rick james said:
One thing I absolutely love is data and data analysis because it gives you amazing new insights. First, let's start with a hypothetical question. Did Froome experience a transformation in 2011? And if so, can we quantify it?
Thanks to Pro Cycling Stats, we have a wealth of information covering his entire pro career. The first step is to define some metrics for our construct. After much deliberation I chose to analyze flat time trials that are not prologues. That's because time trials are the "race of truth" and only depend on the rider's effort and ability. I skipped uphill time trials and prologues to reduce the number of confounding factors. I also rejected an analysis of speed because it isn't internally consistent, is noisy, and is dependent on external factors.
Instead, the metrics I chose were percentile ranking within the peloton (i.e., finished in the top 20%) and relative speed to the winner expressed as seconds lost per kilometer. This reduces the dependence on time trial length.
So using these two measures I collected as much data as possible.
http://www.procyclingstats.com/rider.php?id=140869 I won't lie, it took a while but I was bored...
I've lost the data, but I can remember the results if not the fine details. Basically, Froome went from consistently top 20% and lost 3-4 seconds per kilometer to a permanent shift to top 5% and a fraction of a second per kilometer. This shift happened in a roughly three week span in 2011. The distributions are quite tight with only one or two outliers in the data set. A t-test showed that p was some ridiculously low value which showed that the shift in the data was real.
Further analysis and making some reasonable assumptions, I estimated that Froome's FTP changed by approximately 15% in that short time frame.
Could there be an external factor which caused this change?
- Reduced CdA due to better position, etc? I concluded no, or at least not fully responsible because his climbing abilities changed at the same time and those aren't very reliant on CdA.
- Lost the fat? That doesn't work because flat time trials aren't reliant on power to weight, rather power to CdA.
- He increased his sustainable power? That fits the data quite well.
So now you have to look for reasonable ways to think of how he gained that power in such a short time frame. That would be pure speculation, but I've made up my mind. He's a dopity-dope-doper and it doesn't matter which cocktail of drugs or brand of motorbike he's using. I'm open to other explanations though... I just can't think of any reasonable ones.
And if you don't believe me, head on over to Pro Cycling Stats and start compiling the data for yourself. The data is open source and so was my compilation and analysis.
John Swanson