Well, maybe that isn’t the best title. You see, there is this video going around saying that it is possible that a professional bike racer was cheating by putting a hidden motor in his bike. I am not really going to talk about the cheating aspect (my gut feeling says that it would be too easy to catch, so he is not cheating).
Actually, there was an interesting analysis by Ron at CozyBeeHive. Quite a thorough job. He even used Tracker Video Analysis. However, he didn’t use some of the nicer features of Tracker, so I figured I would do this analysis also. Plus, you know I love bikes.
Although I like bikes, I don’t know too much about professional cycling. I am just going to trust Ron about which guy to look at. Here is the whole video that he takes the clip to analyze from.
There are a couple of things that Ron missed with his use of Video Tracker. First, calibration point pairs are awesome. Basically, you can just take a couple of points in the background and find how the background moves. Then when you mark your objects, it will give you their motion. Here is an example I did with calibration point pairs.
The other thing is that you can actually do linear fits (and more) in Tracker itself. No need to export to Excel.
There are a couple of problems I had with Tracker though. First, I couldn’t see how to show a graph with two different objects. I am sure this can be done though. My second problem was that this track was quite long. For some reason, I couldn’t keep making an infinite number of calibration points. Ok, well, I just broke the motion into two parts. Here is the first part with the guy in question (Fabian Cancellara).
So, during this first part, Fabian has a speed of about 13 m/s (about 29 mph)
What about his acceleration during the first part? I could make a plot of velocity, but it would be easier just to fit a quadratic function to the data. Here is a description of finding the acceleration from position data, in case you are curious. Here is the same data with a different fit.
During this part, Fabian has an acceleration of 0.698 m/s2, or if you like it could be 1.56 mph/s. Also, for comparison, I plotted the speed of the lead biker guy. Here is his position with a linear function fit to the data.
From the fit, he is going about 10.8 m/s (or 24.2 mph). What about the second part of the motion? The two bikers (or should I say cyclists) have about the same speed as they did during the first part.
How does this compare to the values that are posted at cozybeehive? He got a value of around 25ish mph. So, in the same ball park.
Maybe the real question should be how does the speed of Fabian compare to the pack? I still could not figure out how to show both riders positions on the same graph in Tracker. So, I exported that data to Logger Pro. I also fit both a linear and a quadratic function to both sets of data. Here is what it looks like.
Both riders are accelerating just a little, but Fabian is going faster (with a greater acceleration also). I am not going to (yet) go into a power discussion like Ron did, but I will just say that neither of these biker motions seems too crazy.
This is how awesome Tracker Video can handle stuff (even if I didn’t do the best calibration job).