Understanding the Bike Demands of Ironman - Part 1: Kona Pro Men

If you ever use Quadrant Analysis in Training Peaks, (QA), then you likely see how the specifics of a race can really be seen, and the neurological demands on the athlete. I'll avoid explaining the basics of the QA, and instead refer you to here, and dive right into some observations. My hope is to take this one course at a time, and really begin to identify the difference in demands from race to race, course to course, so we as coaches, (and athletes), can train more specifically for the race, as well as learn what might be the best way to ride a course.

With this in mind, let's start with the race everyone is dying to do and know more about, Kona. Obviously this race has the best field of long course triathletes in the world, but there is a difference in the type of races being executed, from the pro's, to the age group competitor, to the lottery winner who is just looking to finish. For this first post on this, I will be focusing on the pro men, courtesy of TrainingPeaks' files of Pete Jacobs, Luke McKenzie and Michael Lovato from 2012.

Ironman World Championships - Kona 

(click on all images to enlarge them)

Most people probably don't know this, but there is about 3000 feet of climbing in Kona, (depending on what source you use to measure), due to the large rollers, and the climb up to Hawi. Add in the quality of the field, and you begin to see the athletes really pushing themselves hard. Let's look at the course profile, because that is a very important aspect of determining training specificity.

So the course is mostly an out and back, with a big climb at the half-way point, called Hawi, (pronounced ha-vee). You can see that there are some other decent uphill sections, with some big spikes within those as well. So there is a constant up and down, never really a flat portion, but the up and down is rarely very steep, as most athletes are able to stay in the saddle for the majority of the race, save for Hawi. 

This is the QA of 2012 World Champion, Pete Jacobs. Pete has an FTP of 370 watts. You'll notice he spent almost 90% in the lower force quadrants. Could there be something to this?

Pete's percentage of samples in Quadrants - 1: 4%, 2: 8.7%, 3: 27.1%, 4: 60.2%

Pete's run split: 2:48:05

Next we will look at Luke McKenzie, who finished 24th in the same 2012 race, and has an FTP of 360 watts, riding only about 3 mins slower than Pete.

Luke's percentage of samples in Quadrants - 1: 0.1%, 2: 36.4%, 3: 58.5%, 4: 5%

Luke's run split: 3:20:32

If we compare with Pete, then quadrants 2 and 4 really standout, as Luke varied greatly from Pete in those. Consider Pete had about 12 times more samples in Q4, which was lower force, higher cadence, and less than a quarter of the amount of samples in Q2 than Luke, which are low cadence high force, (think mashing). Surprisingly, Q1 which means the biggest surges, Pete had quite a bit more in comparison than Luke, but seemed to be able to manage that with such a still small sample, and much less time overall in higher force quadrants. 

Next we will look at Michael Lovato, who finished 25th, right behind Luke. I have estimated Michael to have an FTP of 375, based on the TSS provided by TrainingPeaks, of 274. (Proof that the higher FTP isn't always what matters, but how well you can ride close to it). Michael ran much faster than Luke, but didn't seem to ride near as well, from a time perspective. 

Lovato's percentages of samples in Qaudrants - 1: 1.4%, 2: 13.3%, 3: 49.2%, 4: 36%

Lovato's run split: 3:03:13

Comparing Lovato and Pete, Lovato's Q1 and Q2 seem to be in line with Pete, 14% of samples in those two for Lovato, 13% for Pete. The biggest difference seems to be the Q4 time, as both Luke and Lovato spent the majority of the time in Q3, lower cadence, lower force. For Pete, he spent most of the time in Q4, and about half the time in Q3 that the others did. 

The small yellow triangle represents the average of all the samples, and for Lovato and Luke, they fall in the Q3. For Pete, it falls in Q4. 

So does this mean athletes should be spending more time training with higher cadence, lower force? Not necessarily. It could be that Pete was simply better fit, able to push the gears faster than the other two, and able to run well anyway. Certainly, Pete has been a heck of a runner, having the fastest run split at 2011 Kona. 

I think if there is a conclusion to draw from this, less Q1 and Q2 time is probably best. I do think there is something to be said for having the neurological fitness to be able to hold Q4 for that much of the race. Just for fun I highlighted the samples from the top of Hawi and back for Pete to see if there was a change.

Pete's samples from Hawi and back in Quadrants - 1: 0.9%, 2: 7.9%, 3: 37.3%, 4: 53.9%

Of course, we can expect that he will fatigue, but there really was little change in the percentages for the Q1 and Q2. Q3 saw much more, as it seemed he just couldn't likely hold Q4 as much as he would have liked. (He's racing hard after all!) His average for those samples falls in Q3 now. 

Comparing his cadence from the first hour and the last hour, it dropped by 6.82%. Lovato's cadence fade was actually less, but his power fade for the first and last hour was 10% more than Pete's. Luke had the largest cadence fade, and the largest power fade. See this table which summarizes:

So what does all this mean? It's easy to say athletes should do what Pete Jacob's did, but it's not that simple. This is a small sample size of 3 athletes, and there are a number of factors which could play a role, and athletes are not all the same, far from it. But I do believe this is a small bit of evidence on what it takes to perform well at Kona. It would seem that Pete was trained well to hold a high cadence, and had the neurological fitness to do so. His cadence fade affected his power less since he was still holding a relatively high cadence to start with, possibly saving the higher force outputs for the run. 

In the next part, I will look at some top age groupers from Kona, and see what their QA's look like, if there is any big differences or not. What will we see? Truth be told I'm not sure, haven't gotten that far yet. I would love to do a similar post on the Pro Women, but the data is really limited. Let me know if you find some good files to compare. 

Coach Vance