Amateur Statistical Analysis Pt. 1

I was considering some data today at lunch. I guess I should back up... there was the epiphany this morning about the final sprint - why didn't we have a segment for that? After all, it's clear we're a group of riders with varying strengths. We have courses for hills, time trials, distance... why not sprint? It seemed logical to create a segment out of a sprint we routinely do, so voila! I made a segment from the final half-mile of the TT course.

Reflecting upon my new segment, especially after looking over the leader board, it struck me it might be interesting to compare top speed to sustained speed... specifically, all-out sprint vs. time trial.

This in itself sprang from a conversation I had with Troy the other night, or rather, me railing against him about his mindset. I didn't want him to sound negative about his ability to ride fast on a sustained ride because I believe he's probably one of the best sustained guys we have in the group. He might be inexperienced, but talent is not his problem.

That's when I hit on it... I wanted to divide the TT rate of speed by the sprint rate of speed to get a percentage of top speed at sustained pace. My thinking on this was pretty direct - I felt the higher the percentage, the better.

I did the chart. Check the data yourself.

Cyclist	Sprint	TT	Ratio
Shep	34.00	26.80	78.82%
Miller	31.30	26.70	85.30%
Sowdet	29.90	26.70	89.30%
Deckard	29.50	26.90	91.19%
Heatherly	29.00	26.50	91.38%
Bartley	27.80	24.20	87.05%
Freeman	34.50	25.90	75.07%
Young	26.30	24.20	92.02%
BurrisA	26.30	24.60	93.54%
BurrisL	25.30	23.10	91.30%
Jarrard	25.30	22.50	88.93%

In a perfect world, these times would reflect an individual time trial divided by sprint speed. Obviously some of the rides here were pace line, not TT. The riders in yellow were team rides; the green were individual. Whether or not that invalidates some of the data I don't know, but I think you'd certainly have to value it differently. I'll leave that to the statisticians. Suffice to say, you should be able to compare TT to TT and team to team. Maybe that means we don't have enough samples to be valid, but c'mon, let's roll with it. 

The data somewhat explained, it's time to interpret. This is the part that fascinates me, because there are a lot of ways to approach it. The old saying goes, "Lies, damned lies, and statistics", and one could make these numbers say about anything. For that reason, let's take a bit and explore some of the possibilities. 

Possibility One: riders with higher percentages are in better shape.

Possible, but not conclusive. Looking at the individual times (which are more useful for measuring individual fitness) both Troy and I score high percentages, and I'd say we are two of the guys who put in the most miles. James's time is misleading; he hasn't posted all his Strava data publicly, so his true numbers aren't up. My guess is his ratio is just as high or higher. Scott on the other hand is a relatively new rider. He's talented, but this is his first year. He's still getting his bike adjustments dialed in. His total mileage would not be as high either. 

Possibility Two: riders with higher percentages are more efficient.

Once again, possible. Riding has a skill set to it, and there are things one can do to improve energy expenditure, a.k.a. efficiency. This would explain how someone could sustain a higher output over time, though not have a great top-end speed. In short, this type of rider would have an optimal zone of speed higher than his opponents provided primarily by exercising better technique. 

Possibility Three: riders with higher percentages just don't have high speed ability.

This possibility branches from the concept of fast-twitch vs. slow-twitch muscles. Sprinters have more fast-twitch muscles than other riders. It is a fact that different riders would have differing amounts of each muscle type, meaning riders will have natural tendencies of proficiency that vary from other riders. Some people are built for explosive bursts of energy. They can be taught to stretch the effort over time, but going fast will always feel more natural. 

Possibility Four: riders with higher percentages have higher levels of motivation.

The old adage in athletics goes that winners just "want it more" than the other athletes. While this is a wildly oversimplified way of looking at athletics, there is some kernel of truth in it. Motivation isn't just the ability to accept more pain than other people; it's also the dogged pursuit of overcoming one's shortcomings every single day. What this means is some athletes will do whatever it takes to win at all times. Constantly being aware of terrain, working drills, paying attention to form, pushing one's self when riding alone... these are all signs of high motivation.

To be continued!