Sometimes those online velocity predictors are DEAD ON...
 

I did a 10-mile time trial tonight at Lowe's Motor Speedway in Charlotte, NC. In the days prior to the event I used my power meter and this online velocity predictor to predict my time based on how much power I figured I could produce in the aero position for around 20 minutes.

It's essentially a flat track, and I assumed negligible wind, so I plugged the following into the velocity predictor:

Average power: 310 watts
Rider weight: 163 pounds
Bike weight: 18 pounds
Tire type: clincher
Position: aerobars (I have clip on bars)
Grade: 0.0%
Distance: 10 miles
Temperature: 70 degrees
Elevation: 675 feet
Trans effect: 95.0

The predicted time based on that calculation was 22:02.04.

Anyway, so I ended up averaging 308 watts instead of 310. The temperature was about 70 degrees and there was essentially no wind. According to the velocity predictor, my finishing time should have been 22:05.28. My actual time was 22:05.52. That's pretty amazing if you ask me...

Admittedly, there's a lot of "hand-waving" going on with the predictor (ie. what exactly constitutes an aerodynamic position, and where does the trans effect of 95.0 come from, etc), but still it's pretty cool to see that I can potentially predict future improvement with accuracy if I can increase my wattage or decrease my rolling resistance (i.e. tubular tires or a full TT bike).

--Steve

In the (I believe) September issue of Bicycling, the MIT cycling team talked about how they predicted a team time trial result (College Nationals?) within a few minutes or seconds based on similar information. FYI.

Yeah, I read that article. The science of the team time trial is pretty fascinating, IMO, given that not everyone has the same power output.

--Steve