The anticipation has built. The preparation has been done. The Tour is here! There are many great journalistic stage by stage comparisons, but few opportunities to look at the numbers - as cyclists we know that the length of a stage is only part of the story; adding an analysis of the volume of climbing adds another dimension to the grandeur and suffering. Cycling Tips have published a very detailed stage by stage breakdown of the Tour, which now includes the charts on this page. For comparison, you can see these charts for the 2020 Tour de France at the bottom of this page.
Have a look at my post on Free the Data to learn about why I'm sharing the source data I collect and prepare, and how you can access it to make your own stuff.
As always: remember that you can click on each of the charts and lines and datapoints for more information. You should also know that while you can view on your phone, the larger the screen, the better the view.
A comparison of distance and climbing
This chart compares the climbing metres of each stage with the distance travelled. The size of the circle is a ratio of vertical over distance to reflect average metres climbed per kilometre (larger circles being 'steeper'). The daily update version of this chart will accumulate winners names over the course of the tour.
Stage by stage
An alternative view, showing vertical metres and cumulative vertical metres per stage.
A comparison to last year
This year's Tour de France has more than 8,000 metres less climbing than last year's Tour, however the slope of the curve suggests a lumpier route.
Data quality
I'm reliant on a third party for the existence and accuracy of this data. I evaluate the quality of the data and sources I rely on prior to publishing.
Timeliness
Typically this post triggers a range of suggestions on how I could improve the accuracy of distance travelled or metres climbed, much of which is usually centred on some sort of weighted analysis of pro cyclist Strava files. This is theoretically feasible with one glaring problem. One of the important dimensions of data quality is Timeliness. This can be summarised as:
Is the data required available at the point in time when I need to perform my analysis?
The answer in respect to the harvesting of those Strava files is "no". This method clearly requires a statistically significant sample of pro cyclists to ride the road and post their files so I can collect them, analyse them and write about them. This is problematic if I wish to write about the route they will ride prior to their riding it. If it wasn't for time, everything would happen at once.
Consistency
My data source for the 2021 Tour (the SBS ŠKODA Tour Tracker app) is the same source I used for the 2020 Tour.
Completeness
For the time trials on stage 5 and 20, the Tour Tracker does not publish total metres climbed. As a result I examined the route profile maps; stage 5 has a few lumps, and stage 20 is essentially flat. On that basis, I worked out the elevation difference between start and finish, with a buffer for the lumps mid-stage to arrive at a value. It's unlikely that elevation will be a decisive factor in the ITTs for this year's Tour.
Title image by stokpic
コメント