Throughout the history of cycling, many strategies have been put forward to win cycling’s grandest race, Le Tour de France. It seems that one's best bet at understanding how the winners win is to analyze what we know about the results. The Tour posts finishing times for all of the stages, but real data filtering and aggregation needs to be done to understand what is really going on. These charts and descriptions should display some interesting features about the race and give some insight into the way the best finishers distinguish themselves from the pack and compete for the yellow jersey.

Individual Rider Overviews:

 Stage by Stage Time over Average 

This chart may seem like a mess when you look at it, but drag the left edge to the right and click the blue down arrow next to name. Then select your favorite rider, and you will be able to see how he did against the average on a stage by stage basis. 

 Stage by Stage Placing 

This chart displays different riders' finishing place for each stage. Here I have Lance Armstrong selected, but you can again use the blue triangle to select one or some of your favorite riders.

 Total Time Over Average by Stage Type 

This chart shows the total time that a rider was over the average in a given category. Lance and Alberto are dominant climbers, so they gain a huge amount of time on the field. This is denoted by the large negative bar. Fabian is a great time-trialist, so his time trials allowed him to gain a little more time on the average than Lance or Alberto, but he did not gain nearly as much time as them on the field in the high-mountains. (Again, to change the riders in question, click the blue triangle next to name and check the box next to the riders that you would like to examine.)

Climbers win the Tour:

After finishing my analysis of the rider-by-rider data, I was curious to see what factors a Tour champion must have. To this end, I analyzed a few variables against a rider's final position in the Tour. This yielded some surprising results.

 Finish by Stage Placing across Stage Type 

This x-axis of this chart is the mean of a given rider's placing after each stage. The y-axis is a given rider's final position in the Tour. I have set the frame series to stage type. This chart should therefore tell us if a certain type of stage is more important to winning the tour than another. The high-mountain stage has a much higher correlation than any other type. The riders whose average finishes were low also got low finishing positions in the Tour and visa versa. This clearly confirms the belief that doing well in the Tour depends on good high-mountain stage performance.

 Distribution of Times by Stage Type  

This chart gives us further insight into the importance of the high-mountain stages. Across the x-axis, time distributions against the average are displayed by stage type. As you can see, most riders finish very close to the average on the flats. This shows that there really isn't much time to be gained or lost on the flat stages. On the other hand, the high-mountain stage times have a high variance. This indicates that there is time to be gained by performing well in the mountains.

 Finish Position by Average Time over Stage Type 

This colorful chart reinforces the same principle that was demonstrated by the previous two. The bright blue-green dots display a given rider's time over the average for a given stage. The dots are tightly bunched, which shows again that there is not much time to be made up on the flat stages. They are also arranged in a verticle line, meaning that there is little to no correlation between doing well on the flat stages and doing well in the Tour. The red dots display the high-mountain stage times compared to the average. These are very spread out and display a significant correlation between mountain performance and doing well in the Tour.

Does Having a Good Team Help a Team Leader?

It seems to be a classic Catch-22 in cycling. As a team leader, one wants to have riders who are strong enough to pull you to victory, but long pulls on the front and tiring water bottle runs often leave domestiques fatigued and drained. It would seem that a hard working domestique would help his team leader perform well at the cost of his own finishing position. This would mean that a good finishing position for a domestique could hurt a team leader. In the following charts, I will examine team performance and relate it to team leader performance. 

 Team Performance by Stage 

 Team Performance by Type of Stage 

These two charts display the team performance against the average by stage number and stage type. The metric value is the sum of all of a given team's riders' times over the average time for a given stage or stage type.

 Team Leader Final Position by Team Time Over Average

This chart shows an interesting relation between team leader performance and the total team time over the average total team time. At the extremes, there is a very high correlation between team performance and team leader performance. The team leaders on great teams do well and the team leaders on bad teams do poorly. (Note: Bad teams could simply be teams that focus on the green jersey. Sprinters' teams tend to lose a lot of time in the mountains.) However, between these extremes, there seems to be almost no correlation between team and team leader performance. The team leaders on the decent teams have a wide variability of finishing positions. This shows that, for a team leader, there are advantages and disadvantages to having a team that performs well.

Try for Yourself: 

 Tour de France 2009 Data 

In the above dataset, I have compiled all the information you could need to discover just about anything about the 2009 Tour de France. I have everything from riders' king of the mountain points to what wheels they used, so try it out. All you need to do is click on the data set to start building charts of your own.

Citations:

http://www.letour.fr/2009/TDF/LIVE/us/100/classement/index.html

http://www.slideband.com/slide/2615_Bikes-of-the-2009-Tour-de-France.html

http://www.reynoldscycling.com/index.php?p_resource=articles

http://en.wikipedia.org/wiki/List_of_Tour_de_France_winners

http://www.angelfire.com/realm/cvccbikers/tour/tour_age.html

http://autobus.cyclingnews.com/road/2009/tour09/?id=/results/history/tdfhistory

http://www.cyclingweekly.co.uk/news/tour-de-france/369602/tour-de-france-2009-team-guide.html