Your watch has synced and the finish time is there, seven minutes off the goal. The day after a race that fell apart you want to know why, and a lot of the answer is in the file from your watch. Some of it is only in your head, and no app can read that part.
Don't start with the finish time. It tells you something went wrong but not what. Go through the race in the order you ran it.
Where the pace ran out
Put your kilometre splits next to goal pace and look for the place where the pace starts to slip and never comes back. A single slow kilometre on a hill doesn't count.
Where that point is says more than how much you lost. A marathon that holds to 32 km and then falls apart has a different explanation from one that is ten seconds per kilometre down from 15 km onwards. The first usually points to fuel or a start that was too fast, since a fast first half burns through carbohydrate earlier. The second points to the heat or a goal that was beyond your fitness.
In 2021 Barry Smyth went through more than four million marathon results. He counted a runner as having hit the wall when their pace dropped by at least 25 percent for more than five kilometres after 20 km. It happened to 28 percent of men and 17 percent of women, which is more than one man in four in the start corral.
Heart rate shows whether the start was too fast
The pace early in a race always feels easy, which is exactly where the feeling is least worth trusting. Heart rate says more.
Compare your heart rate over the first five kilometres with what you usually see at the same pace in training, in similar weather. Allow a few extra beats for nerves. If it was clearly higher than that already there, you were running beyond what you had that day from the gun, even if the pace matched the plan.
Then comes the drift. During a long race heart rate climbs at the same pace, because the heart pumps less blood per beat as you heat up and lose fluid. Edward Coyle and José González-Alonso described the mechanism in 2001, and drinking during the race dampens it. Some drift in a marathon is normal. If heart rate climbs a lot in the first half, your body was working too hard early, and you pay for it after 30 km.
The best comparison is your own previous race over the same distance, ideally on the same course. If heart rate rose more this year than last at the same kilometre and the same pace, something has changed, and that comparison says more than two finish times side by side.
The weather takes more than it feels like
In 2007 Matthew Ely and colleagues set the results from seven large American and Canadian marathons, some with more than 30 years of data, against the weather on race day. Times got slower the warmer it was, and the slowdown started on cool days. The fastest men were 1.7 percent off the course record on the coolest days and 4.5 percent off on the warmest. Further down the results the loss was larger.
Temperature alone is not enough to judge the day. Humid air makes it harder to get rid of heat, so look at the dew point too. A race at 18 degrees in humid air is a warm race for someone who did their last long runs at ten degrees.
Headwind and hills slow your splits even when your effort stays the same. Intervals.icu and many watches show a grade adjusted pace, and on a hilly course that is the one to hold against goal pace, not the raw one.
If the day was warm, the goal should have been adjusted on the start line. That is easy to say the day after. But it also means a slower time on a warm day doesn't have to mean worse fitness.
How rested you were
Look at the last three weeks before the race. Did the taper turn into an extra training week because you were worried about your fitness, or was the last long run too close to race day? What those weeks usually look like when they work is in the post on tapering, and it is easier to see what happened when you have something to compare with.
If your watch measures resting heart rate and HRV, read the morning values from the week before. A resting heart rate a few beats above normal for several days in a row proves nothing, but it is worth putting next to the rest, especially if you felt under the weather. How to read those values is in the post on recovery.
Fuel doesn't show in the file
For races between one and two and a half hours the recommendation is 30 to 60 grams of carbohydrate per hour, and up to 90 grams for longer ones. That is from the 2016 joint position statement of the American College of Sports Medicine and the North American dietetics associations. Two gels of around 25 grams each over a four hour marathon comes to about 12 grams per hour. The wall after 30 km often has that simple explanation.
Your stomach is the other thing the file can't see. Between 30 and 50 percent of endurance athletes get stomach problems when training or racing, according to a 2014 review. A toilet stop shows up as a slow kilometre and nothing more, and if it was the sports drink at the aid stations that upset your stomach, only you know.
Fuelling can be trained. Eat on your long runs what you plan to eat in the race, in the amount you plan to eat it.
Sometimes the race was good and the goal was wrong
Many runners set their marathon goal by converting a half marathon time, often with Peter Riegel's formula, which sits behind many race calculators online.
In 2016 Andrew Vickers and Emily Vertosick compared the formula with real results from 2,303 recreational runners. Up to the half marathon it held up well. For the marathon it gave times that were too fast, at least ten minutes too fast for half of the runners. A model that also took into account how much the runners ran each week did considerably better. If you worked out your marathon goal from your half and trained four days a week, it may have been the goal that was wrong, not the race.
So put the race next to your own earlier results. If you ran faster than last year on the same course, on a warmer day, it was a good race whatever the goal said.
What the data doesn't know
The crowd at the start, which made the first kilometre 40 seconds slow and had you chasing it back over the next three. The night before, when you slept four hours. The shoe that rubbed and the decision at 25 km to let the group you were in go. The file shows one slow kilometre and three fast ones after it, but not why.
Write down what you remember today, while you still remember the order things happened in.
What to take with you
Pick one or two things to do differently next time, not everything at once. A start that was too fast can be trained away on the first kilometres of your long runs, and an empty tank can be prevented by eating more on the same runs. If it was the heat that got you, the lesson is more about setting a goal that can take the weather next time than about training differently.
Give your body time as well. A marathon takes weeks to recover from, even when your legs feel fine after a few days. After a 10K a few easy days is enough.
The race review in Coach Fartlek
Coach Fartlek is our own app, so expect bias. When a race you have added syncs from your watch, the coach goes through it on its own and writes a review on the race's page in the app. The result is set against the goals you had for the race and against your personal best. Then the coach reads where the pace changed and how heart rate drifted, with the weather on the day taken into account, and compares with your earlier races over the same distance and earlier runs on the same course. The review also covers the build-up and the taper and ends with what the next build should do differently.
The review doesn't change your plan. Planning does that, and it scales recovery to what the race took: a marathon gets a recovery block of its own, a 10K a few easy days in the week after. Your threshold is recalculated from the race, and if it moves you get a separate message about your new zones.
What the coach can't know, like your stomach at 25 km or the queue for the toilet, you tell it in the chat, and it can then rewrite the review with that included. If you only entered your time by hand, the review is built from the result, your goals and the build-up, because there are no splits to read.
The app needs a watch connected through Intervals.icu and covers 5K to marathon. Pricing is on the pricing page, and you can try it free for a week.
Sources
- Smyth (2021): How recreational marathon runners hit the wall
- Coyle & González-Alonso (2001): Cardiovascular drift during prolonged exercise
- Ely et al. (2007): Impact of weather on marathon-running performance
- Thomas, Erdman & Burke (2016): Nutrition and athletic performance
- de Oliveira, Burini & Jeukendrup (2014): Gastrointestinal complaints during exercise
- Vickers & Vertosick (2016): An empirical study of race times in recreational endurance runners
Questions and answers
Why did I hit the wall in my marathon?
Usually it is fuel that ran out, and a start that was too fast makes it run out sooner. If you lose time early, it points more to the heat or a goal that was beyond your fitness. In an analysis of more than four million marathon results, 28 percent of men and 17 percent of women hit the wall.
How much does heat affect marathon time?
More than it feels like. In a study of seven large marathons, times got slower the warmer it was, starting on cool days. The fastest men were 1.7 percent off the course record on the coolest days and 4.5 percent off on the warmest, and further down the results the loss was larger. Check the dew point too, since humid air makes heat harder to get rid of.
How can I tell if I started too fast?
Compare your heart rate over the first five kilometres with what you usually see at the same pace in training, in similar weather, and allow a few extra beats for nerves. If it was clearly higher than that, you were running beyond what you had that day from the start, even if the pace matched the plan. If heart rate climbs a lot in the first half of a marathon, you usually pay for it after 30 km.
Could my marathon goal have been too ambitious?
Yes, especially if it was converted from a half marathon time. Riegel's formula, which many race calculators are built on, gave marathon times at least ten minutes too fast for half of the 2,303 recreational runners in one study. A model that also took weekly running volume into account did considerably better.