Hmmm. In my day job I'm a professor who usually teaches graduate courses but sometimes teaches an undergrad course. In the undergrad course everyone starts off as a novice but a few will continue on, and I can't always predict who will continue on. In that class I simplify things but I tell them that I'm simplifying. They seem to understand that they're using simple tools in a simple way and that there's more to it if they progress. Do you think the audience here is static, and that no one here will ever progress to needing (or wanting) to do QA analysis, or sprint training, or drag estimation? If so, maybe I could see your point. However, if you think that the audience changes, and some who are starting off will progress on to doing more data intensive things, it seems a disservice to say "yeah, buy a one-sided power meter because canam73 has decided you won't be doing anything that requires high data quality." Except for achoo. I definitely believe he doesn't need high data quality.

Pretty much.

There you go again. See, you just don't need a 2 sided PM to do sprint training. In fact, you don't even need a power meter. It is a nice tool to have, though, so for kicks and giggles, talk me through the terrible scenario where some left leg only sprint data ruins the career of a cat racer.

You keep posting for the 3 percent that will do some QA analysis or drag testing, I'll keep posting for the 97% that won't. I am happy for those who find utility in you pointing out that ability in a more accurate unit. I just think many of your posts sound like a professor who can't see the forest for the trees.

Seems he's posting to toot his own horn, actually. Should be publishing in a scholarly journal somewhere where the readers care about QA analysis and drag estimation exercises.

I believe he has, and is well regarded in it. I just question it's value here. Or at least believe a strong counter point needs to made.

Although I have to agree with you on the horn tooting.

@RChung you don't think UCI pros that have huge contracts, endorsements on the line as someone who cares about high quality data? I guess Team Sky with consecutive wins at TdF are morons. Their training and understanding of cycling hasn't progressed enough to your standards. Team Sky wouldn't sacrifice their training for a $699 powermeter sponsorship if measuring only one leg was as big of an issue as you implied.

Offtopic, amazing to see how dramatic power output is with cadence. I'm not planning on buying a powermeter, but this video is motivation to continue to focus on my cadence over slower, small cogs.

I believe I addressed this back in post #22 . Most WorldTour riders do huge training volume so the importance of a single training session isn't that great, they don't do QA analysis, Sky has access to wind tunnels to do drag testing, at this level they mostly already know how to pace, and most of the riders had been using power meters in the past so they already know a lot about themselves. So, perhaps paradoxically, no, I don't think they care all that much about high quality data from their everyday power meters. They have different needs and different capabilities (chief among them, wind tunnel time), so it doesn't surprise me at all that they're making decisions based on different criteria. Earlier in this thread someone asked about Power Taps in the pro peloton. In a world where a flat means a wheel swap, Power Taps don't make much sense -- but that doesn't mean that that reason is relevant to us.

Uh huh... whatever, not making convincing argument over the importance of using a SRM over stages still if you concede the high power, long training done by UCI pros don't require it. If they don't require it, don't see why anyone else would require better data than they do. Wind tunnel isn't excuse.

Oh, I didn't think you'd find anything I wrote to be convincing -- you seem like you already had your mind made up. However, I think that what ever any pro team achieves or accomplishes this year won't depend on what power meter they're using -- or perhaps even whether they used a power meter or not. That applies whether they're using a Stages or a SRM or a Quarq or a Vector -- or were going "bare." OTOH, I already know of amateurs who would not have achieved their goals over the last few years if they hadn't used a power meter. Different needs, different capabilities, different goals.

To me, the other elephant in the room aside from the top pros at Sky riding with powermeters for years and having a data base to draw upon is...just because the Sky team trains with Stages, doesn't mean every single rider hasn't performed baseline testing comparing power output to say a Powertap with a Stages meter. This would be simple. Install both, each rider perform the test baseline and this would expose any leg bias and absolute power difference. With this understood than any discrepancy in Stages versus a meter that computes aggregate power of both legs would be reduced to insignificant.
So it is what it is. Of course the Sky team trains with a Stages device and who can argue with their success and of course they are being paid to train with it but simple correlation testing would expose any discrepancy and easily be factored into the power numbers they use.

This is what I was trying to get at.

Your position is that

2X = X + Y for the case that X does not = Y.

Have fun with that.

I have a stages power meter. What is said ^^^^^ here is entirely plausible...if you're 100% focused on power during a ride - such as when you're on a trainer. When I first got the PM, I found myself pumping the left leg harder. However, that tendency tapered off after a few rides.

Now that I'm on the road (and off the trainer) - I can't possibly dedicate that much attention to watching my power numbers - and therefore that subconscious asymmetry is completely eliminated.

Actually, it's

2X approximates X + Y for most cyclists with no underlying issues.

There is simply too much hand wringing going on over something that isn't an issue for many people.

Right, I agree that this probably isn't a real concern, except maybe in odd circumstances most people won't see. For example, if you were doing a VO2Max or MAP test with a ramp protocol (where the protocol depends on ramping up your power at pretty precise steps on a trainer) then I could see that someone focused on making the steps might bias their power production to the left leg -- but most people don't do these kinds of tests. When I first got a power meter many years ago I used to try to chase power output but I got over that relatively quickly, and I suspect that most riders will learn that, too, once the novelty of seeing power numbers wears off. There are reasons why the Stages isn't appropriate for everyone but I don't think this particular fear is one of them unless maybe you do all of your riding on trainerroad trying to hit a target.

[Edit:] Which is not to say there's no difference between the Stages and other power meters in terms of data quality. I think the Stages can be a good gateway drug.

You know this how, I have not seen any studies on this other than the one posted earlier in this thread.
Form what I've seen on other boards, the imbalance is less the closer someone is to their threshold. Down at lower power the imbalance is greater. I've already said I don't think it's a big problem, just something to be aware of.

It would be interesting to see some data on this, but we are fairly symmetrical beings so the assumption that one leg is roughly equal to the other doesn't seem far-fetched. While not an unbiased source, Stages claims,

"Measuring power at the left arm requires one assumption — that your left and right legs have a balanced power output. We have found with thousands of miles of testing, that power differential between legs has no significant influence on ride data and its value as a training tool. This assumption is the tenant to our ability to keep the complexity and cost of our system lower, all the while providing power measurement with ±2% accuracy (of the left leg's measured power)."

Even if one has a significant imbalance I don't think it really matters. Yes, your actual power as measured by an SRM or similar and your recorded power from Stages would not be the same, but your daily Stages to Stages measurements are what you working with. In the absence of studies showing an actual imbalance in normal riders that is all over the map, I think it's a non-issue. I would be fun to borrow some Vector pedals for a dozen rides or so and compare the data but I don't know how possible that is. I'm certainly not paying the $1500 or whatever to satisfy my curiosity.

Except there *are* studies that show actual imbalance in normal riders that is all over the map. Worse, the studies show that the relationship between bilateral asymmetry and cadence, power, pedal force, and fatigue status aren't constant or even linear (linearity would make the asymmetry easily predictable even if it's not constant). The conclusion from Smak et al.'s paper on this is: "For example, significant linear relationships existed between pedaling rates and percent difference in total average power per leg for only four of the 11 subjects and the nature of these relationships was different (e.g. positive versus negative slopes). It was concluded that pedaling asymmetry is highly variable among subjects and that individual subjects may exhibit different systematic changes in asymmetry with pedaling rate depending on the quantity of interest." And a recent paper by Bini and Hume showed that bilateral asymmetry doesn't always diminish with power.

But in a larger sense, if all you're tracking is training load for rides you mostly do all the time, I suspect these differences will be relatively small. Basically, you know how people say that as long as your power meter is consistent that's all that matters? Well, as long as you ride the same rides at about the same power in about the same gear ratios then your bilateral asymmetry will be about the same so your power readings ought to be mostly consistent; so if you think consistency is all that matters then this is the way to achieve that. If you understand how each power meter differs in their strengths and weaknesses, you can usually find a way to get around the weaknesses. Problems usually only arise when you think bilateral asymmetry is constant (or maybe linearly related to power) and then use the Stages for things it's not designed to do well on, and most riders don't do those things very often (or at all). Just be judicious in what you're using it for, don't overestimate what it's good at, and you should be okay.

Okay, there are studies, but how relevant are they and what do they actually show with respect to asymmetry and the usefulness of a left arm only power meter? The first study you cite isn't particularly useful in answering this (haven't read the other one yet as I can't find the full text). First, it focused on pedaling rates at a constant power output. It also used simulated riding which is unlikely to be representative of actual cycling and continuously variable power output.

"The Velodyne simulator also enables cyclists to work at a constant work load, or power output. This is independent of cadence. Thus, at a constant workload of 300 watts, if the cadence is increased, the rear wheel resistance decreases. Conversely, if the cadence is decreased, resistance increases. This is a valuable tool for conducting experiments at constant power output in the lab."

Furthermore, they report significant differences but not effect sizes so without running the stats it's hard to know if the effect size is meaningful in the real world. A 49/51 asymmetry is not that relevant. A 45/55 would be if it's not consistent and changes. The study did not address this. They did include data for 4 subjects with two showing very similar power output in each leg and two showing more variability. That's a very small sample but it does suggest that there may be no way to predict how this might affect any one individual. It's also not possible to determine if the experimental protocols or the velodyne had any impact on the data.

Finally, their conclusions basically say asymmetry is irrelevant - granted, for these types of simulated studies. (Yes, I actually accessed and read the full study.)

"With all but one of the asymmetry quantities being unaffected significantly by pedaling rate changes, there is no need to incorporate any systematic asymmetry changes to accurately simulate cycling biomechanics at different pedaling rates."

In other words, not much help for the issue at hand. What we really need is some moderately long term data from actual road riding using something like the Garmin Vector pedals. The second study may address this. If you are aware of any studies like this please share.

This study, Bilateral pedaling asymmetry durin... [J Sports Med Phys Fitness. 2007] - PubMed - NCBI, might be a bit more useful. It also found asymmetry but found it to be consistent (related to the dominant leg) and higher "crank torque" reduced asymmetry. This would tend to support the hypothesis that data from a left only power meter would be consistent (doesn't flip-flop) and therefore usable for training purposes. Further studies are certainly needed. Not my field so I leave it up to others. However, if someone wants to loan me a L/R power meter I'll be happy to collect some data in conjunction with my Stages and share it.

I think I already addressed that: as long as you stay away from the things it's not designed to do and you're mostly just tracking something like training load, you ought to be okay. Problems will only arise if you think (as you seemed to above when you claimed that there was an absence of studies) that asymmetry is either constant or simple to adjust for. Just be judicious and don't overestimate its strengths.
I linked to Smak because it was one of the earlier articles, has received a lot of citations, and showed that the pattern of bilateral asymmetry isn't simple. I think Carpes and Faria have a review of bilateral asymmetry in cycling (and, IIRC, running -- but I assume you don't care much about the running piece), which shows similar asymmetries with changing power output, fatigue state, and either pedal force or crank torque.

Yeah, that's amazing how when he pedals faster in the same gear, his wattage goes up!

No kidding... an 18% increase in speed is going to require more of a power difference than that anyway... you know, that whole air resistance thing we get outside, cube relationship and all that jazz.

Just an observation, I seem to remember Wiggins and Froome staring a lot at their head units for their power output while climbing on many mountain stages. Seems to me with how close in ability all the riders are in the TdF that accuracy in their power readings is pretty important. Sky must think the Stages PM is suitable, I can't believe that they are getting paid that much by such a small company (outside of free equipment).

BTW, Verve's InfoCrank store just opened for those on the preorder list.