Performance: rider data from the 2015 UCI World Championships in Richmond

As increasing amounts of data are made available to us, it’s possible to believe that some performances are predictable; a statistical modelling problem where power, coefficient of drag, air pressure and course profile reside at one end of the equation and a podium place, for a select few, is available at the other.
The elite men’s individual time trial title holders have done little to alter this perception. The rainbow jersey has changed hands among a narrow and slowly evolving group for the last decade. Until 2015, the previous 10 wins were shared amongst just five riders.

In contrast, the World Championship road race is recognised as a lottery. Extreme distance and demanding courses whittle down the field until the winner emerges from among the strongest riders. However, the vagaries of road racing are such that the podium is impossible to forecast with confidence. This year’s 261.4km elite men’s road race was no exception. Peter Sagan and Michael Matthews were expected to be competitive, but who could have predicted that Ramunas Navardauskas would stand on the third step?
Time-trials take place in a more controlled environment. This makes preparation and prediction easier. We can see evidence of this in the team event. Since the revival of the team time trial in 2012, some iteration of the Quickstep team has been on the men’s podium in every edition. BMC and Orica-GreenEDGE have been present on the podium in three out of four events and Specialized–Lululemon, later Velocio-SRAM Pro Cycling, have won every women’s team time trial since the event’s inception.
Performance in the team time-trial has also provided hints to the form of individual riders. In previous years we’ve seen Tony Martin make huge pulls at the head of his team’s train before securing the individual title later in the week. Vasil Kiryienka helped Team Sky to a bronze medal in 2013 before taking a respectable fourth place in the individual time trial. After watching Rohan Dennis’ heroic performance for BMC Racing in this year’s team event, many predicted that the Australian former hour record holder would be in contention for an individual medal.
However, the 2015 elite men’s individual event was a surprise to many. In the mixed zone, immediately following his ride, three times world champion Tony Martin was asked whether he had an explanation for the lack lustre performance of many of the favourites. “No”, was his succinct reply, but he may have simply said that they were not fast enough.
Martin languished in 7th place, off the podium for the first time since 2008 (where he also finished 7th). Tom Dumoulin was an unknown quantity following his demanding campaign at La Vuelta and perhaps the fatigue took its toll as he started strong then dropped to fifth position. Rohan Dennis finished in sixth place, failing to follow-up on his promising performance in the team time trial. In contrast, Kiryienka finally took home the gold medal after a string of near misses: fourth in 2013 and 2014, and third in 2012.

What happened behind the scenes? Which factors accumulated to create this result? Dr B Xavier Disley heads a company called AeroCoach and has published fascinating analyses of time-trial performance, ranging from Hour Records to hill climbs. Disley recently posted a graphic describing the pacing strategy of the top seven riders in this year’s world championship.

It’s clear that Kiryienka’s performance was dominant. The Belarussian rider was fastest at all time checks. Second place Adriano Malori appears to have adopted a very conservative strategy, starting slowly and building speed at each time check. Unfortunately for Malori, his slow start proved to be his undoing. The speed differential to Kiryienka over the first 16km was simply too much for the Italian to recover from. In contrast, Jérôme Coppel started slightly quicker than Malori, but faded relative to the Italian in the final time segment.
It appears that Dumoulin and Martin believed they were capable of medal winning rides. Both riders started with a winning pace. However, on reaching the second time check, their speeds had already dropped dramatically and from this point on, they were never in contention, lacking the power to maintain a consistent effort. Dennis and Jonathan Castroviejo appear to have paced the event well but simply never had the speed to challenge for a podium place.
Dennis’ performance is of particular interest. A rider who appeared so strong in the team-trial was off the pace for the individual event. How do the demands of these two events differ? The willingness of the younger members of the peloton to share their data on social media provides an interesting insight.

Cannondale-Garmin’s Moreno Moser was second in the Italian national time trial championships this year and rode both the team and individual event in Richmond. Soon after finishing, he uploaded his rides to Strava.
The TTT was held over 38.6km. Moser’s Cannondale-Garmin team finished 12th, 1’49 down on BMC Racing. Moser recorded a moving time of 47’10”, maintaining an average power of 348 watts with a weighted average of 367 watts.

The weighted average, relative to the average power, is worth noting as it helps to describe both how variable the effort had been and suggests the true physiological demands. The weighted average is effectively an estimate of the power a rider could maintain for the same physiological cost had the effort been perfectly consistent. Consequently, the larger the difference between average and weighted power, the more variable the effort. In Moser’s case, there was a 5.5 per cent difference between the average and weighted outputs for his team time trial performance.
During the individual time trial, Moser recorded a moving time of 1:04:55. The Italian must have pressed his ‘stop’ button shortly after crossing the finish line as this correlates closely with the 1:04:01 official time recording – good enough for 10th place overall, 1’32” behind Kiryienka. In contrast to the team event, Moser maintained an average power of 380 watts with a weighted average of 385 watts. This represents a 1.3 per cent difference, highlighting the fact that his individual performance was a much more consistent effort and, in all probability, and more accurate representation of his threshold power output.
The difference between the events is also clear in the analysis of the power files. Below, you can see Moser’s speed, power and cadence variation during the team time trial. The power spikes, followed by sudden drops in cadence, describe where the Italian made his efforts on the front of the Cannondale Garmin train before pulling off and dropping back.
This trace reveals one of the key differences between a team and individual time trial performance. In the team event, rider’s must spend much greater amounts of time above their ‘threshold’, which forces them to make a much higher, potentially more painful anaerobic contribution. Following these surges, they will aim to recover whilst riding behind their compatriots.
Again, in contrast to this, the power trace during the individual event is much smoother and more consistent, hovering at the boundary of sustainable effort. However, of interest in both files is the consistency of cadence. Moser appears to be most comfortable maintaining between 102-105rpm in both events, which is to be expected based the findings that higher cadences are generally more efficient as power output increases.
Correlating the speed data from Moser’s Strava file with the time checks and pacing data from the top seven riders in the individual event, it looks like Moser followed a similar pattern: starting fast, dropping off towards the second time-check, building again until the third sector before his speed faded towards the end. However, his consistent power output illustrates that this reduction in speed in the final sector was likely a result of the terrain, rather than fatigue or poor pacing.

For a stark illustration of the difference between road-racing and time-trials, compare Moser’s files with those uploaded by Lawson Craddock following the Elite men’s road race.

The Giant-Alpecin rider recorded an average power of 229 watts and a weighted average of 271 watts (18.3 per cent difference). The power file also illustrates the unpredictable, stochastic nature of riding in a large peloton on an undulating course.
Ultimately, our performance equation can only provide suggestions. Pacing and course profiles intersect with personalities and the will to win. Even in the more measurable disciplines, events conducted on the open roads, on a single day, mean that uncertainty still looms large in the minds of the competitors and audience and Richmond treated us to another entertaining display of power and unpredictability.
202 is a performance coach at HINTSA Performance, Geneva

The post Performance: rider data from the 2015 UCI World Championships in Richmond appeared first on The world's finest cycling magazine.