energy on balance.
So, which argument is correct?
Here’s the answer: in most cases it is better to run at a constant effort—but not always! If you are running relatively slowly (slower than about 8:00/mi or 5:00/km) and you are facing a wind of about 11 mph (18 km/h) or more, it can be better—sometimes significantly better—to run at a constant speed.
At faster running speeds, though, constant effort always wins. The benefits can be substantial—at 4:30/mi (2:45/km) in a 13 mph (22 km/h) wind, constant effort running results in an 1% reduction in average metabolic power output: about one-third the benefit of a typical pair of super shoes.
Do note that this is specifically in the context of attempting to run a specific goal time (or maintaining a specific average pace). Hopefully, it should be obvious that running at a constant effort is the right move if you are trying to maintain a specific physiological state, for example to maintain a metabolic steady-state when doing a threshold workout.
Proving this relationship between speed and effort in windy conditions is not trivial. It requires some careful calculations that balance the energetic costs of running against the metabolic effects of headwinds and tailwinds.
Below, I’ll sketch out what the calculations look like and show some plots demonstrating the two countervailing effects that lead to different optimal strategies depending on your running speed.
The conceptual setup: constant speed versus constant effort in the wind
The reason that running into a headwind is more energetically expensive than running on a calm day is fundamentally due to the additional drag forces on your body.
A headwind applies a horizontal opposing force to your chest, and there’s an additional metabolic cost to overcome it. Conversely, in a tailwind, you get an assisting force, so you do not need to work as hard to propel yourself forward. I’ve got a whole write-up on the biomechanics and physiology of air resistance in running here if you’re interested.
Since drag forces increase according to the square of relative airflow, you always run slower—or have to work harder—on courses that start and finish in the same place, because the headwind slows you down more than the tailwind speeds you up. But knowing that fact doesn’t tell you the most optimal way to pace yourself in the wind if you have some goal time for the course.
As I mentioned in the intro, there’s a countervailing factor that’s at play when you are trying to run a set time over a set distance (e.g. a 20:00 5k at a park run) – if you run at a constant effort, you end up spending more time going into the wind, and less time going with the wind. This could counteract the benefits of allowing the wind to speed you up or slow you down.
Simulating different windy-day strategies
I’ve heard this argument (“run harder into the wind so you spend less time in it”) often enough that I’ve wanted to actually test it out for a while now, and thanks to the research I’ve been doing over at my apps site, I finally have the tools to do it.
To empirically answer the question of whether to run at a constant speed, or run at a constant effort, on a windy course, I set up a series of tests using the model that powers my headwind and tailwind calculator, as well as my track wind calculator.
The “test course” was a 1000 meter out-and-back run: 500 m straight into the wind, then 500 m back to the start with the wind. We ignore any time delays from the turnaround.
Since this is all just computer code, we can conduct dozens of tests across different running speeds and wind conditions. I swept a grid of running speeds from 3.0 to 7.0 meters per second (8:55/mi or 5:35/km down to 3:50/mi or 2:25/km) and a grid of wind speeds from 0 to 8 m/s (up to 18 mph or 29 km/h).
Our “reference class” is the average metabolic power[1] required to maintain the target speed across the test course. For example, if the target speed is 5.0 m/s (3:20 for 1 km), we find the metabolic power required to maintain 5.0 m/s on the way out (into the wind), and the metabolic power required to maintain 5.0 m/s on the way back. Then we average them (which is fine to do, since you spend the same amount of time going out and coming back—constant speed!)
The more interesting case is the constant-effort case, which is our comparator. We want to know the following: what constant metabolic power output results in our runner achieving the same time for the course (e.g. 3:20 for 1 km), given that they’ll slow down in the headwind and speed up in the tailwind?
The way to solve this problem is, essentially, to do a search across different possible metabolic power outputs, finding which one results in the necessary target time.
Then, the comparison becomes: is the constant-effort metabolic power lower or higher than the average metabolic power in the constant-speed case?
Results: constant-effort is usually better in the wind
Because of the airflow-velocity-squared term in the drag equation, I was expecting constant-effort to win out every time. But this is not what happened!
It does turn out that constant-effort is usually better: Below about 6:20/mi, constant-effort running is always superior, no matter what the wind is doing. And for moderate winds, below about 5 m/s (11 mph or 18 km/h), constant-effort running is better at any speed.
But the interesting and counterintuitive result was for slow running speeds in very strong winds: in these cases, it is substantially better to use the constant-speed strategy! Check out this plot of absolute metabolic power savings:
A more intuitive way to look at these results might be to look at relative change in metabolic power output, as a percentage—that’s exactly the units that research on super shoes uses (e.g. the famous “4%” number from the original Vaporfly 4% shoe—that number was literally the relative decrease in metabolic power output, in W/kg, at a given speed).
Here’s what that looks like:
So, if you figure a typical super shoe is more like 3% (the original 4% number was a bit higher than later studies), the benefits of constant-effort running at high speeds in intense winds are almost 2/3rds the advantage of a super shoe!
Why slower runners benefit from constant-speed running in strong winds wind
Now, how to explain the surprising findings for high winds and slow speeds? The explanation that I find most intuitive is actually not the “spending less time in the wind” argument—it’s the fact that, at slow speeds, small changes in your running speed result in a big change in your finish time over a set distance.
Here’s a concrete example, using some unrealistically large and whole numbers to illustrate the effect:
Suppose you are targeting a speed of 3 meters per second. That’s a 5k time of 27:45, or an average pace of about 8:55/mi or 5:35/km (all rounded a little bit). You face a stiff wind, and you slow down to 2 meters per second. That costs you 83 extra seconds over the 500 meters on the way out.
How much faster would you need to run on the way back, with the wind, to make up for those 83 lost seconds? Six meters per second, double the speed you were attempting to run! Even a very strong tailwind can’t help you out that much.
In contrast, an elite runner targeting 7 m/s (3:50/mi or 2:25/km) who slows to 6 m/s in a strong wind ends up losing only 12 seconds over 500 meters. Making it up on the way back requires increasing to 8.4 meters per second – only a delta of +1.4 m/s instead of the +3 m/s in the slow runner’s case.
These examples are extreme to make the math easy, but even in more modest cases the math works out significantly favoring constant-speed running for high winds when we’re talking about speeds slower than around 8:00/mi or 5:00/mi (just from eyeballing the plots above).
Recap
If you are targeting a specific time on race day, and you’re running on a windy loop course (starting and finishing in the same place), you should run at a constant effort, no matter the wind, if you are targeting a pace faster than about 8:00/mi or 5:00/km.
If it’s very windy (>11 mph or >18 km/h winds) and you’re targeting a slower pace, it may make sense to use your watch and stick to your goal pace, even though you’ll need to work harder going into the wind.
Of course, none of this applies to workouts with a physiological goal—there, you always want to be using a constant effort, whether you’re fast or slow, because achieving a constant, targeted metabolic power output is the entire point!
And this advice also does not apply to point-to-point races where you are running with (or into) the wind the whole way. In those cases, a constant effort is going to result in a constant pace (though a faster or slower pace than on a calm day), and the only question to answer is what constant effort level you should target. For that question, you can use my headwind and tailwind calculator.
Though these findings mostly matched my intuitions I'm glad to actually work out the calculations - both to have some real numbers, and because I learned something new about the effects of wind at slower running speeds!
Learn more about the science of running
If you happen to be in Minnesota, join me on April 30th in Minneapolis for my first live event: a lecture and Q&A on the science of marathon training!
If you or any of your friends are local, you can get tickets here 🎟️.
If you can't make my live event, make sure to sign up for my email list to find out when I have a new article or new web app coming out.
If you're a marathoner (or an aspiring one), you should also check out my new book, Marathon Excellence for Everyone, which details a scientifically based approach to modern marathon training.
If you live outside of the United States, Marathon Excellence is also available in a special Metric Edition with all workouts and paces in kilometers.
If you aren't up for the marathon just yet, I published a shorter book in 2013 on a simple scientifically based approach to 800m to 10k training, Modern Training and Physiology.
Footnotes
[1] We measure metabolic power in watts per kg of body weight. You might be familiar with this power metric from cycling. If you want a more intelligible unit, you could easily convert it to “food calories burned per minute, per pound (or kg) of body weight” – watts is just a way of measuring energy expenditure per unit time, “calories per minute” is a different measure of the same physical property.
Related articles
Elite marathon training with Victor Urquhart Smångs
Talking marathon training on the Life in Stride Podcast
New web app: Predicting LT1 pace and Zone 2 pace from 5k time
New horizons in iliotibial band biomechanics in runners
Designing a plyometrics program for improving bone strength in young runners
Three theories of tissue damage accumulation during running
Podcast: Marathon Excellence on Running Book Reviews with Alan and Liz
There's also a crosswind, which is a hindrance. I hate running on the track because the wind changes four times per lap. It's much easier on the straightaway street.
If you're running for a long time against a strong wind, it's easy to move into a heavier intensity zone (for example, from marathon pace to 5K pace effort), from which you won't be able to recover without a significant loss of pace, and even the wind at your back won't help you.