Predicting Climb Times – Results

First off, let’s just get one thing out of the way. No, this wasn’t a scientific experiment. Instead, let’s call it a fun experiment and get on with it.

I set off with my good friend Jody from Ukiah around 9am Saturday morning. Temperatures were already around 80 degrees Fahrenheit and on the way towards the upper 90’s. We rolled across town, cyclecrossed through a vineyard, through a drainage ditch, and onto Orr Springs road. We hit the climb full speed and I settled into a tempo. My legs felt good and I tried to keep a steady pace, but my heart rate remained about 5-10 bpm faster than it was on Cavedale. I crested the first summit at 24:55, beating my predicted time by about 2 minutes. Success or failure?

Here was my relative power in Watts/kg on Orr Springs:

Gradient factor = 2 + (% grade / 10)
2 + (7.9% / 10) = 2.79

Relative power (Watts/kg) = VAM (meters/hour) / (Gradient factor * 100)
1180 / (2.79 * 100) = 4.2 W/kg

My predicted relative power was 3.9 W/kg, a difference of about 7%. This seems like a fairly large discrepancy. Any number of factors played part, but here are a few:

  • Higher heart rate – possibly due to warmer temperatures, but also an indication I was putting in a harder effort.
  • Lower weight – although I didn’t weigh myself before each climb, I did notice some weight loss over the course of the week.
  • Smoother road surface – Cavedale had several potholed sections, while Orr Springs was in good condition.
  • Mental factor – since I knew I was being tested, it’s very likely that I simply went harder.

Many of these factors can never fully be eliminated when attempting different climbs on different days in different conditions, but I think the mental factor could eventually be eliminated through successive tests. Eventually I should get to the point where I am pushing myself close to my limit each time.

I guess I’ll just have to put together another test!

This entry was posted in Analysis. Bookmark the permalink.

One Response to Predicting Climb Times – Results

  1. Fred Moyer says:

    A 7% error for that segment seems like a reasonable discrepancy. The grade levels out and goes negative for a little bit towards the end which will produce some error in power level calculations. The climb speed is low enough that drag won’t be a factor, but a 30 minute climb will mean some more variation in output as opposed to a 10 minute climb I would think.

    I’ve been doing some small experiments with calculating power output relative to heart rate on the McCullough Rd. segment ( The grade is a bit steeper, but it is surprisingly steady, and largely free of wind. I calculated theoretical power for a rider in my club and compared it against his power meter measurements for that segment, and it was within about 1%. I think there are some good segment candidates on Strava that you can use to get a pretty accurate measure of your power output at lactate threshold.

Leave a Reply

Your email address will not be published. Required fields are marked *