Iroman Hawaii - Power data analysis

Data from the Ironman Hawaii 2008

After having ridden the fastest cycling section last year, the Dane, Torbjørn Sindballe got off his machine in first position again this year. He started more aggressively as can be seen on the power readout (see graph below): the performance drops off towards the end and is nowhere near as consistent as last time. "I felt as though my form would allow it," explained the 2007 bronze medal man. The Viking had a bad time during the run, however and says this was due to a stomach problem. The heat was also a problem. Despite a clever cooling programme " ice cubes in his gloves, long, net vest top", Torbjørn had a 3° higher chest temperature than the year before. He fought through to the finish though and ended up dead beat in 45th place.

In fact a similar thing happened to all the athletes, who were strong on the cycling section. Faris Al-Sultan and Normann Stadler came in in 11th and 12th place, after they both had problems in the run. The only fast cyclist to reach the top ten was Eneko Llanos, who was second at the finish behind Craig Alexander. Craig Alexander purposely held himself back on the cycling section. "I was in exactly the right bunch, as I had planned," reported the Australian after the race.

It is risky to start the cycling section at too high a tempo. The reason is the disproportionate increase in aerodynamic drag on the bicycle. The energy required to counter an increase in drag increases with the square if the speed. So minutes "won" on the bicycle cost more energy than those gained at running speed, where aerodynamic drag can be almost discounted. (see also the data further down.) br>
So conceding 11 minutes to the fastest bicycle split time saved Craig Alexander 10% of his power. 10% less power is a world of difference for a top sportsman. In this way Craig Alexander probably rode evenly in the fat burning range and save his glycogen reserves for the run.

Torbjørn Sindballe, 45. Platz Kona 2008

Power readout of: Torbjørn Sindballe. In the first hour the average output was a massive 330W, in the last hour 280W. The overall average was a power output of 300.5W.

TIPP: We advise you as always to start out at an even tempo on the bicycle no matter in what category you race. To overdo it by a couple of percent costs dearly later in the event. It is better to invest the energy saved in the run, where usually the outcome of a triathlon is decided. The best chance of gaining a good overall time is when your running potential can be fully realized.

For professionals, who are aiming for a top placing, the case is a bit more complicated. A strong cyclist such as Torbjørn Sindballe must really try to get a lead on the cycling section, because that is his only chance against the top runners. So he has to take a calculated risk, which last year paid off. The "Cycling victories" of Normann Stadler and Faris Al-Sultan also demonstrate that an aggressive cycling strategy can bring results. In those cases, though, it was the far and away strongest athlete who won the triathlon in any case. The opposite argument is this; the more compact the talent in an Iron Man field, the less the chance is of winning by attacking on the bicycle.

Cycling vs. Running Performance Kona 2007

This was the fourth consecutive year that we are fortunately able to gather race data from the top contenders in Kona. This, together with what we have learned over the years from other events and what we learn daily from your feedback data, allows us to improve our methods. Recently we have studied if it's better to hold back a bit on the bicycle or go flat out from the start read on.

A possible answer is to be found in the energy equation. We have computed the numbers based upon the power data and race time of Torbjørn.
The theoretical question: What is the ideal ratio between cycling and running speed to minimise energy over the whole distance? The following graph showing Torbjørn Sindballe's actual race data, provides the answer: The graph shows the correlation between cycling and running energy for a given total (cycling + running) time as a function of running speed. Sound complicated? The simple message is this: The faster you run, the less energy you need overall.

The reason is that running energy increases less with speed than cycling energy due to the lower drag at low speeds. Going very fast on the bicycle wastes total energy unless the speed results from superior aerodynamics.

What we learn:
=> running speed is the number one limitation for overall performance
=> saving energy on the bicycle is a good strategy
=> low aerodynamic drag is crucial!
=> refuelling (easiest on the bike) is very important

So, what is my ideal cycling speed? The speed which allows me to fulfil my maximum running potential. Here you see Torbjørn's running output compared to his cycling output (vertical mark). His running output was higher, despite cycling being his strongest segment. Here is the equivalent graph, with data from age grouper Andreas Bodenmann (50). Adjusted for body weight difference (69 vs. 80 Kg from Torbjørn) his energy expenditure was exactly the same as Torbjørn's. Admittedly, Andreas needed more time to use the energy, as of course he didn't produce the same performance or speed as the pro. Andreas comparison of cycling vs. running output tends even more towards running output compared to Torbjørn (vertical mark). His total time was 10:40:42 with a 5:46:21 cycle segment and an excellent 3:16:42 marathon; the swim was his weakest discipline due to an injured shoulder).

Torbjorn Sindballe, 3rd place overall Kona 2007 (8:21:30)

Under hot, but almost windless conditions and without being challenged by the other two strong riders Normann Stadler and Faris Al-Sultan suffering from digestion problems, Danish Torbjørn "Thunderball" Sindballe rode the fastest cycling segment of the day. This segment started very fast. Torbjørn took his time and began to reel in the unleashed US-American Chris Lieto on the climb to Hawii and caught him an hour later. By then he had already put 11 minutes on the chase group from which he rode away earlier. His data reveals how he took and held the lead: Torbjørn did not slacken during the ride. He produced the highest and most consistent average power output so far measured and published at the IM on the Big Island: An average of 309 watts over the whole cycling leg. Impressive! (Average values for cadence: 82, heart rate: 139 both with a slight falling trend over the distance). Until the turning point in Hawii, Torbjørn averaged a power output of 313 W and was able to hold an excellent average of 305 W on the way back. In this way his power profile differs radically from most other pros whose data 2PEAK has been able to collect and analyze so far; they usually have a sharp fall off in output beginning after the turning point.
With his closing marathon in 2:57:25 Torbjørn proved that he did not even ride at his limit and was able to secure himself a well deserved podium spot, so far his best result in Kona.
It is also interesting to note that Torbjørn needed less power output in 2005 to establish a new cycling course record (286,6 W),(later broken by Normann Stadler in 2006). In 2005 his power output was also less consistent. The numbers prove a dramatic improvement in his riding strength.

Rutger Beke, rank 898 in Kona 2007 (11:13:58)

Rutger Beke wanted to race for the crown and after several top 5 placings his goal was realistic. Unfortunately he got hit badly in the head during the swim which affected his day. On the bicycle he put in an effort to finish just a minute behind a group with Chris McCormack despite having lost time during the swim. Stunning to see that he managed to do the cycling leg in 4:36 at a surprisingly low average power output of 232 W. He must have exceptionally low aerodynamic drag.Rutgers power profile also shows the typical dropping off trend. During the run his legs failed him, but he fought his way to the finish nevertheless!

Wolfgang Teuchner, rank 211 in Kona 2007 (9:49:22)

Wolfgang is Faris Al-Sultan's training partner. The speciality of the 81 kg (179 lb) heavy giant is cycling. Apparently he wastes his energy like nobody, as you can see in his power graph: until the turning point Wolfgang was able to keep the pace of the best pros and averaged 336 watts. After the downhill from Hawi he lost his pedalling action as his average output falls to 212 watts (average overall: 272 W) so his resulting cycling segment time of 5:01 is disappointing for a cyclist of his caliber. With a steadier power output he could probably have finished faster.

Faris Al-Sultan, 3rd place overall, Kona 2006

Faris averages the same power he achieved when he won in 2005: 283 Watts. His cadence was a bit lower this year averaging 81,7 RPM. The race again falls into two sections: High power on the way to Hawi, averaging 301 W in the first half and a drop to 267 W in the 2nd half. It can be seen in Faris' data that 7 meters distance between the athletes isn't enough to avoid drafting effects. Luke Bell, who was riding in a group in Faris' wake, especially in the 2nd half of the race, needed remarkably less power (15 - 30 Watts less)to hold the same speed. This also shows in the energy equation: Faris burned 4592kJ (mechanically) vs. Luke who spent 164 kJ less despite being taller and (2kg) heavier. Our spotter who followed the leading men, confirmed what we measured. The real cycling shootout was between Normann at the front and Faris behind him.

Luke Bell, 7th place overall, Kona 2006

Luke had a good start and went with Faris right from the beginning. His power profile is very similar to Faris'. In the 2nd half drafting effects can be seen. This results in a slightly lower overall average of 273 Watt at 82,9 RPM.

Mike Montgomery, Rank 281 overall, Kona 2006

Mike Montgomery (Age Grouper) delivers a solid and remarkably stable performance. In contrast to the Pros, who did a tactical race, he starts the cycling rather slowly and paces himself very well so that he doesn't slow down in the 2nd half of the race. His power output is almost maintained after the turn in Hawi. The resulting cycling split is very good: 4:53:45 Rank 124.

Faris Al-Sultan, Ironman World Champion 2005

Data from the champion: the graph shows power (green), speed (pink) and cadence (blue) from the bavarian pro Faris Al Sultan. The power output over the cycling section averaged 283 watts, the average speed: 40.45 km/h, the average cadence was 85 rpm.


Faris, who had the third best swim, got himself ahead and led the field after just 14 minutes and produced an average of 324 watts until the first split time. Thus he started much more conservately then last year when he averaged 343 watts for the same split. His bike race can be divided into two segments as regards power output and these altered at the turnaround point in Hawi. During the first 90 km his average power kept close to 309 watts, in the second segment however he had a few weak moments but the rest of the field must have had similar problems, as he was still able to defend his advantage. Not until the end end of the cycling section was he passed by Torbjörn Sindballe whose data can also be seen below.

Torbjorn Sindballe, sets new bike leg record in Kona

New record: 4:21:57 hours with hot but pretty calm wind conditions.
Torbjorn Sindballe breaks the record of 4:24:50 which Thomas Hellriegel established back in 1996. The graph shows how well Torbjorn paced himself. His average power output of 300 watts in the first half of the cycling section is only slightly above that in the second half when the Danish express produced an average of 273 watts. With his consistent and strong effort, he managed to pass Faris with 10k to go after having spent most of the race just 3:00 behind. It is clear that he has an excellent aerodynamic position. Despite his 10kg weight disadvantage to Faris to haul up Hawi, he needed on average only 3 more Watts then the champion for the same speed. We can assume the difference is the effort to catch and pass Faris. (286 watts).

2004 World Championship data

Data from Faris Al Sultan, 3rd place overall

The graph shows the power (green), speed (pink) and rpm (blue) of German Pro Faris al Sultan. Faris averaged 286,9 W and 37,0 km/h on the windy and difficult course. The interval marks (vertical lines) show average data for the first and second half of the race. Average power dropped from 309 to 268,5 in the second half. Faris started out very fast, hammering way above 300 Watt for the first 1,5 hours. After this brave phase, although at a slightly lower level, he stays pretty constant. Variations in the speed graph are indicative of the climbs and descents. Speed is generally higher in the second half. Look how Faris keeps his power output up in the descents by using a huge gear. Apart from the descents, where he reaches 100 rpm, he is pedalling at a steady 80 rpm.





Smoothened signal: The average power output is clearer in this graph. The sharp drop in the last 20 minutes is mainly due to the smoothing.



Data Distribution: This graph shows the time spent by Faris at which power/speed/rpm. Most frequent speed (highest peak in the pink curve) is 30 km/h - the total range is between 20 and 60. His power band is wide - from 180 to 400 W. Most frequent power is around 300 W.

Data from Fernanda Keller, 9th place

Fernanda's speed signal is similar to Faris' but at a lower level due to the lower power. Fernanda averages 161 Watts and 30,6 km/h. Taking her light weight into account she has a power/weight ratio of 3,22 W/kg - approx. 20% less than the top men.







Smoothened graph: look how constantly she pedals! After 3:11 hours something happend: Power dropped off.





Power distribution: compared to the men the power curve is shifted to the left and it is less wide because of the more controlled and constant riding style.

Data from Thomas Hellriegel (DNF)

Not a good day for the champion: Although Thomas kept himself in check at the start, he couldn't maintain the desired power level of 300W. In the end he averaged 268 Watt. (no speed signal)



Power distribution

Data from Markus Forster, place 27

Too fast at the beginning: Markus started with 400W, averaged 312,5 in the first 1,5 hours (first split only 2 minutes down to the winner) but had to pay for his effort later. In the last 2 hours he averaged 243 Watts.



Fluent style: Markus pedals between 80 and 90 RPMS. His Power band is wide - from 140 to 400 W.




Cycling training plan | Triathlon training plan | Running training plan | Crosscountry ski training plan | Mountain bike training plan