Okay, now for the comparison.
- Stage 17: 4h 21' 36" (actual), 4h 42' 42" (prediction), 21' 06' slow (8.07% error)
What's bizarre about what our model did today is that it nearly predicted the peloton's arrival to zero error. The peloton came in 20' 10" after Trentin. In other words, our model missed the peloton's time by just 54". But we want to be able to model the winner, not the peloton!
Despite our model being slow today, it was great watching Trentin's ride. Below is the moment when today's stage was won (click on image for a larger view).
I got that screen capture when Trentin broke off from the lead group and took a peek back to see if anyone was chasing him. He accelerated past 60 kph (37 mph) to get in front. The race was over at that point. Check out Trentin's average speed.
- Stage 17: 12.74 m/s (45.87 kph or 28.50 mph)
The winning cyclist was once again about 2 kph faster than what the Tour de France organizers estimated for the top speed of the day.
- Stage 18: 5h 38' 19" (prediction)
The Tour de France has befuddled me a bit this year, sometimes in exciting ways. Regardless of how our model performs tomorrow, I'm anxious to see the monster climbs in the Alps!
No comments:
Post a Comment