Trying to make sense of nonsense

Beware of the tweeting cyclist

Tour de France - Grand Depart

Which of these riders will be the next to get up close and personal to the asphalt? Picture by Laurie Beylier

Written for Significance


In the lull between the end of Wimbledon and the start of the football transfer window, two all-time favourites of the British, one of the most magnificent sporting events of the year takes place: the Tour de France (TdF). Unfortunately, it also appears to be one of the most underrated events, by the British at least. It might have something to do with France being the sole centre of attention for three whole weeks, the popularity of track cycling (even though road cycling can ride in circles around the track variant, very big circles where the TdF is concerned) or the lack of a British Tour de France winner. But with a rider like Mark Cavendish and his ability to actually win – as opposed the ‘cursed’ British tennis players – the spectacle ought to be at least as appealing as the Wimbledon finals.

One feature that does seem to grasp the attention is the spectacular falls that are bountiful in La Grande Boucle. The first week has been notorious for years: traditionally, it is filled with flat stages and the just under 200 riders are all but too eager for some early results in the race they have had their eyes on all year. This year was off to a maybe even more nervous start than normal as there were more flat stages before the riders reached the first real mountains: the Pyrenees. Moreover, the weather wasn’t exactly what you’d expect during summer in France, and wet roads and nervous riders are not the best combination. This shows in the number of riders leaving the Tour early, mostly due to injuries sustained by falls.

Figure 1: Proportion of riders still competing in the Tour de France by stage

Figure 1: Proportion of riders still competing in the Tour de France by stage

That some stages can be particularly dangerous is shown in Figure 1. The line depicts the proportion of riders that remain in the race. During stage 9, a very rainy day with a couple of small, but tricky hills, no less than 8 riders had to leave the race injured. Even more fell but managed to carry on, among whom were the victims of the most spectacular fall: Johnny Hoogerland and Juan Antonia Flecha who were driven off the road and into barbed wire by a car trying to pass them. Accidents like this can happen of course: in the past, riders have had to leave the race after collisions with anything from dogs, motorcycles to not-so-innocent onlookers. However, there might be a trend as to what sort of riders are more likely to get seriously injured. Maybe it’s the younger, less experienced riders who don’t know what to expect? Or maybe the opposite: the older riders who are defying age and maybe better judgement in trying to finish on the Champs-Élysées one last time?

There only one way to find out: throw some data and analyses at the problem. Wikipedia helpfully has a list of riders of this year’s Tour with some additional data on each rider such as country of origin, team and, more importantly, their age. After the first 10 stages, a trend does appear to be developing: older riders seem to be more likely to leave to the Tour: each year increase in age, increase the risk of leaving the Tour by around 10%. However, the difference is not statistically significant and with just 20 riders having been eliminated at this point, this finding is likely to be due to chance. The same holds true when looking at number of previous Tour entries; and there is no one ill-prepared team whose riders have a tendency to get up close and personal with the asphalt.

It gets more interesting when you add Twitter to the equation. Sporza, a sports channel from Belgium, has compiled a list of tweeting riders and a quick analysis learns that these social media savvy men are much more prone to fall: their risk of leaving the Tour early is over twice as high as their non-tweeting colleagues. The effect just fails to reach statistical significance, so when looking just at the numbers we can’t be sure whether we should issue a warning to be on the lookout for all tweeting cyclists.

Survival analysis: riders surviving in the TdF by Twitter presence

Figure 2: proportion of Tour de France riders still in the race, by Twitter presence (mind the change in the values on the y-axis compared to the first graph)

The picture changes when we take the whole Tour de France, so flat and mountain stages, into account: Twitter loses its influence on the riders’ capability to remain on their bikes, but age does seem to explain away a part of the apparent randomness in the quitting behaviour. It’s only a very small effect, but the older riders do seem to be a tiny bit more likely to give in to injuries and fatigue. When taking a closer look at the Twitter behaviour of the riders, it becomes apparent that the younger riders are just a bit more likely to be on there, thus explaining the initial result. If anything, this shows that if you include enough nonsense in an analysis, you will find something. With Twitter and Tour de France it’s obvious that there is no real cause and effect relation, in trials and medical research it might be a lot less clear.

So should we beware of the tweeting cyclist? Perhaps not, but as it just so happens that the entire top three in the overall classification has a Twitter account, you might want to keep an eye out for them.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: