Beyond the concept of bragging rights as long as a power meter gives consistent results it doesn't actually matter if the figure if gives is accurate imo.

Yes I would care if the PM I used told me i was only making 100W instead of 300W but as long as my training meant the numbers went up it would not actually matter what the number was

You mean the Polar power meter?

From the feedback I've seen, the relevant players are: SRM, Powertap, Quarq and the Garmin Vector (dual sided)

The article looks at its reliability which is a much more serious issue. Accuracy is a problem if you are switching between PM's between years or have different PM's on bike. The reliability is a huge issue which is what this article was discussing. Also 3% off can be a huge issues, if i do a 600w 1 minute interval with 3% error that could mean 580 or 620, there is a huge difference between 580 and 620.

Yep, but the reason i posted this article is highlighting the importance of actual "real" testing of power meters. If you read the entire article there is a good discussion of testing protocols and how the only power meters to undergo lab testing is Quarq and SRM.

That doesn't sound right. Even if you lab test a product doesn't mean it's equal/superior to another that's been lab tested.

I'm sure all of the power meters have been "lab tested"

Are we really going to play semantics here? You know what i am talking about. I am discussing a third party professional lab, familiar with proper testing protocols, using a decent sample size, and posting the results in a peer reviewed journal. SRM and Quarq have undergone these kinds of tests.
If you can point me to an journal article where stages/P2M/Rotor/4iii etc was used i am all ears.

Can't find them either! Those weren't in my list for a reason :).

I am not trying to knock some companies, my general point is that we should do some research before making our purchases. Just because x bike company goes to the wind tunnel or x power meter company says they are lab tested does not necessitate they are providing good or truthful data.

I think one flaw in this "study" is that they went into it trying to determine if the Polar Keo meter would match the SRM. I think anyone familiar with powermeters isn't surprised that there is some difference between the two when they are located on different parts of the bike. the fact that the mean difference between the 10 riders in 2 separate tests were within 3 watts shows consistency in the meter; which is what you want. you would get the same exact variation with a powertap vs an SRM. there are certain losses which take place from where the power is inputted on the system all the way to the tire at the pavement.

I would add powertap to the list of well conducted testing proving their accuracy.

---

---

Kat Hunter reports on the San Dimas Stage Race from inside the GC winning team
Aeroweenie.com -Compendium of Aero Data and Knowledge
Freelance sports & outdoors writer Kathryn Hunter

Exactly, look at the PM reviews DCrainmaker does. Runs multiple PMs pretty much all the time and although he see's a variation in readings they are all pretty much the same. To me that suggests that all the supposedly non accurate PM's are in fact accurate.

From DC Rainmaker:


If all you need from a power meter is for the average power for a ride to be within 10 watts of correct or so, most of the time, then yes, most power meters will do the job. But if that is all you need from a power meter you could just use heart rate.

If you use normalized power (and if you use TSS, you do) or want to work on sprint or anaerobic power, or want to pace perfectly during races, or anything more complex than "go out and ride zone 2 for 3 hours"...then some of them aren't as good as others.

---

---

Kat Hunter reports on the San Dimas Stage Race from inside the GC winning team
Aeroweenie.com -Compendium of Aero Data and Knowledge
Freelance sports & outdoors writer Kathryn Hunter

The analysis was a bit more involved than just that.


Pedal vs Spider should be within 0.5 watts of each other


No, that does not necessarily follow.


Assuming you had a pedal meter, spider meter, and wheel meter, all working perfectly, you would expect much larger difference between the wheel meter (2% lower or so) than you would the pedal meter ( 0.2% higher)

---

---

Kat Hunter reports on the San Dimas Stage Race from inside the GC winning team
Aeroweenie.com -Compendium of Aero Data and Knowledge
Freelance sports & outdoors writer Kathryn Hunter

I'm with you on the criteria, but can you point to any such test of the Quarq?

You are talking about the difference between precision and accuracy. The power meter needs to be precise but not necessarily accurate in order to do what you need it to do. Unless we won races by comparing wattage numbers :-)

http://www.ncsu.edu/labwrite/Experimental%20Design/accuracyprecision.htm

That's only true if you have low expectations/standards.

In any case (before jackmott says it), if a device is precise, accuracy is only a matter of calibration.

Ok, this is my pet peeve

There are three reasons why a power meter being consistent/repeatable is not sufficient

1. The obvious reason is that should your consistent power meter ever break, get lost, or should you decide to switch for whatever reason, your old and new data will no longer match. This is a pretty big problem for anyone using a power meter halfway to it's fullest. Imagine: "How did I pace kona last year, when it went really well? Oh well shit, I don't know any more" or "What CTL was I sitting at when I crushed the state TT in 2010? oh shit, don't know"

2. Imagine the following 4 hypothetical power meters, which are perfectly consistent/repeatable, but inaccurate in the following 4 different ways. Each one will cause problems in evaluating the progress of your training, trying to hit zones in training, or trying to paces races, or field testing:


3. Comparing the overall aeroness of yourself vs others. I've actually used this once to find a setup mistake I made on my P3. Someone the same size as me was going faster on the same watts, with worse equipment. This caused me to look at my fit. It was a new bike, I thought I had replicated a proven good position, I had not. I fixed it, got faster. yay. In other words the ability to do field tests and get an accurate CdA can give you an idea if you are as aero as you should be.



And of course everyone should keep in mind that any power meter that was good at consistency would ALSO be accurate with a simple calibration protocol in software. So in practice it is very rare that you have a repeatible but wrong power meter, and if you do you can at least fix it yourself (with SRM and Quarq) or send it in to be fixed.

---

---

Kat Hunter reports on the San Dimas Stage Race from inside the GC winning team
Aeroweenie.com -Compendium of Aero Data and Knowledge
Freelance sports & outdoors writer Kathryn Hunter

Yikes, that Stages curve is ugly at < 100 seconds! I'd put this product in a category similar to the Polar chain sensor PM I once had, a gateway drug that will lead to the eventual purchase of a real PM. (I could probably use stronger terms than gateway drug for the Polar PM.)

So much emotion! ;-)

Please don't get Tom Anhalt started on the Polar chain sensor -- he'll sow you how wrong you can be. Weigh your chain! ;-)

---

AndyF
bike geek

Do you even want to touch single leg power meters on this one as well??? might as well let it all out going into the new year : )

No, just refer to the alex simmons blog post on power balance =)

---

---

Kat Hunter reports on the San Dimas Stage Race from inside the GC winning team
Aeroweenie.com -Compendium of Aero Data and Knowledge
Freelance sports & outdoors writer Kathryn Hunter

At this point, choosing to publish a comparison like this says more about the journal than about the comparison.

Some things you can do with a power meter only require consistency, that's true; however, there are things you can do with a power meter that do require accuracy across a wide range of power outputs. People who make the claim that consistency is the only thing that matters don't know what they don't know.