^^^ the app does say that it uses weather services and factors in the conditions, which Strava doesn't do (to my knowledge), but yeah, it's still gotta be piss poor in terms of accuracy.

That much? You're feeling generous.

If it truly worked one of the big boys would have bought up the software and would be selling it for a lot more than $8.

Yeah, you would think so.

I actually played with the trial version of this app and it was okay.. better than a basic bike computer and convenient since it's in your pocket, the speeds etc matched up fairly well with a wired computer. I think you just have to think of the wattage reading as an added novelty, I did really like how many variables it keeps track of though.

I mean, heck, I can probably get close to 15% by "it hurts this much".

I use an RPE Meter. There is too much variation if you try to calculate your perception using just your perception.

I think if you are grinding up a steep hill at under 10 mph on a day without a lot of wind, it could probably be fairly accurate.

I think there's an interesting philosophical argument about what "measuring something directly" really means. Does an SRM or Quarq measure force directly, or does it measure changes in resistance in a network of strain gages, which are then converted to torque? Does it measure velocity or does it measure time between reed switch activations? Or does it count quartz oscillations between reed switch activations, which are converted into time and then average velocity is derived from that?

I work for a large agricultural manufacturer, and grain carts are part of our product line. As an option, you can purchase a scale kit, which is basically a special axle spindle with a strain gauge built into it. In a nut shell it tells a farmer how much weight is in the grain cart, and from that extrapolates the total bushels of grain.

Your question about the innards being expensive...the scale kits we use operate on the same general principle as a bike power meter; just on a smaller scale. Our cost alone per unit is several thousand dollars, but the most important thing to take away is that you wouldn't believe the amount of development and engineering that goes into these things. The materials and processes aren't what's expensive, it's the research, testing, software development, and trying to make something extremely delicate but at the same time rugged.

You aren't going to rapid fire thousands of power meters a day off of machines in a sweat shop and have them be accurate and reliable. That's why you don't see a $200 power meter out there.

As far as your question about used units being so cheap, that chalks up to market forces. That $200 used power meter was probably bought at $800 when it was new, and now that there are smaller, lighter, wireless units out there there isn't much demand for them.

It could be done, but the market isn't large enough to justify the capital that would have to go in to creating the manufacturing processes that could pump out $200 power meters.

Low tech method:

Ride your bike. See who you naturally keep pace with. That's how fit you are. If you need more precision, ask one of the people who's about your same size and rides like you what their numbers say. That will be much closer than any dumb app.

You get the same accuracy as the new models.

Sealing against water has gotten progressively better with the new ones.

Weight matches the wireless models which followed, as in 412g for the SL wired like the SL+ ANT+ model, although you'd have the 12mm aluminum axle some big riders find flexy compared to the 15mm they upgraded to if it hasn't been back to the factory for service. The current G3 ($800 new) is 330g.

BUT the wired units only talk to a CycleOps "Cervo" aka "Little Yellow Computer" via a stay mounted shark-fin antenna, not your sexy Garmin. That uses a proprietary download cradle and software versus the Garmin's mini-USB cable which lets it mount like a hard drive for processing or upload.