I was being tongue-in-cheek, of course, but do you cook? For various reasons, I have had to cook from American, English, and French recipes and sometimes I work in cups and tablespoons, mililiters and grams, fahrenheit and celsius and gas marks. Yes, it’s rather messy – but the underlying food tastes the same (or should), and after a while you calibrate the measurements so you hardly have to measure at all. This is pretty close to what training with various power meters is like.
However, there are some things you really do need to have an accurate thermometer for, for food safety, or accurate measurements for like when you’re using brining chemicals. This is a more demanding usage of measurement. (And, of course, not everyone brines their food, or needs to make sure the thing they’re cooking reached the right temperature.)
Doubling one side is something manufacturers do, but it obscures a potentially consequential source of error. It’s often better to be up front and not to obscure things: then the end-user, if they want to, can double the data on their own–but then they’re aware that’s what they did.
I’m a bit meh on the need for absolute accuracy in my power. I’m more consistent with a permanent turbo bike, and keeping my single sided power meters for outside. I want the power meters themselves to work and be accurate and consistent, but will suck up balance issues I may or may not have. If I’m doing outside workouts, it’s a range rather than a specific power target. If I’m in a race, I either have the legs or I don’t regardless of my theoretical FTP. Ideally/ money no object, I’d have total or dual power, but I’m not in that position!
Yeah, I totally get that. And to your cooking them you start with that super precise use (mixed units and all) and then admit to calibrating yourself to a point of no measurements.
I get your proposal here and think it has some merit, but in a rather narrow use case like your aero testing. Outside of that, the old doubled data is likely sufficient for the greater use by most users. I just don’t see the error present as a massive problem broadly speaking when we look at what people are actually using their PM’s for most of the time.
If/when people head down the aero testing road, it would be interesting to at least have an option for people to swap modes like you suggest, but it seems that is a rather narrow and limited frequency use case considering how few people we see even bother to test like that in the first place.
This is pretty much how I approach it. I have a Kickr bike for inside and my bikes have sram axs left crank PMs. Outside I’m typically doing longer zone 2 stuff and that’s easy enough with or without a PM and any interval stuff I’m targeting a range. And same as you with races. I think races I’m more using a PM to make sure I’m not overdoing it. I’m just using it to make sure I’m not going to be hammering away and then completely blow up
Yeah. I’ve tried to be clear that use case determines how much accuracy you need, and that right now most users’ use case doesn’t require much accuracy.
Here’s a fly in the ointment since you brought up TSS: that’s based on NP, which is dependent on a 4th order vector norm. Vector norms in general, and higher order ones in particular, aren’t terribly robust to errors in the data. The idea behind NP is to normalize variable efforts to a steady effort, but if you have variable accuracy across the range of power, the NP calc will also be variable. That hasn’t been much of a problem since TSS “dosage” doesn’t appear to be that critical, but it could be an issue if you were trying to use TSS in a more critical way. For example, early on in the development of NP, pre-ANT+ standard, there used to be power meters that would (or could) report at different time intervals. It turns out that NP is sensitive to that.
Of course, nowadays, there is ANT+, and almost everyone has standardized on 1 Hz reporting, so we don’t see the problem anymore. But not seeing the problem is different than there not being a problem, which is sort of like doubling a single-side: if you don’t see the problem, you don’t think about it.
It’s a good thing that most riders simply use their power meters for training.
No doubt. This seems to appear around here most often when people are comparing data from a LS PM to some other device, frequently Total Power (trainer most commonly) or maybe dual-side pedal for another example.
In those case, understanding the fudge in play is SUPER important and not everyone knows that up front. We see it nearly monthly and have to make sure that factor is included among the many others when people try to do these PM comparisons.
My mother told me that her grandmother cooked over a wood stove. She never used a thermometer, or thought about Fahrenheit or celsius. She could pick up a piece of wood, weigh it in her hand, and know how much wood she’d need to cook the meal. When the meal was cooked, the fuel had been burned: you didn’t waste wood. I think we had that conversation one day when I complained that our oven thermostat was off.
I have dual sided power pedals, but I think single side power would be just fine. The difference in the sides is minimal and I only really care about the one power number.
I do mine as part of my final prep before roll. The final beeps when my Bolt fully loads are the trigger for me. I do miss it on occasion, but fix it the next time I stop at a light or any other break.
I’m terrible about it. I really need to get into some sort of routine. But it’s hard to do that when the bike is stored inside and the outdoor temp is vastly different … or at least that’s my go-to excuse
I have both single and double sided PMs. Hard bunch ride this morning and balance was 51/49 and I have never seen the difference bigger than 52/48. I’d probably save the cash next time and go left only. Interesting reading @the_cog and @RChung takes on it though - appreciated.
Good timing on your question. I had an interesting conversation just earlier today with someone who’s been working on an aero sensor and app. He said not checking the zero offset was a source of error that made it look (to his customers) like his aerosensor wasn’t working instead of a user error. If you’re familiar with the operation of most strain gage based power meters, there are offset errors and slope errors. Zero offset errors can affect both single-sided and total power PMs, so that’s a pain, but most well-designed strain gage PMs nowadays (that is, excluding Shimano) don’t have much of a slope problem (Most.) However, differential pedaling asymmetry could produce variable slope error, which, I suspect, is why single-sided PMs don’t produce as reliable drag estimates as reliable total power PMs.
I keep my bike inside the house, so I manually zero offset after a few minutes of riding.
But if you’re just using your PM for training, it’s probably not that critical.
(There was an interesting case several years ago. You remember PowerCranks? The inventor of PowerCranks used to tout cases where his customers talked about how much more power they produced after training with them. (No, we’re not talking about that case – we’re talking about a different one). It turns out that in this case, the “personal best” power this guy had put out was almost surely a zero-offset problem).
I do a calibration after 15 minutes every indoor workout on my original hammer. I know they say you don’t necessarily have to, but temp and humidity can vary a lot day to day here.
For outdoor spins, while the bikes live inside, the bike goes outside when I get up, so probably half an hour or more to adjust to temp. Calibrate before I go.
A bit of a contradiction, not fussed about dual or total power, but do try to limit the inaccuracy of single sided by calibrating!
I think single side is just fine. Personally, I took up road cycling a few years after a knee replacement ended my ability to run and play Soccer. Started with a Stages left pm, but realized that I had a big power imbalance. Training with a dual sided PM has helped work on equalizing the leg strength to 50 /50. First dual PM was a Pioneer which had incredible real time metrics. Now using Assiomas
Those are a lot of words that don’t answer the question.
How do you use a DS power meter to reduce drag in a way you can’t use a single sided PM sufficiently? How do you adjust tactics using dual sided data that is superior to a single sided?
Google the ‘Chung method’ of aero testing (and note who you are replying too’. Needs accurate power data. I tried with single side PM… useless. Tried with double sided … Superb.
