We’ll start this blog with a caveat, this forecast is not an accurate prediction but a rough guide of the speed the riders might get come race day out of their 24 hour TT effort. The gpx file used to make the forecast essentially covers one lap of each leg, which come race day the riders will cover multiple times. The file that we’ve used for this is a total distance of 197 km and there’s a couple of things that will increase the uncertainty of these forecasts.
Caveat number 1 – The Road Surface
myWindsock allows users to set the rolling resistance during their forecast, which on a TT where you know exactly the route, you can make a reasonably educated guess at. However, during a 24 hour TT the riders will be doing multiple laps of the various legs so it’s quite difficult to know whether they’ll end up doing 3 laps of the leg with the dodge road surface and 5 laps of the leg with the amazing (for Britain) road surface or vice versa.
Caveat number 2 – the wind
It won’t surprise you to know that the wind, during a 24 hour time trial, can change quite dramatically. In the UK, average wind speed over the day and prevailing patterns follow some consistent trends due to the country’s maritime climate and westerly wind influence. Usefully, the National 24 hour championships are in July. July near Wrexham brings relatively calm conditions, with average winds between 7–13 mph (≈11–21 km/h). The most reliable data suggests a typical value around 9 mph, making it the least windy month of the year.
Even with relatively low winds, there will be legs that have faster wind conditions than others and we don’t know the proportion of time riders will spend on each loop and such. That said, using all loops equally with average historical wind data for the date of the event may actually average out all of these problems and we could get a reasonable estimate.
How to win
Now the caveats are out the way, let’s turn to how we might go about winning this race.
In 2024, the National 24 hour TT championships were won with 426.19 miles for the women and 546.36 miles for the men. In 2023 the men’s top three covered between 501–519 miles with the women’s podium riding between 381 and 465 miles. Hard to know exactly how far you’ll have to ride in order to podium this year but the men will most likely need to exceed 500 miles and the women will need to exceed 400.
This is the average speed required for a given distance over 24 hours – bare in mind that this also includes fuelling and toilet stops.
Now we have the speed, let’s check out the power needed for each of these speeds. For the sake of this forecast, I’ll use a 72kg rider with a 10kg bike, no pacing rules but sitting up on the climbs. If you want to plug your own numbers in, you can sign up to myWindsock here.
Using a myWindsock forecast of the course we can see how much power we need to ride at various speeds. Please bare in mind the values inputted to the forecast will not be your own but you can sign up to myWindsock here and have a go yourself.
While we aren’t quite crazy enough to enter the 24 hour national time trial championships, myWindsock has decided to sponsor them and as such, we thought it would be a good idea to take a look into what it is, how you might prepare for a 24 hour time trial and go through some practical tips. It’s weirdly hard to get the “mainstream cycling media” if there is such a thing to cover these events but personally, I find the idea that anyone would even want to enter into a competition where you have to ride for 24 hours continuously interesting.
24 hours like this? Go on then…
What is a 24 hour time trial?
On the face of it, the rules are quite simple…
Riders start individually at one-minute intervals.
Each rider attempts to ride as far as possible within 24 hours, starting from their individual start time.
The winner is the rider who covers the greatest distance.
There are a couple of circuits used – one for daytime, a safer one for night time and a simple finishing circuit to help timekeepers count laps. Each rider’s distance is measured by counting completed laps and using the rider’s last known position when 24 hours elapse.
Other than that, they start in the morning and ride until the following morning.
How can you even prepare for such an event?
A 24 hour time trial is an extremely complex problem. Your main physiological limitation is energy availability – you’ll be burning more energy than you can take in and what you take in will be limited by what your gastro-intestinal system can tolerate (and one might imagine that 24 hours of gels would cause a bit of stress in that department but we’ll get onto what to eat later). Completing back to back long rides (i.e. 8 hours Saturday then another 8 on Sunday), GI training on long rides (12+ hour rides with race nutrition) and strength training to improve muscular endurance and glucose uptake will all help this.
There are other things to consider too, with the race being so long there are psychological and neurological factors at play – anxiety over the distance, the very real effects of sleep deprivation. A study by Jugovac and Cavallero in 2012 showed that 24 hours of total sleep deprivation significantly impaired executive control in attention tasks, while leaving other attentional networks largely unaffected. This means a cyclist’s ability to make decisions, plan, and manage effort (executive control) will be significantly impaired, even if basic alertness or reaction to stimuli (like steering or responding to road conditions) remains relatively intact. A robust plan is key here in order to limit the necessity for executive control, your decision making will be massively compromised so simply limit the number of decisions you have to take in the moment and take them in the run up to the event. These are things like when and what to eat, how much power to ride at and making sure you have the route in your head unit or very squarely in your memory.
How should I pace it?
You are unlikely to negatively split a 24 hour time trial and typical pacing advice of going a little over threshold on the climbs and recovering on the downhills doesn’t really apply. The key to pacing a 24 hour TT is to start slow and slow down as little as possible. Conventional advice is to start at “normal endurance ride” pace – around 55-60% of your FTP. Most riders will go out way too hard.
Coming up with a distance schedule of reasonable low and high estimates (along with average power for these) is one way of ensuring that you don’t start too hard as well as alleviating the inevitable panic when people are ahead of you at 6 hours.
This leads us on to mental pacing – a 24 hour race is long but splitting your plan into 6 hour macro-chunks each of 2 hour micro-chunks will make it more manageable. It also allows you to know where you should be roughly during each block, what you should eat and when you should stop if needed.
What should I eat?
A good fueling plan for a 24-hour time trial is the difference between riding strong into the night… or falling apart at hour 10. The main difference between a 24 hour TT compared to any other, even a 12, is the fact that you have other considerations than just getting down as many carbohydrates as humanly possible. There are consequences to just eating gels for 24 hours in terms of muscle preservation, recovery, mental function, and gut comfort. A low level of protein intake every 6-8 hours (around 15g would probably do it – roughly half a scoop of protein shake) could reduce muscle breakdown and improve mental clarity. Fat isn’t needed in large amounts during the ride, but a total absence can cause GI discomfort (lack of fat can delay gastric emptying in some) and contribute to brain fog. You need to fuel your brain AND your legs – this is unique to a 24 hour TT.
Keep the plan simple – a peanut butter sandwich and a bit of protein every 6 hours, 60-90g of carbs (more if you can hack it) on the bike with electrolytes. Make sure you have food that you like – you’re going to consume a LOT. Many riders will easily surpass 10,000 kcal over the duration of the race – some will burn double that. That’s a huge amount of food. Prepare a large variety with known macronutrient contents.
So, what is “the truth” about ultra endurance time trials?
The reality is, ultra endurance time trials are not just a physical test. It’s a game of who can be the most prepared in all aspects, from the gastro-intestinal to the psychological. You will slow down, you may even hallucinate and your ability to operate on zero sleep and a large calorie deficit will be tested.
If none of that has put you off, and you’re a UK based reader of our blog, you can enter the National 24 hour TT championships here – entries are open for another 4 weeks at the time of writing. If you want to be as prepared as possible for your next race, whether or not it’s a 24 hour TT – sign up to myWindsock!
If you’ve ever looked at your ride on myWindsock, you know how important the weather is for cycling. Headwinds, tailwinds, air pressure, and even temperature can affect your speed and effort. But where does this weather data actually come from?
What Is a Weather Forecast?
A weather forecast is a prediction of what the weather will be like in the future — whether that’s in the next hour or several days ahead. Forecasts tell us things like:
How strong the wind will be
What direction it will blow
Whether it will rain
What the temperature and air pressure will be
These forecasts help cyclists plan everything from clothing choices to pacing strategies. But how do we know what the weather will do?
Forecasts are made using something called weather models. These are complex computer programs that simulate the atmosphere using current weather observations. There are many different weather models used around the world, and a rider in the UK might see forecasts based on models like the UK Met Office’s UKV model, the European Centre for Medium-Range Weather Forecasts (ECMWF) model, or the Global Forecast System (GFS) model. Each model has different strengths — some are better at short-term, very local predictions, while others excel at longer-range forecasting across wider areas.
Here’s a simplified version of how it works:
Observation: Weather stations, satellites, balloons, aircraft, and ships collect data from all over the world. This includes temperature, wind, humidity, and more.
Analysis: This data is fed into supercomputers that look at current weather patterns.
Modeling: These computers run simulations based on the laws of physics to predict how the atmosphere will behave in the future.
Forecast Output: The result is a weather forecast — including maps, numbers, and wind arrows — like the ones myWindsock uses to simulate your ride
myWindsock uses forecasts, not just current conditions. That means when you plan a ride for tomorrow or analyse one from last weekend, the weather details are based on predicted conditions that change across time and distance — exactly what you need to understand how the wind or rain helped or hindered your ride.
myWindsock uses forecasts, not just current conditions. That means when you plan a ride for tomorrow or analyse one from last weekend, the weather details are based on predicted conditions that change across time and distance — exactly what you need to understand how the wind or rain helped or hindered your ride.
What do we do with the weather forecasts and why isn’t it always perfect?
Forecasts often disagree, for example today where I am it is simultaneously 13 degrees with light showers as well as 8 degrees and snowing. This is a short term forecast for the next few hours, as you can imagine weather models disagreeing with each other is a cause of headaches for us.
As you’ll know, weather forecasts aren’t always perfect and these models have errors associated with them. These errors propagate through the myWindsock physics models (the things we use to work out how fast you’ll go over a given distance (so if the weather model says the wind will blow faster than it ends up blowing, even if our physics engine is perfect, the prediction will be wrong). Of course, here at myWindsock we spend time accounting for this and doing what we can but it’s a reality of trying to make an extremely complicated prediction.
How do we make predictions?
The best way to think about it is by considering the “equation of motion” of a cyclist. An “equation of motion” is something physicists use in order to model something. Now, I’ll let you into a secret – there’s always stuff that physicists just don’t count because the effects are small or including it just makes the mathematics a bit too hard. What you have to do is think of all the things speeding the cyclist up (pedal force, tailwind force etc) and all the things slowing the cyclist down (headwind force/air resistance, rolling resistance etc) and add these up. Once you do that, you’re left with a number which will either be positive or negative which tells us whether or not the bike is speeding up or slowing down.
This diagram, from a paper published in an academic journal, shows us roughly the forces acting on a bike (but it leaves out drivetrain resistance). The main point I’m trying to make here is that there’s quite a lot going on all at once. The air resistance force that we predict is a function of the weather forecast and the rolling resistance force that users enter using the myWindsock advanced settings is actually a function of the pavement so is rarely (almost never) a static value for the entire ride.
We have a physical model similar to the one above which does all of this – some of the inputs come from the user (like power numbers and the gradients from the GPX file you use as well as rolling resistance in advanced settings) and others from the weather forecast models. We then add all of these up, calculate the forces at each step along the course, use the result of this calculation to figure out the speed then add these speeds up to get a time prediction at the end of the course (if you’ve already don’t the course and upload a ride file, we do all of this backwards and figure out the resistive forces instead).
What can you do improve your predictions?
If you find that our predictions are off, there’s two steps you can take…
Make sure the forecast you use is as recent as possible – so update the forecast. It will change from the one you made 2 weeks out, as you’ll have experienced weather forecast accuracy decays with time from the date you’re trying to forecast.
Check your settings – bike weight, cda, rolling resistance and such, take some time to make sure these values are as close to accurate as possible. Methods for doing this could (and probably will) make an entire blog in their own right.
While weather forecasting and physics modeling will never be perfect, myWindsock brings you as close as possible to understanding how the conditions around you shape your ride. We are the best in the business. By blending cutting-edge atmospheric data with detailed physics simulations, we offer a powerful tool to plan, analyse, and learn from your efforts on the bike. And with a little help from you—up-to-date forecasts and dialled-in ride settings—we can make those predictions even better. After all, cycling may be part science and part suffering, but we’ll do everything we can to take the guesswork out of the wind.
European race season is upon us, the National B road races, CTT time trials, IRONMAN Triathlons and World Tour stage races have hit Europe and according to our data – you’re all using myWindsock to prepare. As you’re doing it anyway, we thought we’d go through how we, here ay myWindsock, use the app to get ready for events that we’re doing. This will be quite comprehensive and will go over the following:
Course recon
Duration predictions
Pacing plans
Fuelling
Running the risk of being self indulgent, I’ll base the contents of this piece on a course that I’m doing – namely the bike course at the Peniscola Infinitri Triathlon – my first race of the season.
Course recon
Unfortunately, we don’t all have unlimited money and time so can’t often recon a course in the flesh – but myWindsock is the next best thing for it. It allows us to review the terrain, weather and their combination.
The plot above shows the gradient as a function of time, which is typically where I like to start. This allows me to think about gearing on a high level and decide between a 1x and 2x chainring set up. For this race, it’s on the cusp but I prefer to pay the aero penalty rather than running the serious risk of being undergeared so I will use a 2x chainring.
The second plot I take a look at is then “Where Power Matters Most” – this essentially shows us where on the course larger investments of energy are best rewarded, typically the steeper gradients. The colour on the plot represents the wind direction where the outline of the elevation sits. This allows us to see where attacks might come and what parts of the race are going to be the hardest – in this example we can see that the bike leg is essentially only 55km long and that from this point it’s just about recovering for the run. This leads us to start thinking about how long the race will take and what pacing strategy we should make use of…
Predicting duration
In order to predict duration I do a worst case scenario and best case scenario which to me, looks roughly as follows…
From this we can see I’ll be around the two hour mark plus or minus a couple of minutes depending on how I feel. You’ll notice a wide cda range as well – this is because on a bad day I also typically sit up a bit more and shift around on the saddle. With these constraints we can come up with a pacing plan and look into the course demands.
Pacing plans
Firstly we can look into the distribution of speeds and we can see it’s a very wide range – we spend a lot of time under 30 kph, so we will look to minimise this by going harder and setting a rule that if we are under 30 kph, we’ll do 400W, if under 20 we will sit up and do 450W. To balance this out, we’ll also stop pedalling over 55 kph. Let’s see how this impacts a middle case scenario (remembering that in triathlon we are looking to balance going fast with saving energy for the run).
The above case is without the aforementioned pacing rules, and with them we get…
We notice it’s actually 14 seconds slower (probably within the margin of error). Changing out the speed at which we stop pedalling from 55 to 65 kph adds 20W to the average power and saves 2 minutes over the course of the event. This information allows me to make better decisions during the race, if I want to behave aggressively I know what the time savings for going hard on the descents are and I also am fully aware of the energy penalty I’ll pay for doing this.
Fuelling planning
We need to plan to consume 2 hours worth of fuel for this race, which for me is around 240g of carbohydrates, 1000mg of sodium and around 1.5-2L of water. Now, we don’t want to be eating when going full gas and we can use dips in the power profile to think about when we’ll likely be able to get some food in. Around 15 minutes, 40 minutes, 50 minutes etc. I’ll have the chance to take a gel or a sip. Also, looking at this tells us the first 15 minutes is likely to be pretty full on so maybe a gel in transition would be wise.
This blog gave us just a brief insight into some of the tools available in myWindsock for planning performance but it was by no means exhaustive. To prepare for your next race and leave no stone unturned, sign up to myWindsock here.
This week, we speak to Simon Warren. The author of the iconic 100 Climbs books.
If you’re considering a Christmas present for the cyclist in your life, these books might just be the perfect stocking filler, take a look at the full range here.
This week we speak to Alex Welburn, the founder of the performance project – you can find more about this here https://www.myperformanceproject.com/about.
This week we speak to India Lee, listen to the full episode on Spotify.
India Lee is a perennially underrated athlete constantly appearing as if she’s the underdog despite a slew of fantastic results. We spoke to her in the wake of a disappointing race at the Ironman World Championships and an amazing race at the T100 World Tour race in Ibiza.
Grant Sheldon – listen to the full episode on Spotify.
This week we speak with Grant Sheldon, a former elite funded GB triathlete. We had a great time recording this episode and thanks to Grant for being so open and honest with us about what it’s like as an athlete to be supported (or not supported) by your federation.