Power differences between two people

brede

BigMat wrote:

Does your Dad ride on the front most of the time? Even with aero bars, he would be having to work harder than you if that was the case. Maybe you are outing yourself as a dodgy wheelsucker?!

Nope, we usually ride side by side, if there is any drafting he usually tucks behind me. (I mean, he is my dad who rides awesome bikes with his two sons, he deserves to have us do the work and him enjoy)

MiddleRinger

Tom M wrote:

... all you're saying is it takes your father more power to ride at the same speed as you, which is a fair indicator you are a bit fitter

Or Dad is as aero as a fridge.... lol

Bobbinogs

I am still not sure what the point of all the debate is? Two numbers appear on a headset after rides with a huge amount of variables, none of which are controlled or calibrated and yet everyone jumps in with a view on the cause of a minor discrepancy. My ten bob's worth...what the heck, ride your bikes. If all you and your dad can find to chat about is some numbers then pick some nicer routes ☺

philbar72

SloppySchleckonds wrote:

brede wrote:

He has garmin vector pedals, and I have a stages. Both are left only.

Sell them and buy a decent powermeter

sell them both, yes. 4III or Pioneer for single side. Power2max for chainset, powertap for hub.

thomasmorris

Bobbinogs wrote:

I am still not sure what the point of all the debate is? Two numbers appear on a headset after rides with a huge amount of variables, none of which are controlled or calibrated and yet everyone jumps in with a view on the cause of a minor discrepancy. My ten bob's worth...what the heck, ride your bikes. If all you and your dad can find to chat about is some numbers then pick some nicer routes ☺

The two numbers that appear on the headset have cost the users about 600 quid each. They've paid this money because presumably it's important to their training and / or performance. So it seems fair to discuss the differences.

Yeah sure, if you're just riding your bike with no interest in performance / speed / fitness then sure, it doesn't matter. But in which case why have power in the first place? Just because your primary interest is talking about the scenery, it doesn't means that's everyone's focus.

Plus, 20-30 watts is not a minor discrepancy. There are a few variables, but not really a huge amount and can basically be boiled down to five (weight, aerodynamic drag, rolling resistance, drive train efficiency and power meter error). Like I posted up thread you could pretty quickly narrow it down to find out where the difference is coming from. Knowing where that difference is coming from may give you an insight in to improving performance... which is the whole point of having a power meter.

thomasmorris

meanredspider wrote:

there's plenty of prejudice against single-sided PMs and Stages in particular.

Prejudice: preconceived opinion that is not based on reason or actual experience.

There is so many threads and posts online about issues with stages in particular, and single sided power meters generally which are proper reasons not to buy a stages.

In my actual experience of all seven people I know who have used stages, every single one has had problems.I do not now any in my personal experience who has not had a problem.

I think it's unfair to brand those warning against stages as prejudice. It's based on reasoning from evidence and actual experience.

Sure, there will be examples of people who have had stages and they've been fine. But in my actual experience that would make them a minority.

cougie

If stages is pants- what's the deal with sky using them ? I can't imagine they're that short of money that they'd be happy with a guesstimate of power ?

wongataa

cougie wrote:

If stages is pants- what's the deal with sky using them ? I can't imagine they're that short of money that they'd be happy with a guesstimate of power ?

Sky get to use Stages dual sided power meter that you can't buy.

DeVlaeminck

Tom M wrote:

In fact, a better question would be why would you expect there to not be a difference? Disregarding bikes and powermeters, all you're saying is it takes your father more power to ride at the same speed as you, which is a fair indicator you are a bit fitter, and possibly not altogether surprising assuming he's a good few years older and a little bit heavier. If I rode side by side with a pro of similar stature, then I would need to produce more power than them to ride at the same speed as I'm not as fit and have much less impressive muscular and cardiovascular system.

Come on now.

meanredspider

thomasmorris wrote:

meanredspider wrote:

there's plenty of prejudice against single-sided PMs and Stages in particular.

Prejudice: preconceived opinion that is not based on reason or actual experience.

There is so many threads and posts online about issues with stages in particular, and single sided power meters generally which are proper reasons not to buy a stages.

In my actual experience of all seven people I know who have used stages, every single one has had problems.I do not now any in my personal experience who has not had a problem.

I think it's unfair to brand those warning against stages as prejudice. It's based on reasoning from evidence and actual experience.

Sure, there will be examples of people who have had stages and they've been fine. But in my actual experience that would make them a minority.

Amongst all the noise to which you refer (which are valid about battery life and drop-outs) there's very little actual experience of power measurement issues. You're conflating the two. The OP isn't complaining about those reliability issues. There is, though, and has been lots of prejudice (and I choose the word carefully - I do know what it means) about single-sided PMs (of which, Stages is by a very long way the most popular). There have been a few questions about absolute accuracy but mostly corner-cases. My own tests vs my Neo (independently tested to +/- 1%) suggests, for me at least, the Stages is as accurate as claimed.

Personally I think it's funny to see suggestions if replacing two single-sided meters with single-sided meters.

I think the OPs question is interesting but entirely academic. Unless the OP and his dad are willing to do a load of tests (swapping meters, swapping bikes, swapping just tyres, swapping clothing, including helmets, riding different routes etc) then we won't know. Even then, if the meter contribution won't be clear as to which is contributing what. But, as I say, let's just blame the meters. Absolutely no evidence for this but it's more fun.

fenix

wongataa wrote:

cougie wrote:

If stages is pants- what's the deal with sky using them ? I can't imagine they're that short of money that they'd be happy with a guesstimate of power ?

Sky get to use Stages dual sided power meter that you can't buy.

Has that always been the case though ? I thought the double sided is a new thing ? Given that we know Froome rides to power for the critical parts of the Tour - I can't see that such an attention to detail team would be happy with a PM that was less than accurate ?

thomasmorris

meanredspider wrote:

I think the OPs question is interesting but entirely academic. Unless the OP and his dad are willing to do a load of tests (swapping meters, swapping bikes, swapping just tyres, swapping clothing, including helmets, riding different routes etc) then we won't know. Even then, if the meter contribution won't be clear as to which is contributing what. But, as I say, let's just blame the meters. Absolutely no evidence for this but it's more fun.

Sorry, I thought you were referring to a prejudice against stages in general, rather than just a prejudice against them in this thread. A lot of people wouldn't use a left sided power meter because of leg imbalances, and a lot of people wouldn't use stages in particular because of reliability issues. I don't think you can say either are prejudices as they're based on pretty reasonable beliefs.

My L-R balance is 44-56. So when I'm doing 200 watts a stages would read 176. That's 26 watts right there, before you consider the imbalance of the other rider which could add further error if also using a left side only meter. So why, when someone comes on a forum and is using a single sided power meter and is struggling to reconcile a power difference is it 'prejudice' to suggest the error is from the measurement? That's not prejudice, it's pretty reasonable.

The OP wouldn't have to do loads of test either. Put the bike on rollers and you'd know in ten minutes whether its a power meter / imbalance issue.

The discussion is hardly 'entirely academic', the OP has noticed a 20-40 watt difference. That's massive. The whole reason he and his dad bought these tools was to improve performance I presume, so why not use them to analyse and improve performance? They can very easily use the expensive tools they've bought to find those 24 watts then his dad will be going several miles per hour faster.

I think it's much more productive discussion to be having than "why don't people wave at me".

thomasmorris

Fenix wrote:

wongataa wrote:

cougie wrote:

If stages is pants- what's the deal with sky using them ? I can't imagine they're that short of money that they'd be happy with a guesstimate of power ?

Sky get to use Stages dual sided power meter that you can't buy.

Has that always been the case though ? I thought the double sided is a new thing ? Given that we know Froome rides to power for the critical parts of the Tour - I can't see that such an attention to detail team would be happy with a PM that was less than accurate ?

Pretty much right away Sky had a dual sided PM. http://www.bikeradar.com/road/news/article/stages-prototype-power-meter-spotted-video-44698/

For general training and pacing in events I don't think L-R balance would be huge issue. It would be relatively constant to the individual.

They do all of their aero testing in wind tunnels where power meter accuracy wouldn't be an issue.

It's interesting that Bradley Wiggins used oval chain rings when SRM and Stages over read with oval chain-rings. At one point most in the team were using oval rings. But once it became clearmost switched back to round (only Chris Froome still on oval?). An example of a power meter reading incorrectly leading to equipment choice mistakes?

okgo

I think I preferred it when only people with half a clue would buy powermeters. Now everyone has them and thinks they're Dr Ferrari.

Stages are total dogshit. Vectors hugely sensitive to installation torque.

meanredspider

thomasmorris wrote:

meanredspider wrote:

I think the OPs question is interesting but entirely academic. Unless the OP and his dad are willing to do a load of tests (swapping meters, swapping bikes, swapping just tyres, swapping clothing, including helmets, riding different routes etc) then we won't know. Even then, if the meter contribution won't be clear as to which is contributing what. But, as I say, let's just blame the meters. Absolutely no evidence for this but it's more fun.

Sorry, I thought you were referring to a prejudice against stages in general, rather than just a prejudice against them in this thread. A lot of people wouldn't use a left sided power meter because of leg imbalances, and a lot of people wouldn't use stages in particular because of reliability issues. I don't think you can say either are prejudices as they're based on pretty reasonable beliefs.

My L-R balance is 44-56. So when I'm doing 200 watts a stages would read 176. That's 26 watts right there, before you consider the imbalance of the other rider which could add further error if also using a left side only meter. So why, when someone comes on a forum and is using a single sided power meter and is struggling to reconcile a power difference is it 'prejudice' to suggest the error is from the measurement? That's not prejudice, it's pretty reasonable.

The OP wouldn't have to do loads of test either. Put the bike on rollers and you'd know in ten minutes whether its a power meter / imbalance issue.

The discussion is hardly 'entirely academic', the OP has noticed a 20-40 watt difference. That's massive. The whole reason he and his dad bought these tools was to improve performance I presume, so why not use them to analyse and improve performance? They can very easily use the expensive tools they've bought to find those 24 watts then his dad will be going several miles per hour faster.

I think it's much more productive discussion to be having than "why don't people wave at me".

Yeah but why does it matter that the meters read differently? What would he and his dad do differently if the two meters read the same? My guess: absolutely nothing (why, IMO, it's academic) - they aren't going to choose the OP's dad over the OP to represent his street in some team or other, for instance, based upon those numbers.

I'm ONLY interested in my meter: did I work harder today? could I sustain a power number for longer? do I need to ride at a lower power on this endurance ride? I don't care if the the units are Watts, HP, or Spiders. I don't care what your number is. It's just a number. Which is why, for me at least, the whole L-R balance thing is immaterial. Would I ride any differently? The answer is No. I am, after all, measuring my overall aerobic cardio vascular system and not solely my individual legs. All I want to know is that the number I get is repeatable and, based upon several years (3 now) of use, I have absolutely no evidence to say it isn't. I don't do a ride one day and get 250W and the next only 230W for the same ride. Changes, when they occur, are gradual and in line with expectation (working hard on fitness, the number trend upwards and vice versa). But this is old ground and been covered over and over.

meanredspider

okgo wrote:

I think I preferred it when only people with half a clue would buy powermeters. Now everyone has them and thinks they're Dr Ferrari.

Yeah - because the concept of power is really difficult to get your head around.... :roll:

thomasmorris

meanredspider wrote:

Yeah but why does it matter that the meters read differently? What would he and his dad do differently if the two meters read the same? My guess: absolutely nothing (why, IMO, it's academic) - they aren't going to choose the OP's dad over the OP to represent his street in some team or other, for instance, based upon those numbers.

Regardless of whether the power meters read the same or not, if you're going to make comparisons between the two it's worth checking you're comparing apple with apples. Chucking the bikes on the roller for 5 minutes would tell them whether leg imbalance or accuracy of their power meters is skewing that comparison.

The fact they do have different numbers, however, suggests that there could be significant gains to be made. So all the more reason to investigate it. Of course, they might just find out one of them has a leg imbalance, but at least they'd know.

The dad may find he needs to tuck a little lower, change his wheels, or helmet, or all of the above which may allow him to cycle 1-2 mph quicker. The extra speed may be useful in competition, it may just allow him to set some PBs on local segments. But given he's bought a power meter to improve his performance wouldn't that improvement be a good result ?

meanredspider wrote:

I'm ONLY interested in my meter: did I work harder today? could I sustain a power number for longer? do I need to ride at a lower power on this endurance ride? I don't care if the the units are Watts, HP, or Spiders. I don't care what your number is. It's just a number. Which is why, for me at least, the whole L-R balance thing is immaterial. Would I ride any differently? The answer is No. I am, after all, measuring my overall aerobic cardio vascular system and not solely my individual legs. All I want to know is that the number I get is repeatable and, based upon several years (3 now) of use, I have absolutely no evidence to say it isn't. I don't do a ride one day and get 250W and the next only 230W for the same ride. Changes, when they occur, are gradual and in line with expectation (working hard on fitness, the number trend upwards and vice versa). But this is old ground and been covered over and over.

Well, you might only be interested in your own power meter and whether you can work harder or sustain power for longer. However, if that's only what your using the power meter for then you're missing a big trick. Left only is okay for pacing yourself and setting training zones; however, it is fraught with difficulties for aero testing and making comparisons to other people. But when people suggest that the power meter could be the source of the difference you label it as prejudice; yet you then go on to admit you don't use your power meter for aero testing or comparison to other people.

An example; I was reasonably well trained when I got my power meter three years ago. Since then I've done about 1,500 hours of riding with power, of which about a third is structured training. This added at very best 20 watts in output, that equates to about 1 mph. Through 6-7 hours ours of aero testing (plus a few hours of analysis) over the last year I've added 2 mph.

meanredspider wrote:

okgo wrote:

I think I preferred it when only people with half a clue would buy powermeters. Now everyone has them and thinks they're Dr Ferrari.

Yeah - because the concept of power is really difficult to get your head around.... :roll:

The fact you don't seem to realise that power meters can be used to increase performance in other ways than just increasing fitness, and you question whether any discussion along those line is worthwhile seems to suggest you are struggling with the concept of power.

meanredspider

@thomasmorris (to avoid endless pointless quoting)

You're right, rollers would reduce some of the variability (if they have some - I doubt it) though might increase the impact of tyres and still wouldn't tell you which meter was more accurate than the other. And, should the numbers be closer together, you're still going to have to measure the effects of all of the other variables that doubtless have an impact. They could spend a grand on a Neo and remove the tyre variability...

You'll have to explain to me why left-only isn't OK for aero testing if its precision is OK. Provided the measurement is repeatable, you can assess aero changes. The absolute numbers are meaningless.

I don't know where you got the idea that I don't understand that performance other than fitness comes from power testing (I listed a whole load in my original post...). Again, everything can come from your own number and that comparisons with others are full of potential confounding issues. I'm not interested in that though. And, in any case, aero testing is best done in the wind tunnel.

thomasmorris

@meanredspider

Rollers or a mag trainer would completely remove the aerodynamic drag of the rider; which is the largest variable. Any differences left at the same speed would then be drive-train efficiency and rolling resistance. Friction facts website tells you that the variance from the most efficient drive train to a very average one is 7 watts. Rolling resistance between an excellent tyre and an average tyre would be in the range of 10 watts. The rest of the resistance is the trainer, which is a constant.

Of course they could just buy a Taxc Neo for a grand, but if they hopped on each others bikes that would remove the variable of tyres, power meter and drivetrain to and leave only leg imbalance.

It would take 5 minutes to do.

Left is not okay for aero testing. I find in the more aero positions (tight hip angles) my balance skews further. It would be hard to differentiate an aero gain against an change in L-R balance. Yes, those 1-2% might not matter for pacing on a climb, or measuring a fitness gain. But it sure matters for aero testing. That's why lots of people wouldn't use a single sided power meter.

Yes, everything can come from your own number. But equally, if someone rocks up to a time trial and goes faster than me with less power, it's useful to look at why. It would take a long time of trial and error aero testing on your own to find out. Looking at difference between yourselves and other could lead to quick gains (which obviously you'll have to go and test yourself, but it could send you in the right direction).

Aero testing is not best in a wind tunnel either. It's more repeatable in a wind tunnel with less variable. But it's also hugely expensive and still needs to be validated in the field. You can achieve the large majority of what is possible in a wind tunnel with a good power meter, a calm day and quiet road. The power meter is 500-700 quid and once purchased you can collect as much data as you can find time to test. A wind tunnel costs each and every time and there is limited availability.

meanredspider

Still assumes they have rollers and can ride on them consistently..

If 1-2% matter, that's the accuracy of most meters anyway (and precision is rarely quoted which kinda makes a nonsense of it anyway).

What you say about aero testing flies in the face of the received wisdom on here and elsewhere.

And, when comparing yourself with a guy that rocks up and outpaces you at a TT on less power, how do you know his number is accurate? If that's a common approach, I'd offset the calibration of my meter to confuse anyway. It's a common trick in the car racing I do to do things to confuse the opposition over your pace.

thomasmorris

meanredspider wrote:

Still assumes they have rollers and can ride on them consistently..

If 1-2% matter, that's the accuracy of most meters anyway (and precision is rarely quoted which kinda makes a nonsense of it anyway).

What you say about aero testing flies in the face of the received wisdom on here and elsewhere.

And, when comparing yourself with a guy that rocks up and outpaces you at a TT on less power, how do you know his number is accurate? If that's a common approach, I'd offset the calibration of my meter to confuse anyway. It's a common trick in the car racing I do to do things to confuse the opposition over your pace.

You're right, the answer should be, ignore the power numbers and talk about the scenery.

Most people have rollers or a trainer. He works in a bike shop. It's a suggestion of an easy way to see the difference, I think you're splitting hairs now to save face!

Most people can ride steady on trainer. But if that was really an issue you could just do a power profile test for the rollers. The point is, they'd be able to see instantly whether the 20-40 watts was real difference in aero, drive train and tyres or just power meter / imbalance. That would answer the OPs question straight off.

What have I said about aero testing that flies in the face of 'received wisdom' specifically? I'm pretty sure it doesn't.

As for 1-2% the answer lies in the type of error.
1) A power meter may over read always by 1-2% due to calibration. This is a sytemic error, but as you're only interested in relative improvements it's fine.
2) Or it may randomly fluctuate by 1-2%. This is more of an issue, but can be accounted for by doing more repeats to increase statistical confidence.
3) If however, the error is in built and correlated to the thing you testing (position), then this is an issue. Like the example I gave above, my imbalance increases with my hip angle, which is related to position which i am testing. It then becomes impossible to tell apart whether I am getting more aero as I'm increasing my hip angle, or whether by left right balance is just changing.

It is a common approach in time trialing to look at other peoples numbers and time and make a judgement about how aero they are. You can pretty quickly see if someone is slippery or just very powerful. If they are slippery you may look at their kit or position, evaluate what the differences are to you, and it gives a few new areas to investigate. Some things may work, some might not. But it gives you avenues to explore which you might not just doing testing on your own.

As for their power meter being off, I tend to look at what power meter their using and discount left sided only.

meanredspider

He works in a bike shop and is posting this question on here about a 20W difference in power reading without giving relevant info and without figuring it out for himself...

As for "saving face", firstly I'd have to care far more than I do about what people on the Internet think of me and secondly your roller test still presents the question of which meter is contributing what to the offset (if anything).

Yup - plenty of sources of error both related to the measurement system and the person doing the measurement. I'm interested to know how you can hope to measure something to an accuracy of 1-2% when the device itself is only rated to 1-2% - unless you've had your specific device tested against a known standard and, even then. You've talked about (without using the terms) accuracy and precision. I've not even seen precision quoted for any power meter (yet it's by far the most important element for your aero testing). I used to work as quality director for an in vitro testing company. I understand measuring and error - which is why I find the whole meter thing both amusing and tedious in equal measure.

I'll let all the guys who've said aero testing is best done in a wind tunnel defend their position - I can see why they'd say that on the basis that you're looking for marginal gains that are likely to be swamped by external noise. But that's why I used the term "received wisdom". And I think it's way outside the scope of the OPs question.

If that's the way it's done and it's even remotely competitive TT'ing you're doing, I'm surprised. Why would you want to give away your competitive edge? I'd throw a known offset into my calibration to keep my opposition guessing.

ETA - in fact the funny thing is that you're arguing (seemingly) that the biggest difference is from the meters yet, you yourself, look for other differences between riders and assume the meters are accurate.

thomasmorris

For the final time, it's the type of error! 1-2% if that is a random precision error then by doing more duplicates you can still find a differences smaller than 1%. If that error is accuracy error, but the PM is precise then you could find still smaller differences. From your background you should understand this.

But the OPs difference isn't 1-2%, he doesn't give the averages but it's more likely 10% difference. So easily distinguishable with devices that have an error of 1-2% even with very few tests.

However, if you have an error which correlates with the variable you're measuring (i.e. left right balance varying with position) then you can never, no matter how many replicates you run, resolve differences. This is the problem with left only power meters.

I'm not arguing that wind tunnel testing is not more accurate / quicker or anything like that. I am saying it is not best for everyone as it is expensive and has availability issues. A power meter you can do a lot of the same things at a fraction of the price and in your own time.

As for looking at what other people are doing, I'm saying as a guide it's useful to have a comparable number.

meanredspider

thomasmorris wrote:

For the final time, it's the type of error! 1-2% if that is a random precision error then by doing more duplicates you can still find a differences smaller than 1%. If that error is accuracy error, but the PM is precise then you could find still smaller differences. From your background you should understand this.

But we don't KNOW the precision of any of the meters as far as I know (I've written it twice already) - only the accuracy.

I'm not suggesting that this isn't good for the OP just responding to the point that you are looking for 1-2% differences in your aero testing using a device that has an accuracy of 1-2% and an unknown precision.

The OP has a huge difference in AVERAGE power and, as I've said before in this thread, DC Rainmaker often makes the point that even the crappiest PMs make a pretty good stab at average power - and he compares a lot of meters. So, there is either a gross issue with a meter or they are broadly accurate and the issue is coming from some other source (or somewhere between). Most other contributions can be identified by inspection: weight, clothing, tyres, riding position - if not entirely quantified.

thomasmorris

Yes, I don't know the precision on my power meter. However, the error is random. So I don't need to know the precision. I just test position A against position B enough times that I get a statistical difference in the mean CdA for each position.

The error of a left sided meter is not random. It correlates with what your testing position. So no matter how many test runs you do you will not be able to resolve a difference.

The op has a huge difference in power to someone of similar size and build. He asks if this is usual and what it could be because his dad would like the extra speed.

I maintain that the difference could be in power meter accuracy (leg imbalance), aerodynamic drag, rolling resistance and drive train efficency.

Considering they're both using left only meters and the kit they're using is similar it's no wonder people have pointed to the power meter.

The OP could do a simple test on the trainer / rollers to work this out.

Instead, anyone who suggests the error is in the power meter you are labeling as prejudice and your saying trying to work out any other difference is impossible.

The problem isn't other people's prejudice, it's your ignorance of what can be achieved with a power meter. But that's not surprising as you said yourself you only use your power meter for pacing and measuring fitness.

meanredspider

I think you're kidding yourself with your testing and just measuring noise and attributing meaning to it. But don't let me stop you.

We didn't know most of what kit they were using before people jumped on the left-only meters. That's pretty much the definition of prejudice. I said I expected the meters to be contributing something as any two different measuring systems will - especially where humans are involved. I just think that it's funny. Ultimately I don't care.

thomasmorris

Fine, I'm kidding myself with my testing, but I've gained 1:40 on 20 minute course in a year with no gain in power.
There are lots and lots of people doing the same.
Left sided power meters are known to have the issue with imbalance. So if you have two similar sized people the first thing to check is for that error. It's just common sense. If you rule that out you can look at other stuff and start finding performance differences.
That's not prejudice it's based on sound reasoning.
You're just ignorant to the issues with single sided meters and the possibilities of using devices in this way.

thomasmorris

Just a word on error:
My Power2Max is accurate to 1% and precise to 2%.

The P2M records a reading every second. So even a short 5 minute test has 300 samples over which the precision error averages out (hence the accuracy of 1%).

So to conclude, using a spider / wheel / dual sided power meter is actually one of the smallest contributors to uncertainty in my CdA calculations.

The biggest uncertainty comes after the power measurements are taken in calculating CdA, these errors come from holding a constant position, getting accurate environmental measurements, gusts of wind, knowing accurately Crr, knowing drive train efficiency. I know from the variance in my results measuring a single position that these other errors of 3-4%. However, these errors average out (random error) so you can still have statistical confidence but you just have to run duplicate / repeat tests until you have confidence that the mean CdA for position A is different from the mean CdA for position B

If you ask any experienced person about aerodynamics it's the above factors which make outdoor aero testing difficult and mean if you want to resolve very small differences (1-2%) you need to have very accurate local weather readings. You may have noticed several bike companies are trying to address this with on-board sensors.

So you're completely wrong that it is the unknown precision of power meters that makes my aero testing silly.

Single sided power meters:
A single sided power meter may have an accuracy of 1-2% and an unknown precision but lets say it's similar to P2M at 2%. However, you also have the built in systemic error of the measurement system, which is taking left leg power and doubling it. Many people have a leg imbalance of 1-3%, and people like me have an 6% imbalance. This error is then magnified when the power meter estimates total power and will become an accuracy error of 12%.

Then the worst bit is the 12% error is not random. It can be related to the variable you are trying to test. I do two loops, position A is 208W for 30kmh. Position B is 192 W for 30kmh. It is impossible to tell whether position A is less aero, or whether my leg balance has just changed from 52-48 to 48-52. And no matter how many tests are run it is not possible to average this error out as it is correlated with the variable you are testing.

meanredspider

thomasmorris wrote:

Yes, I don't know the precision on my power meter.

thomasmorris wrote:

So you're completely wrong that it is the unknown precision of power meters that makes my aero testing silly.

Well, I'm glad you got that off your chest. You've now found out the precision of your meter - that is at least a start.

BTW - have you calculated the number of repeats you need to measure based upon the accuracy of your measurement system and the difference you're trying observe? I absolutely see why committed testers buy wind tunnel time.

And your thinking on this stuff is a mess. Precision is pretty much everything for the "aero" meter - you need to know that your measurements are repeatable and that you have confidence that the difference you observe is real. All things being equal, leg left measurement is absolutely fine and L-R balance is immaterial. But, you're correct, things aren't equal if you're constantly changing position. You have what seems to be quite an extreme imbalance- you don't share any data on how that imbalance changes during your testing - that would be interesting in this context. But still, absolutely accurate power numbers for the testing don't matter. You want absolutely precise numbers so you can be sure the difference you see is real. It's interesting that 4iiii have called their meter "Precision" - I've not been able to find precision numbers on their site nor Rotor's.

What I also find strange is that you suggest that there's an balance change but you think that a 5 minute test is sufficient - yet you have no idea about your ability to sustain that level of power if, for instance, the leg you normal generate less power with now has a greater demand. It all comes across as deeply flawed.

BTW - I've never suggested a left-only meter is good for aero testing and I'm on record as saying I'm sure they're poor for track cycling too. What I have said (and I'm still a deep sceptic) is that I don't think your method of aero testing is going to reliably find differences of 1%. But please feel free to carry on - you obviously have the time to expend. This really doesn't interest me at all and there's no suggestion that it does the OP either.

thomasmorris

Precision in the power meter is not absolutely everything for aero testing. I've explained why, as the error is random, so you can still be confident in your average power for each test run (within the 1% accuracy stated). Each short test contain hundred of measurements with the power meter where the precision averages out. You are not using individual measurements, but the average of many. Do you not understand this? Any precision error in individual measurement precision error will average out as it is random error. You are then left with an average which is accurate to 1%.

This is why precision is probably not often quoted for power meters as very rarely do people look at individual power measurements.

Confidence in environmental conditions is of far greater concern for outdoors aero testing, as ultimately you are calculating CdA and these conditions have a big impact on speed.

If you want to pick holes in my aero testing, you should start with the environmental factors, not the power meter I'm using. Anyone with any knowledge of aero testing knows this. It's these errors which require me to do multiple runs to have confidence in my results. If I was working in a wind tunnel or indoor velodrome I could do less duplicates or resolve smaller differences as I would have more control over environmental variables. But it would cost much more money.

For what it's worth, Aerocoach, one of the industry leaders in aero testing do outdoors testing. They work with professional athletes and many of the UKs top amateurs.

My L-R balance varies a lot. Both with position and effort. Typically it's 46-54 on a average club run. On an easy ride on the road bike it is more 44-56. On the TT bike those number change to 45-55 and 43-57. However, they can vary by 1-2% day to day for no discernible reason. The vast majority of people, have an imbalance. It would be unusual to have 50-50 balance for every single ride and effort. However, if your using a left side only meter you will never know this.

As for the impact on sustained power in the position I do test this but separately. I do an all out effort on a known circuit to see how my ability to put out power is affected by the new position. Interestingly, this is another area single sided meters fall down, as you wouldn't know whether a new position is making you put out less total power, or is just changing balance balance, therefore inaccurately reporting you total power.

I agree, none of the above is directly relevant to the OP, but it might be he he wants to find where those 20-40 watts difference are.

Basically, what you objected to was people saying the 20-40 watts was probably coming from the power meter measurement and you labelled that as prejudice. I just pointed out that it's not, and it's well reasoned.

You then went on to infer that it was futile trying to find out where the differences were using power meter data, which I disagree with as well, as I know from my own personal experience that it is possible with quite remarkable results.

The long and short of it is: people aren't prejudice against left sided power meters, you're just ignorant as to how the limitations of left sided meters impact what can be achieved with powermeters. This is largely because you are ignorant about aero testing with a power meter and only use your's to measure your effort for pacing and training.

I gave the OP relevant good advice which could quickly eliminate the L-R imbalance from the list of factors causing the difference. Put the bike on a trainer or rollers, swap the riders over and look at the speed to power relationship. If at that point they don't find any difference they can start looking at performance factors.