Analytic Cycling Accuracy?
 

Has anyone with a powermeter compared their recorded power output on a climb to the generated output of analyticcycling.com? Plugging in the data for a local climb and all of my info, it puts me at 4.5W/kg. My gut says this is pretty darn close to accurate, but I'm just curious if anyone has compared the two side by side.

Well, there are two problems:

1. You need to know what the wind was blowing at and which direction
2. It only allows one setting for gradient, so if you rode out over many hills and rode back over many hills, what do you put for gradient? For it to truly be accurate, it requires you to know your speed (and the wind direction/speed) over every diff. gradient on your ride.

1. I recorded my PR on a windless day.
2. I put in the average gradient for the first 3 miles of the climb. That's what I always time / record. My time rarely varies more than 30seconds slower when it's windy.

So basically, is it accurate for shorter, consistent efforts where the weather normally doesn't factor in?

Not the analyticcycling algorithm per se, but there have been papers published validating the model used there. The results show that the model is accurat to better than 2%, which is about the same as powermeters.

The greater inaccuracy is going to be getting the precise data to enter, i.e. the exact slope of the road, exact length, exact length at each gradient, etc.

I just did a steady grade hill sprints workout. Chances are, sprints are not going to be as accurate as a long steady, lower wattage effort. While my power meter (SRM) was consistantly averaging 700+ watts over 30+ seconds, Analytic cycling predicts 590 watts. Does this mean it is no good? Nope. I don't know the exct air density or what my drag coefficient was, etc. If you know all of those numbers, then great, otherwise, you should take it as a rough estimate.