After an hour playing with the results from Challenge Roth 2012 I came to the startling conclusion that the fastest Ironman athletes are fast on the bike, fast on the run and, probably, fast in the swim. You may feel I should have better things to do with my evenings than sit at my computer, Excel open, charting datasets and looking at correlations. Unfortunately, you are wrong. Not that I’d set out to demonstrate the seemingly obvious, I was actually considering the relationship between bike and run performance for a blog on the impact of cycling too hard, but as I compared more of the data I became distracted.
What was more interesting was – as the charts above demonstrate – the high correlation between either bike performance or run performance and overall time (the charts include all athletes who finished). Hence the obvious conclusion that faster athletes are, basically, faster. If you look at the sharp end of the field you see a tight focus of bike and run times, but the pattern is not as strong for the swim. I hadn’t abandoned my bike-run comparison, plotting that chart demonstrated the weakest correlation of all; as we might expect, a fast bike won’t necessarily mean a fast run, but generally those who bike faster also run faster.
Clearly the fitter, better prepared athletes are going faster across the board. If you want to win your age group you need to be strong in all three sports. It’s also clear that swimming is the discipline where athletes vary the most in ability, you get slower swimmers with fast overall times and some fast swimmers with slower overall times. I know many age groupers will be relieved to learn that their swim is a poor indicator of potential race performance, but before they get too excited, I have more to add in a moment. Otherwise the faster you can bike and the faster you can run, then the faster you will go. A good reason to train hard.
The relationship between bike and run is relatively weak, showing a much greater dispersal of points on the chart; from the Pros, fast in both, through to the back of the pack with a broad spread of abilities. Looking at the middle of the data set there is a range of marathon times for a given bike split, chances are many of the outliers, those on the upper fringe, did bike too hard and pay for it on the run. Of course some may have mitigating circumstances – an injury or illness, but many simply made pacing errors as I’ve indicated in the graph below.
There’s no real answer as to how much impact riding too hard has on the run (my original motive for producing these charts), experience suggests it does and the dataset hints at the same. We can tell that being fast on a bike doesn’t necessarily mean being a fast runner; though we might expect that a strong cyclist has the fundamental fitness to run well, whether that actually translates through technique and efficiency to a fast marathon split is another matter. The Roth data brings me no closer to demonstrating how bike impacts run. It’s a good job I found something else to do with it.
The Sub-10 Athlete
When I tweeted these graphs yesterday evening, the real interest came from those at the sharp end, looking to refine their race further or to break the 10 hour barrier. So to better appreciate the race at the front I produced a second version, this time focussing only on those who broke 10 hours.
With this smaller data set the correlations are weaker, but what was more interesting was how the typical sub-10 age grouper puts their race together. Bearing in mind this is course and condition dependent, in Roth this year they would have been swimming between 55 and 65 minutes, cycling between 4:50 and 5:20 and then running between 3:10 and 3:40; allowing for transitions the slow end of those ranges give a fairly balanced race and some indication of the standards to aim for. And back to that swim, it may not be strongly correlated with finish time, but you really want to be able to swim a 65, or at the the outside a 70, to have a comfortable shot at sub-10.
Correlation is not causation. I’m not suggesting that hammering the bike always leads to a faster overall time, it could be quite the opposite. Fitter athletes tend to go faster, so we see this relationship between performance in all three sports and the finish time; however the relationship between bike and run is much weaker. Fitter athletes, those going under 10 hours, are pacing correctly. They are fast on both bike and run because they have trained to the point where they are fast at a sustainable effort for a 10 hour race.
Looking at race statistics and times is an interesting diversion. It gives a sense of the standards required, but what should be taken away is the need to train to reach those standards. What really matters is the balance of effort across the three stages, courses and conditions change times, it’s all about managing yourself.