London 2012 & The Mens Elite Time Trial - A practical exercise in event modelling

Why do some people develop a passion for endurance sports such as cycling and triathlon while others prefer the ever popular field sports that we are force fed by television such as football, baseball or cricket? It's a discussion we've had many times with a few conclusions that have a real danger of looking like a case why "cycling is the ultimate sport".

You can love cycling for the equipment. Regardless of how good the rider cycling has a certain appeal for the bikes - the aerodynamics, the engineering, the aesthetics, the aerospace grade materials and the sheer cost. To this end it captures the imagination of those who might otherwise love motorsport, where the physiological attributes of the driver are seldom mentioned. Conversely field sports may appeal for their simplicity, accessibility and portability. When all you need is a pair of boots a sport can easily achieve mass appeal.

You can love cycling for the beauty. Televise a football game and you have a field of turf surrounded by an anoynmous crowd, every time. But televise one of the major bike races and you have near endless helicopter shots of some of the most spectacular landscapes on the planet. You have intrusively close motorbike shots of the key riders at all the key moments and almost frighteningly close contact with the fans. To their defense field based sports can claim some impressive stadiums usually located on the doorsteps of their fans...and it's never too difficult to find a beer or a hotdog.

You can love cycling for the physiological aspect. Sure, all high level sport involves a great deal of training and preparation but cycling takes it to another level. The hours in the saddle, the periodic monitoring of physiological variables in a lab environment and then the perfect distillation of all of this - the ability to directly measure riders power output at the pedals using a device such as the SRM. What other sport can put such an accurate magnifying glass to it's athletes? We almost return to the appeal of motorsport here... power, torque and test rigs used to run diagnostics on the latest engine.

But there is another reason you can love any endurance sport and it's the lack of randomness. Endurance sports reward all the hard work and preparation with a certain inevitability to the results. The upset is rare and perhaps 90% of the race is really in the taking of raw talent and preparing it to perfection. When Fabian Cancellara shows up at a prologue time trial the smart money wont bet against him. Nor Alberto Contador at a grand tour. (Nb. This article goes out in the winter of 2011-12 with the greatest respect to Tony Martin.) We're at the level of horse racing here where horses can have such consistent form the odds speak for themselves and handicapping is common-place, just to mix things up a bit. Back to cycling though and it's the relative importantce of the preparation, not the execution, that makes it so popular at the amateur, mass participation level. It's hard to win but it's easy to feel satisifed with a job well done in terms of training and conditioning.

The lack of randomness in cycling combined with our ability to model performance on a course offers some exciting possibilities. As an exercise in the application of course modelling and in recognition that 2012 is well and truly upon us we set about predicting key riders performances in the forthcoming London Olympic Time Trial based on their finishing times in the Copenhagen 2011 Worlds Time Trial.

Copenhagen 2011

To their credit the organisers of the Copehnagen worlds made it extremely easy to understand, with accuracy, the profile of the time trial course. We were able to build a course model based directly on the profile spreadsheet posted here. The essentially flat 46.4 km course looks like this...

The weather in Copenhagen on September 21st 2011 we recovered from a historical weather service. And the top 10 finishing times in the mens time trial were as follows:

Placing Rider Time
1 Tony Martin 53:42
2 Bradley Wiggins 54:59
3 Fabian Cancellara 55:04
4 Bert Grabsch 55:15
5 Jack Bobridge 55:57
6 Richie Porte 56:13
7 David Millar 56:29
8 Lieuwe Westra 57:02
9 Alexandr Dyachenko 57:03
10 Jakob Fulsgang 57:14

The Riders

In order to model these riders on some other course such as the Olympic time trial but without access to their power and aerodynamic drag data we had to convert their finishing times in Copenhagen into some useful numbers. We can certainly know their weight because most teams advertise rider height and weight statistics and we can add a fairly typical 6.8 kilos of machine weight into the model (thankyou UCI). But we are left with the two unknowns of CdA and watts of power output. For the purposes of this analysis we take the view that on predominantly flat courses the riders individual values of power and CdA are not so important, rather it is the ratio of the two which determines finishing time. We therefore assumed all riders had the kind of low CdA that befits a world class time triallist, 0.23 m^2, and solved for their power outputs. We calculated the following:

Placing Rider Weight (KG) Watts/m^2 CdA Watts
1 Tony Martin 75 2,089 481
2 Bradley Wiggins 69 1,943 447
3 Fabian Cancellara 82 1,961 451
4 Bert Grabsch 78 1,935 445
5 Jack Bobridge 65 1,983 424
6 Richie Porte 62 1,813 417
7 David Millar 77 1,817 418
8 Lieuwe Westra 74 1,761 405
9 Alexandr Dyachenko 63 1,739 400
10 Jakob Fulsgang 69 1,725 399

London 2012

The course that will be used for the elite mens time trial has been set out by the organising committee here. It is, by the way, a very nice part of London. We can see from the following GPS profile that it is significantly hillier than Copenhagen so it's a good thing we have ballpark figures for rider weight if not their exact power to weight ratios due to estimating power based on estimated CdA.

Now assuming typical weather conditions for London in August we can predict the following results. This is if the event were viewed as some sort of "rematch" among only the top 10 at the worlds - same weights, same ratio of power to CdA, different course. In fact, with the riders all using the bulk of their power to overcome aerodynamic drag at race speeds, we can expect the results to be very similar, notwithstanding the hillier course.

Placing Rider Time
1 Tony Martin 50:24
2 Bradley Wiggins 52:05
3 Fabian Cancellara 52:11
4 Bert Grabsch 52:21
5 Jack Bobridge 52:59
6 Richie Porte 53:14
7 David Millar 53:30
8 Lieuwe Westra 54:03
9 Alexandr Dyachenko 54:03
10 Jakob Fulsgang 54:14

It will certainly be interesting to see what happens. And if August 1st doesnt play out as above, well, it would be tough to blame the laws of physics. The truth has to lie in the riders numbers...just as it lies in all of our numbers - power, weight & aerodynamic - every time we get out on the bike.