The Holy Grail of Power Calculations

bbrauer

I had a mtb race today and I was reviewing the results this evening. Based on time, for me to match the leaders of my category, I would need to improve my times by 20%. However, I think most of us can conclude that power doesn't correlate very well with actual time in an event where there are so many variables, ed. total elevation gain, descending ability, rolling resistance, etc. So, simply trying to, say, improve my FTP 20% isn't going to cut it as an actual predictor of performance.

I found a link to Kruezotter on this board and was playing around with it a little. Pretty cool little app. Much better than analyticcyling. There are still some gaps in the data input fields that make predicting the actual average or normative power needed to achieve a specific time impossible.

• Kruezotter uses slope to determine time and power. There's still no way to factor in total elevation gain, which, unless you're doing an uphill time trial, is really the more important variable. For example, this race profile consisted of a lab of 8.3 miles with around 1500 vertical feet of climbing per lap. You finish at the same spot, so the total slope is still zero, but the substantial climbing still skews the numbers (as does slowing for the switchbacks and other course vagaries). I wonder if there's a formula that takes this into account.
• Rolling resistance on this calculators are still pretty much rough estimates and based more on slick tire figures. I was wondering if there was a way for DIY rolling resistance calculations of different MTB tire combinations on differenct types of terrain, ex. sandy, hardback, gravel, rocky and loose....etc., perhaps even a sampling to get a rough estimate.
• a database of power figures for particular body weights for particular course profiles would provide maybe the best way to bypass all the variables. I wonder something like this on the horizon, based on the popularity of sites like geoladders.com where users can upload their Garmin data to show rides, course profiles and times.

I guess I'm just trying to estimate what the leaders average power, or normative power was for a given time for a given course so I can have a figure I need to achieve to match it.