Wattbike pedal shapes - real life examples

bahzob

I've mentioned the Wattbike in several posts as helping develop pedal technique for those interested in doing this. The reason it so useful, is that it provides real-time feedback in a straight forward visual format, in what it calls its “Polar View “ showing you exactly how the way you are turning the pedals results in producing power. Currently this isn't possible with power meters fitted to road bikes. (However this is changing, some new power meters are already collecting the data needed for this feature e.g.
http://www.factorbikes.com/files/6192793f-507e-4a67-89c6-a243011f2875/DOWNLOAD%20PDF.pdf
and I'd expect it to be included in future head units and versions of products like WKO.)

A picture can be worth a thousand words and the Polar View can be used to experiment seeing how modifying the way you pedal affects your capacity to generate and sustain effort. Basically it allows you to train using biofeedback which is a long established means of developing skill and far more effective than trying to follow written instructions or being told by someone else what to do.

So if possible I'd suggest finding a Wattbike near you and giving one a try. That said I appreciate not everyone has access to a Wattbike so I have put a small video up here showing how the Polar View changes with different pedal styles
https://www.youtube.com/watch?v=plF4kp5LUOU&feature=youtu.be

Even if you don't use a Wattbike this should be helpful as it shows the clear cut difference between pedal styles.

A full description of the Polar View can be found here
http://wattbike.com/uk/guide/cycling_tests/pedalling_technique_test/what_the_polar_view_shapes_mean

Briefly though:
>> The quadrant from 9:00>6:00 is the drive phase for the left leg, the quadrant from 03:00>12:00 drive phase for the left. The other two quadrants show the transition between drive phases.
>> The distance red line:centre shows the amount of force applied to the cranks through 360degrees as they go through a rotation. As each rev is completed the picture is refreshed which gives the impression of a movie.
>>The area within the red line represents the total force of each revolution. Power is force*rev. (i.e. there are two ways to increase power, press harder to make the force area bigger or spin faster keeping force area constant)

The video shows the following types of pedalling done at a reasonable power level (c. 300W). It's pretty obvious which is which hopefully.
“Circle”- This is “pedalling in circles” or “souplesse” where force is smoothly distributed and changes during transition between legs are small
“Peanut” - This is "pedalling in squares"/mashing the pedals where force is concentrated at 2 points of each rev,in a big peak one for each leg with large transition changes
“Standing” - Even more accentuated differences in force for illustration (if actually standing for long periods during a climb I'd be aiming to smooth this curve out)

Following these, for illustration resistance is decreased and there are some one-legged drills (not that I recommend them btw, time is much better spent working on technique while riding normally).

Finally there is a short section upping power a bit to the higher 300s.

Hopefully some will find this helpful in clearly illustrating the differences in pedal styless.

As for these differences:
>> The “circle” and “peanut” shapes feel quite different. This is reflected in the way the various muscles operate. Specifically, the peanut tends to emphasise the quads. I found that to produce a circle I had to make more of a conscious effort to use of other muscles e.g. glutes and calves and to time their action more precisely so their effect was co-ordinated with that of the quads. Also I had to to be more focussed in ensuring the recovering leg was turning over smoothly and not providing resistance to the drive leg. (Others may find different ofc illustrating the benefit of actually riding a Wattbike).

>> Circles are easy to produce at low resistance but a lot harder, until fully trained, at high power.

>> Top cyclists will typically have the “circle” shape, while poorer cyclists tend to pedal in “peanuts”.

>> Given this I can't help but point out that imo it would be strange if both circle and peanut were identical in all respects in terms of efficiency and sustaining power.

>> Just riding a bike will not necessarily result in changes to pedal style at least in my experience. I had ridden 10s of thousands of miles when I first tried a Wattbike out and assumed I would pedal in nice circles just like a pro. I didn't and it wouldn't surprise me if most other cyclists were similar. It took time and focussed attention, aided by the Polar View, to adapt before I was comfortable. However the result was not just being able to produce circles but to do so at higher power for longer than my old style. Which is sort of the point.

(Just for the record, I have absolutely no connection with Wattbike. I am enthusiastic about the product because of my experience with it. Using one has made me a better rider which is why I'd encourage others to give it a go.)

Tom Dean

bahzob wrote:

>> Top cyclists will typically have the “circle” shape, while poorer cyclists tend to pedal in “peanuts”.

Evidence please.

imposter2.0

Tom Dean wrote:

bahzob wrote:

>> Top cyclists will typically have the “circle” shape, while poorer cyclists tend to pedal in “peanuts”.

Evidence please.

Surely anecdote is enough for you..?

charlie_potatoes

This topic is just going round in circles.
It is driving me peanuts

SwainDogg

My stroke is more of a cashew nut ...

JayKosta

When trying to pedal in 'circles', do you have any particular 'feeling' about how your muscles in the middle area on back of thigh are tensing and relaxing?
Do you have a feeling that any of the leg muscles have beneficial periods of relaxation?

Jay Kosta
Endwell NY USA

amaferanga

Yawn.

Stalin

I use a Wattbike for much of my training. I don't use the shape graphs much, but when I do the shape for my right leg is more round than my left leg. But paradoxically the left leg is the more powerful leg. This is probably due to some serious injuries many years ago to the right knee and hip. I conclude that the right leg does not apply as much force on the downstroke as the left.

My left right power balance is consistently 53/47 ish at lower powers. At near maximum 20 minute power it balances out though. I assume the right leg does its fair share of the work when needed. I talk as if the legs have a mind of their own, but obviously what is happening is the brain in monitoring the legs, getting feedback and firing muscle fibres and distributing the work in such a way as to protect the long term injuries.

I find the power read out on the Wattbike fluctuates all over the place. This is so much a criticism of the bike, more a reflection on my own rather uneven power output.

I may well pay more attention to the polar graphs in future.

xavierdisley

As I've mentioned before, as the Wattbike only has two location magnets it only knows where the cranks are twice out of 360 degrees. The rest is a guess and influenced by changes in pedal velocity which isn't measured, much like a Computrainer SpinScan.

Xav

Stalin

https://wattbike.com/uk/guide/using_the ... sing_watts


The amount of power produced is measured 100 times per second by a load cell, located on the chain, and is a measurement of the sum of all forces applied to the chain through the cranks. The Wattbike’s Performance Computer then displays the power output every 3 seconds, if you are connected to the Wattbike Expert Software every pedal revolution can be seen in real time with data sampling 100 times per second.

This high rate of data sampling means that the force exerted on the cranks can be recorded every 10 milliseconds (1/100th of a second), allowing the Wattbike’s unique Polar View to display how force is being applied at all points in every pedal revolution. This allows the Wattbike to show your pedalling technique as well as your power output.

Accurately measuring power output is the basis for all of the other parameters recorded on the Wattbike, and means that for the first time it is possible to race between indoor bikes and produce accurate and repeatable results. This is true of every session completed on the Wattbike, no matter when or where the session is done; the results will be comparable with all other Wattbike results. This is made possible because the Wattbike is factory calibrated, and is accurate throughout its life.

Alex_Simmons/RST

High frequency torque sampling + low frequency rotational speed sampling = low frequency power sampling.

It's the same deal for the SRM Torque Analysis system (200Hz) and Computrainer's SpinScan (I forget the torque sampling frequency but it's enough to give readings split into every 15 degrees), and the somewhat lower resolution left-right side power balance data provided by Quarq and Power2Max power meters**.

However sampling frequency is not so much the issue.

The primary issue is this is a downstream measurement of the net forces from both legs combined.

That alone means the nature of charts produced will be different to what you would see from inspecting independent crank or pedal force vectors (which is what you actually need to see before making any decisions about interventions wrt pedalling).

This is not a matter of level or precision of data, as has been previously suggested, but rather it's showing you something quite different and can readily be misinterpreted (as is clear from threads such as this one).

e.g. what might be reported as say a dominant right side, may not in fact be the case. You may very well be producing more power with your left leg but even though the net forces suggest you have a dominant right side. Alternatively you may very well be right side dominant. Problem is, with such downstream net force data you just can't say.

So if you see on such a chart that your left side is less powerful, and decided to act on that in some way through training intervention and deliberately work harder with that leg (say), it just might be that what you end up doing is the opposite of what you actually should do.


As for how the elite/pros pedal, well I suggest people inspect the peer reviewed published data actually collected from such riders using independent force vector measurement methods with high frequency torque and rotational speed sampling. Then you will see that, in general*, more powerful elite riders push down harder relative to their average torque per revolution than do lower level/less powerful riders (and their true polar charts are less circular than their less powerful lower level buddies).

None of which of course infers anything about how the rest of us should attempt to pedal.

I'm with amaferanga. Yawn (and I'll leave it there).


* of course every rider has their own unique pattern.

** There are several newer cranks emerging that are beginning to provide independent force vector measurement, however the software to display/record such data is not consumer friendly or compatible with data transmissions protocols in common use (e.g. ANT+).

Stalin

My left right imbalance shows the same using Look Keo Power Pedals and Garmin Vector.

As I may have said before, on a Wattbike my weaker right leg shows a much rounder graph than the left leg. I can only assume that the left more powerful leg produces more force on the downstroke.

I got the same numbers and power splits using cleats and ordinary shoes, even ordinary shoes on both Keo and Vector.

I'm not sold on pedalling technique. I am sold on pressing as hard and fast on the pedals as possible. Pressing harder seems to increase cadence and increasing cadence tends to cause pressing harder! Funny that.



I

imposter2.0

Stalin wrote:

I'm not sold on pedalling technique. I am sold on pressing as hard and fast on the pedals as possible. Pressing harder seems to increase cadence and increasing cadence tends to cause pressing harder! Funny that.

So pushing harder and more often on the pedals makes you go faster? Hopefully this is something we can all agree on...

Stalin

Imposter wrote:

Stalin wrote:

I'm not sold on pedalling technique. I am sold on pressing as hard and fast on the pedals as possible. Pressing harder seems to increase cadence and increasing cadence tends to cause pressing harder! Funny that.

So pushing harder and more often on the pedals makes you go faster? Hopefully this is something we can all agree on...

But does doing that smoother with better technique make you go even faster or as fast for longer?

Stagbeetle juice anyone? Works even better than Japanese Hornet juice.

dw300

I was always told peanuts were full of energy. :?

bahzob

Fairly predictable views, really if this subject bores you the subject and author give a clue as to what is in the post, just ignore it.

On the other hand if the subject does interest you the bottom line is the answer to this question posed above

"But does doing that smoother with better technique make you go even faster or as fast for longer?"

In my experience the answer to this is yes. What's frustrating is spending years training before discovering this.

Which is why I'd suggest others do as I have (and many above have not I guess), open your minds and actually try using a Wattbike.

Tom Dean

I find the subject interesting.

What is boring is your repeated assertions, without evidence, and contrary to peer-reviewed data, that smoother=better. I am open to any EVIDENCE you might have to support your view (advertisements for Wattbike don't count).

ncr

bahzob wrote:

On the other hand if the subject does interest you the bottom line is the answer to this question posed above

"But does doing that smoother with better technique make you go even faster or as fast for longer?"

In my experience the answer to this is yes. What's frustrating is spending years training before discovering this.

What do you mean by smoother technique ? Describe what you are attempting to do at 3, 6, 9 and 12 o'c. How does it make you go faster.

BenderRodriguez

Tom Dean wrote:

[you assert] without evidence, and contrary to peer-reviewed data, that smoother=better. I am open to any EVIDENCE you might have to support your view.

Much of the 'peer-reviewed data' relating to this issue is poorly designed. For example, expecting differences in efficiency to be apparent by simply telling people to alter their pedalling style in the lab without there being a period of adaption.

I would say that one aspect of pedalling in 'smoother' way would be to apply torque to the cranks when pushing down in a more even manner, but with a lower peak torque value. There is evidence that pedalling in such a 'smoother' manner is associated with a higher level of efficiency, at least close to threshold. This is probably because it minimises the number of easily fatigued 'fast twitch' fibres that are recruited. This is, those fibres that would be recruited if the rider used a 'punchier', less smooth style with a higher peak torque value and a less even distribution of torque.

Influence of pedaling technique on metabolic efficiency in elite cyclists.

DOI: 10.5604/20831862.1003448

Biol. Sport 2012;29:229-233

Our objective was to investigate the influence of pedaling technique on gross efficiency (GE) at various exercise intensities in twelve elite cyclists… at the LT, there was a significant correlation between GE and mean torque and evenness of torque distribution (r=0.65 and r=0.66, respectively; p < 0.05).

http://biolsport.com/fulltxt.php?ICID=1003448‎

bahzob

Just to deal with Alex's point.

"The primary issue is this is a downstream measurement of the net forces from both legs combined."

This both misleading and based on a very naive understanding of how scientific measurement works. It basically asserts that if you cannot measure something directly and perfectly then you cannot measure it all.

This is ofc rubbish. Every scientific measurement has to factor in a degree of variance that will be caused by limitations on the accuracy of the measuring tool and noise from unrelated sources to that being tested.

Many experiments have, of necessity, deal with this. BICEP2 was just a recent example. It did not actually "see" gravity waves or post big bang inflation. It merely detected some, extremely downstream, patterns in the CMB. Alex I guess will be writing to the Nobel prize committee expressing his concerns.

What is important is that a tool is capable of producing a reliable significant discernible signal above background noise that is consistent with what is being measured.

The Wattbike does this. Yes there is some noise generated by the opposing leg. However, contrary to the misleading/erroneous opinion of Alex this is at a low and inconspicuous level and does not detract significantly from the resultant functionality.

What happens when you actually use a Wattbike is that small variations in terms of pedal stroke reliably, consistently and instantly appear on the display. This is sufficient to help act as a guide to understanding and improving the way you pedal>more sustainable power.

As I said I know there are new meters that will do this more accurately. However I am quite happy to stand by the prediction that they will do so in a manner that will correlate to the Wattbike.

bahzob

Tom Dean wrote:

I find the subject interesting.

What is boring is your repeated assertions, without evidence, and contrary to peer-reviewed data, that smoother=better. I am open to any EVIDENCE you might have to support your view (advertisements for Wattbike don't count).

I get bored as well, especially from those with entrenched opinions on this subject who have never seen a Wattbike or made any effort to improve but still seek to advise those with a more positive approach. As repeatedly said it is pretty much the same boring "all you need to do is just ride your bike and you will get better" advice that various dinosaurs kept spouting to the minority of us training with power meters a few years back.

If you really find the subject interesting go find a Wattbike. Do a FTP test on one and post the results here along with the polar view. If it looks like a peanut then I am reasonably confident you could improve by making it smoother and would be happy to off advice on how to do this if you are willing to listen.

(Wattbike will be able to let you know where the nearest one you can try is, there should be one nearby you.)

imposter2.0

bahzob wrote:

If it looks like a peanut then I am reasonably confident you could improve by making it smoother and would be happy to off advice on how to do this if you are willing to listen.

I suspect we are all willing to listen - - HOW..??

xavierdisley

Bahzob - pedal velocity is not measured and pedal location only known twice per pedal stroke. That's a fact and has large scale implication for the reliability of the polar plot (in that it's estimating pedal velocity 99.4% of the time, which is not an accurate way of going about things). Not differentiating from left/right pedal force in order to create the graph is something that should bother you as well! And if it doesn't then there's not much anyone can do to convince you otherwise. The only way to get real data is independent left/right sensors - we constructed three axis force pedals in our lab to explore this kind of thing, and these two reasons (non measurement of pedal velocity and combined force data for both cranks) render their polar plot (and the Spinscan) unusable for pedalling analysis. You must have high speed video, infra red cameras or some other timing system to measure pedal velocity, you absolutely cannot estimate it in the way presented here.

The Q Factor is also monstrously high, which will affect the way you pedal. I've tested this in the laboratory too

Xav

Alex_Simmons/RST

bahzob wrote:

Just to deal with Alex's point.

"The primary issue is this is a downstream measurement of the net forces from both legs combined."

This both misleading and based on a very naive understanding of how scientific measurement works. It basically asserts that if you cannot measure something directly and perfectly then you cannot measure it all.

This is ofc rubbish. Every scientific measurement has to factor in a degree of variance that will be caused by limitations on the accuracy of the measuring tool and noise from unrelated sources to that being tested.

I said I'd leave it, but feel it might help others to respond on this.

Firstly, it's not an accuracy issue. The Wattbike may can be as accurate as you like in its downstream force measurements but it's still not measuring what's actually needed for the purpose of examining pedalling for the intent of recommending a training intervention. The same issue applies to an SRM Science power meter.

As for "noise" of the opposing leg in the signal of downstream measurement versus upstream measurement - well it's not so much that it's noise, as these are measuring something quite different.

The question then is whether one can reliably infer things from downstream torque data (e.g. Wattbike, SRM) that leads to sound advice about training intervention wrt pedalling.

Well here's an example of why this matters at even a low level of angular resolution*.

Compare the left and right side power split when measured using a downstream method (Quarq - which samples downstream data just like a Wattbike does) and an upstream method (Garmin Vector which is measuring forces at each pedal independently):

https://www.youtube.com/watch?v=k0i_jV9ygLI

So even splitting the torque data into just two segments we are getting significantly different results from two different measurements techniques (upstream and downstream) - and importantly these differences are not consistent (put aside the single leg stuff). And the differences are so significant such that one might easily interpret the data quite differently and what training intervention one might consider. This is not noise - it's actually a different measurement.

Xav has made the other points which affect how such data appears, let alone how we might interpret it.


* while the resolution of angular segments is low (i.e. just two segments, left and right), the torque data is sampled at a high frequency

Stalin

Alex,

Thanks for posting that video. Some observations .

If you look at the total power output you will see that at the start of the video the Vectors are consistently showing more power ( 7 to 10 watts) than the Quarq. At the end of the video the Quarq is measuring consistently a few watts more than the Vectors. The change is at about 1 min 10 seconds into the video.


Now this is the sort of thing that causes me to be sometimes wary of power meter data. Having seen that video I am not impressed by the accuracy or consistency of either the Vectors or the Quarq. One power meter is obviously inconsistent and inaccurate but which one? Or are both inaccurate and inconsistent.

It looks as if the swing from the Vectors measuring higher power than the Quarq to less power is when a foot is clipped back into a pedal. Might this point to the Vector giving different readings and the Quarq being more stable? What I don't like is the way there was drift, if one were always higher than the other all the time at least you would have consistency.p


The video does show how the two systems provide a very different example of left right distribution. Perhaps it is the way I pedal but I found my left right distribution very similar on Look Keo Power, Vector and Wattbike. However, I did not test either pedal system on a Wattbike.