As frequently requested by tri fans following the broadcasts of the T100 races, the PTO have added more athlete data to their 2024 broadcasts, mainly using HR as an indicator for how hard an athlete is working at the moment. One of my interests is using the available data to enhance the broadcast and to explore what analysis would be helpful to tell the story of an unfolding race. As an example, this post looks at the women’s Dubai T100, the race that decided the first T100 World Championship title. It discusses what happened in the race and how that is reflected in the HR data.
Key Race Moments
Dubai T100 occurred last November, so let’s quickly refresh what happened in the race.
Here’s a Race Development Graph for the female race, showing who was leading at certain points in the race and how far back everyone was:
(As always, click on the graphs for a hi-res version that makes it easier to discern the details.)
The Race Development Graph highlights three athletes and key moments of the race:
- Taylor Knibb (blue line at the top, i.e. leading the race from the start of the bike all the way to the finish line) building a commanding lead in the last two laps of the bike.
- Ashleigh Gentle (green line) making up time to Taylor in the first laps of the run but then falling back into third place across the line.
- Flora Duffy (orange line) losing time on the bike but then posting the best run split and moving into fourth.
Below, I look at each of these athletes and how their heart rate data reflects what happened in the race.
Taylor Knibb
Let’s start with Taylor Knibb’s HR data:
I was especially interested if it was possible to see that Taylor Knibb picked up her effort towards the end of the bike. At the end of lap five, her father shouted at her, “You’ve got to pick it up now, let’s go!” and it was clearly audible on the broadcast. Here’s a screenshot of that moment:
Here are my observations on Taylor’s graph:
- For most of the bike, you can see Taylor’s rate slowly declining (the table below shows that it went down from about 89% of her max heart rate down to about 85%). She stopped that decline in the last two laps, and her heart rate stayed roughly the same for the sixth and seventh lap as it was for her fifth lap.
- Taylor significantly increased her lead to Ashleigh Gentle and Julie Derron in the last two laps of the bike. The table below the HR data shows how much she was faster in each lap. In T2, Taylor’s lead was 2:42 over Ashleigh; she created half of that lead in the last 25 minutes on the bike.
- On the run, Taylor could increase her heart rate, but her highest HR for the whole race was right after T1 when she worked hard to close the gap to the front after the swim and a somewhat slower T1 to put on socks.
Ashleigh Gentle
Let’s compare Taylor’s graph to the one from Ashleigh Gentle:
- Ashleigh had her highest heart rates during the “Aussie Exit” in the middle of the swim and running into T1. This could be a sign that she swam a bit harder (to stay with the good group she was in) and that she also made an effort to have a quick T1.
- Throughout the bike and in the first half of the run her heart rate is slowly declining. This is a bit surprising, especially given the heat in Dubai which usually leads to an increase in heart rate the longer the race goes on. Since Ash’s graph is quite typical of a few other athletes, it might indicate that the biggest heat stress was already during the swim in the warm water. Another possible explanation: A slowly declining heart rate is often a sign of fueling problems. As it’s unlikely that most athletes didn’t have a proper fueling strategy, it could mean that the high temperatures diminished the ability to absorb calories. However, this would require a more specific and individual analysis than what can be seen in a simple HR graph.
- The fact that Ashley isn’t able to lift her heart rate after the bike is probably an indication that her absolute HR numbers are quite similar between the bike and the run. (Taylor’s data indicated that her max HR is higher on the run than on the bike. While this is the case for a lot of athletes, all HR numbers in individual disciplines and absolute values are very individual.)
Ash was making good progress in chasing down Taylor’s lead, but her progress stalled after about three hours of racing. Ash had to slow down and was eventually overtaken by Julie Derron. Here’s a detailed look at the closing part of the race:
- As in the other graphs, there are no obvious signs of heat stress, even as Ash had to slow down (at about 3:07 into the race) and stop and walk (at about 3:24 race time).
- After about 3:07 Ash slowed down and she was no longer able to run faster than Taylor. (Their gap stayed about the same for a while.) You can indirectly see this in her heart rate: It was slightly lower than before.
- When Ash had to walk, her HR came down considerably. She was able to rally herself a bit when Julie caught up with her, but once Julie passed and she wasn’t able to stay with her, her pace and heart rate came down again.
Flora Duffy
Flora Duffy displayed another interesting HR graph:
- As with all other athletes, her heart rate spiked at the Aussie Exit and around T1, showing how stressful these race points are for the athletes even if they seem to be moving relatively slowly.
- Similar to the other athletes, her heart rate slowly came down on the bike. She was able to lift her heart rate in laps five and six when she tried to stay with Kat Matthews and Laura Philipp. Apparently, that felt a bit too hard for her, and her heart rate came down again when she decided to let them ride away.
- On the run, she was again able to lift her heart rate, even to the point of having her highest HR towards the end, an indication that she fueled and paced well in Dubai.
Closing Comments
I have done my best to be respectful of the athletes and their actual values. Most of the “interesting” aspects can be derived from the “shape” of the graph – that is why the graphs above don’t show a proper axis with absolute numbers. In the broadcast, the PTO also doesn’t show absolute numbers but rather a percentage of an athlete’s maximum numbers. (This also makes it easier to compare values between athletes but relies on athletes submitting the max HR numbers which are sometimes off and need to be “calibrated” on the fly to be useful.) For the 2025 broadcasts, the PTO are exploring the use of power numbers (maybe as a percentage of FTP/threshold or as W per kg), but these numbers might be even harder to reliably collect/calibrate and more “revealing” about the strengths and weaknesses of an athlete. There will be tension between the interest of the fans to have compelling data points and some athletes’ understandable reluctance to share all their numbers with everyone.
But even the analysis of heart rate data (as shown above) is probably not easy to automate to be usable for the broadcast. Creating graphs takes a while, properly analyzing them a bit longer, and explaining what one might be able to see takes even longer – not easy to do that while you see the leader being chased down and trying to figure out how much longer it’ll take before a lead change happens. But maybe an analysis after the race – such as the one I have tried to do here – could be interesting for the hardcore triathlon data nerds?
Course Conditions