carleton

It's a computer that sits on my handlebars and tells me how sorry I am.

zigmeister

Strava tries to estimate power off uploaded data. I use a Garmin Edge 500. How accurate, I don't know?? But I can't believe that I'm only putting down 500w/hr...lol


Power w/kg hr Elev Gain
1 min 489 w 6.1 182 bpm 59 ft
2 min 476 w 6.0 177 bpm 50 ft
3 min 452 w 5.7 175 bpm 94 ft
4 min 417 w 5.2 175 bpm 107 ft
5 min 399 w 5.0 180 bpm 100 ft
6 min 403 w 5.1 179 bpm 108 ft
7 min 407 w 5.1 179 bpm 144 ft
8 min 404 w 5.1 179 bpm 182 ft
9 min 390 w 4.9 178 bpm 188 ft
10 min 377 w 4.7 177 bpm 210 ft
11 min 368 w 4.6 176 bpm 246 ft
12 min 360 w 4.5 177 bpm 264 ft
13 min 352 w 4.4 177 bpm 288 ft
14 min 346 w 4.3 177 bpm 297 ft
15 min 338 w 4.2 177 bpm 340 ft
16 min 330 w 4.1 176 bpm 333 ft
17 min 322 w 4.0 176 bpm 373 ft
18 min 312 w 3.9 177 bpm 386 ft
19 min 305 w 3.8 174 bpm 373 ft
20 min 307 w 3.8 173 bpm 396 ft
25 min 292 w 3.7 174 bpm 510 ft
30 min 270 w 3.4 173 bpm 575 ft

wjclint

I've only done it that way a few times and it was just to check the formulas I had come up with, but yes you can do it with Excel, with or without floating averages. I wouldn't use floating averages on the calculations, but if I wanted to smooth things out would apply it to a chart or tell whatever software I was using to smooth it for visualization purposes. I don't think I would use Zero if the speed is decreasing, unless it was decreasing above a certain amount - just because you are slowing doesn't mean you aren't pedaling. If you have cadence information you could export that too and when cadence is zero set power as zero. On a trainer though I don't think that would make much difference since you are almost always pedaling. Once you have the new calculated power column in Excel you should be able to import it back into whatever program you are using.

It sounds more complicated than it is. Once you have the formula the whole thing amounts to 1) export; 2) apply formula; 3) import.

carleton

I'll give this a try later tonight. I'm a data geek. My trainer work is based around cadence and HR anyway. As a wannabe track sprinter, the only real power numbers I really care about are Max, 1", 5", 10", 15", 30", and 1 minute.

jmX

You know, I bet it'd take somebody (me?) only a few hours to make a program that takes a .FIT file from a garmin edge, lets you tell it what trainer you were on, then spits out a new fit file with wattage numbers embedded in it.

Maybe I'll play with that idea this week.

graphs

It would be cool to get an idea of just how feeble my output really is without dropping the coin on a PM. This is of great interest.

CWM

I have been using trainerroad.com, but am interested to learn about Golden Cheetah. I will certainly give that a look, and if I can get it to work, will cancel my TR subscription.

Trainerroad is good software, but my problem with it is the lingering concern over how much it will cost me over the next few years. Five years of use will cost $600. Golden Cheetah is what, free? And as Android devices continue to incorporate native ANT+ support, I am sure a slew of clever mobile apps ala Endomondo will provide similar functionality for $2.99 or whatever.

If trainerroad had a one-time fee of $79 or so, I would buy it, but as it stands, its hard to justify the price given alternative offerings. I think they are doing a great job, but I think the business model is a bit flawed. All it will take is the aforementioned $2.99 app and the whole thing will seem ridiculously overpriced.

ancker

Nobody likes subscription based anything (I hate it). But unless you are buying a single good or service a single payment is never going to be a good business model. It's impossible to build a company when you have unpredictable revenue each month/quarter/year especially as a startup. There are ongoing fees associated with running a service like this. Not limited to hosting the webpage, data storage, buying test equipment, etc. And perhaps most significantly, time. No one can work for free. Even if you have a 'day job' no one starts a side business for the purpose of just breaking even.

The reason smartphone apps work is because it's known that app usage falls off after time. Studies have shown than a large percentage of apps downloaded to phones get used once or twice then never again. Usage is slightly higher for paid apps. They will release the first version and charge little for it. They will probably even update it a few times. But if there is a significant update or addition of significant features a new app will be released which you will have to pay for. While this isn't technically a subscription model it works the same way since they can predict what the uptake of the new app will be and can plan for regularly scheduled releases. They can also bank on a significant amount of people buying the app (Hey, it's only $1.99, why not!?!) use it once or twice, then never again. This is essentially free money as the customer paid for the service but aren't consuming any resources that the company purchased. Think of it as similar to the 'fat tax' of all those people who pay for gym memberships and rarely go. That's how gyms keep their monthly fees down.

TrainerRoad is actually unique in that it understand that most people are only slaves to the trainer/rollers during the winter months. They have an option to suspend your account for as long as you need to and cease payment. When winter comes back around, you simply reactivate and start paying again. Your data is all still there. (I don't see this on their website but I've read it from one of the owners. So I'm guessing this is still an option.) The only catch is that when you reactivate, if there has been a fee increase, you pay the new fee, not what you were previously paying.

(I'm am not affiliated in any way with TrainerRoad nor have they compensated me for anything I say about them or their service. But I am a user and a fan.)

recursive

My cycleops fluid trainer, FWIW, seems to slightly increase resistance at a given speed after it's warmed up.

CWM

I like trainerroad too. I understand why they are going with the subscription model. The new Texas Instruments wifi chip in the majority of new Android phones has ANT+, but it is disabled in firmware in all but the Sony Xperia phones. In the future that will change and I predict most if not all Android phones will have native ANT+ support.

The garmin APIs are published and it would be easy for a developer to build an Android app that does exactly what traineroad does, and charge some marginal fee. As soon as that app becomes available (and Endomondo almost does this already), trainerroad will be incredibly overpriced in comparison. A user who has an android phone only needs a 2.99 app to do this, he doesn't want to pay $40 for an ant stick and have monthly fees.

The amount of work involved in this app is not that much. An android developer could crank this out in a few weeks and would be happy to sell it at $2.99, even if he only sells 2000 copies. I respect what they are doing with trainerroad, and will support it until a better more cost effective solution presents itself. And it will.

Computrainer 5K

TrainerRoad 100-500??

Android Ant+ App 2.99


All do essentially the same thing.

jmX

A computrainer is completely different to what trainerroad or a simple app are.

Anyway the App already exists, from Garmin, and its $0.99.

wjclint

Can you point us to the Garmin App? All I could find was "Garmin Fit" which is "completely different to what trainerroad" is.

jsutkeepspining

i have the same thing. Over the course of about 20 minutes a certain gear can gain as much as 20 watts for the same cadence.

CWM

jmX. Your app looks good. The only problem I see with your model is that you require a custom hardware module that does the same thing as the Garmin ANT+ speed/cadence sensor. I realize you probably built it for your own edification with an Arduino board, but you should consider integration with the garmin sensors, as you already support the ant+ heart rate monitor.

This is a similar problem to the traineroad model, in your case you are requiring a proprietary "dongle" hardware, where trainerroad requires an ongoing subscription. The value added is that trainerroad provides a social network of other riders, and yours essentially adds graphics to the golden cheetah freeware app.