With Tony Martin's two stage wins and a stage win for André Greipel, German cyclists won 7 of the 21 stages. Below is Kittel's winning time and a comparison with our prediction.
- Stage 21: 3h 20' 50" (actual), 3h 17' 50" (prediction), 03' 00" fast (-1.49% error)
We'll take that error on such a hard stage to predict! Kittel's average speed is given below.
- Stage 21: 11.41 m/s (41.08 kph or 25.53 mph)
Cyclists hit speeds around 63 kph (39 mph) a few times on the streets of Paris. That's faster than they would drive on those streets! Seeing the cyclists loop around Paris made me want to return there for another visit.
The man of the hour is of course Vincenzo Nibali, winner of the 101st Tour de France. There were times when he simply looked on a different plane of athleticism compared to his competitors. I grabbed a screenshot as Nibali crossed the finish line (click on the image for a larger view).
I also grabbed an image of Niabli on the podium with Péraud and Pinot and the Arc de Triomphe in the background (click on the image for a larger view).
Nibali was the only cyclist to finish the entire Tour de France under 90 hours. His winning time was 89h 59' 06". With the 3 km removed from Stage 5 because of two dangerous cobblestone sections, the total distance biked came to 3660.5 km (2274.5 mi). That gives Nibali an average speed of 11.30 m/s (40.68 kph or 25.28 mph).
It was a great Tour de France! Stages were well planned, and there was plenty of cycling variety. Kudos to Ji Cheng, the first Chinese cyclist to compete in the Tour de France. There were 164 cyclists who finished this year's Tour de France, and Cheng came in last. But, he did what 34 cyclists who finished Stage 1 could not do; he finished the entire race. He may have been just over six hours behind Nibali, but he will surely return to China amidst cheers. Tour de France athletes are as good as it gets. I couldn't even finish a single stage of the Tour de France, much less come close to what Ji Cheng did in July. Congratulations to all those who finished!
My research student, Chad Hobson, helped make modeling this year's race a lot of fun. We are happy with the improvements we made to our model. Except for Stages 4, 5, and 6, where massive tailwinds made our predictions too slow (the rain-soaked and shortened Stage 5 at 7.79% was our worst error), all of our predictions came in under 5%, including five stages under 1%. Predicting stage-winning times isn't easy!