Login required to started new threads

Login required to post replies

Proof bike fitting works???
Quote | Reply
You are professionals asking your customer to pay several hundred dollars claiming you can make then faster (among other things).
Have you ever proven your customer or yourself that your approach does make the rider more efficient?

I do not question the general use of bike fitting but i believe there are few scientifically proven concepts and there is also a lot prescribing snake oil going on in bike fitting. We want to improve and I think a good first step is questioning our knowledge. I am therefore currently working on testing protocols to validate the concept and I am looking for inspiration. So my questions:
Have you tested your concepts?
Which protocols did you use?
What were your results?
How do you judge the uncertainty of your test protocol?
Which ideas do you have on improving the test protocol?

I know I phrased the beginning of this post a little offensive but I did this in the hope to spark some interest in the conversation. I think Dan created something truly great with this subforum and I am a little sad about the little participation. So let's get this place to life :)

Ingmar
Quote Reply
Re: Proof bike fitting works??? [IJ] [ In reply to ]
Quote | Reply
One of the truly difficult parts of doing this sort of research is he number of samples (clients) you need to be able to get "Statistically Significant" results. The smaller the effect the larger the sample size required. The second issue is to do multivariate analysis gets very challenging to do in this sort of arena. So from a scientific peer review point of view the cost of the program would be quite high. So I agree it would be great to have lots of research on methods, to date most research goes more towards one specific cause-effect system. Further confounding the problem is the training effect and accommodation/ learning effect. Some people may adapt faster than others. This means seeing longitudinally if the effect is solid or is it a short term effect. All of this leads to research problems. So there are all sorts of problems or processes that should be looked at but the actual doing of solid research is very challenging, both mathematically and in resources. I would love to see more research though.
Quote Reply
Re: Proof bike fitting works??? [IJ] [ In reply to ]
Quote | Reply
I'll throw the first wrench into this. The result of a proper fit is not simply "make them faster." Injury prevention or management, comfort, longevity in the sport, the ability to stay in the aero position longer, better run off the bike, making the rider ITU compliant, "something ain't right," and even the occasional medial reason--I've had quite a few recommendations by MDs and PTs.
To take just the "make them faster" fits would require years of data to be statistically significant. I use the referral method as proof. If my fit clients are happy and tell their friends, we're doing it right.


Brian Grasky
Grasky Endurance: World Championship Triathlon Coaching; Professional Training Camps
RETUL fitter, Biomechanist, USAT Level 3 Coach, USAC Level 2 Coach
Quote Reply
Re: Proof bike fitting works??? [seebritri] [ In reply to ]
Quote | Reply
seebritri wrote:
I'll throw the first wrench into this. The result of a proper fit is not simply "make them faster." Injury prevention or management, comfort, longevity in the sport, the ability to stay in the aero position longer, better run off the bike, making the rider ITU compliant, "something ain't right," and even the occasional medial reason--I've had quite a few recommendations by MDs and PTs.
To take just the "make them faster" fits would require years of data to be statistically significant. I use the referral method as proof. If my fit clients are happy and tell their friends, we're doing it right.

+1 agree 100% same here.
Quote Reply
Re: Proof bike fitting works??? [s5100e] [ In reply to ]
Quote | Reply
Sorry for the delayed response, had a medical situation in the family...

So when we assume that we need a multivariate analysis, wouldnt it make sense then to also have a multivariate fitting approach? To my knowledge, all the major fitting concepts treat the goals as independent variables.
Also I think "proving" it works might be a little bit much to expect from a first study, if no prior work has been done.( I know I named my post this, but mostly to spark some discussion) I think good first studies could be directed at finding out how multivariate bike fitting actually is, and which the strong interactions are. Also I think some interaction could be reduced by using smart definitions. For example if we would be interested of the effect of hip angle on maximal average power over a certain duration. If we define hip angle with the leg at bottom dead center my guess would be that it interacts strongly with Saddle height and crank length. When defined at top dead center, I would expect the interactions to be much less.

Also I found this great page.
http://www.medicineofcycling.com/research/
There is a lot of literature on there, but Ill need a while to sort it all out.

The other good point you made is that an adption period might be needed. Do you think that is always true? For example, if the client comes in with a hip angle that is too tight, would you expect, that opening the hip angle would not show instantenious improvement?
Quote Reply
Re: Proof bike fitting works??? [seebritri] [ In reply to ]
Quote | Reply
I am not at all saying, that faster is everything in bike fitting. But some athletes come in with this being the major goal. I see a lot of bike fitting pages that throw around numbers like "our bike fit shaves around 2 minutes of a 40k time". I think if we make claims like, that we should be able to support it.
Also even if we go away from looking at the individual parts of a bike fit, it would be interesting to see if the sum of all the fitters decisions did make the athlete faster. This should be possible to estimate by the fitters who are also coaches. For example they could compare the 4 weeks improvements after a bike fit with to a control group who did not get fitted during this time.
If you say that it is hard to show with statistical significance that bike fitting makes an athlete faster, you are saying that the variance between fits is bigger than the gain. This then either assumes, that bike fitting does not show big improvements in speed or that the gain varies a lot (some athletes get faster after a fit, some get slower). If this is the case, we should overthink our fitting strategy.
When the client is happy this is a very good sign, that the comfort part and the psychological factors are done right in a fit. I do not think it is a good indicator that the athlete are indeed faster. I am thinking of loads of non-aero aero equipment, where the athletes swear they are feeling faster.

Another question: Fitters go to fit classes where they pay quite some money to get told how to do it "right". I think these companies then should be able to show some evidence that their concepts work.
Quote Reply
Re: Proof bike fitting works??? [IJ] [ In reply to ]
Quote | Reply
So when we assume that we need a multivariate analysis, wouldnt it make sense then to also have a multivariate fitting approach? To my knowledge, all the major fitting concepts treat the goals as independent variables.

when you do a bike fitting you are varying multiple variables by definition, I assume you do not only raise or lower the seat but you may wedge the shoe, add a shim, change the length or angle of the stem, change the set back or angle of the seat, etc etc. As well you may suggest training to improve bilateral flexibility such as tight hip flexors and the list can go on... so bike fitting is very much a multiple variable manipulation, some may be correlated and some may be fully independant. In addition you may not change all of the same thing on every client so that further confounds the use of statistical inference. I have done a few research projects using multiple variable and measures and hen the change is small or the variability large you need enormous data sets to ensure that you can reach a modest 95% confidence interval. In reality that still says that 5% of the time the model you have developed does not work, and as well if there are outliers you have to deal with that effect. Let's also assume those who come in for a fitting are not the average person who rides a bike but rather a very specific sub set, so you have already started with a biased sample which statistics does not really like too much. This is all to say that to "prove" that a fitting does, x, y or z is something that is a bit difficult. So though it may be true anecdotally that it improved something for one individual to say that can happen for all is maybe stretching it a bit.
Quote Reply
Re: Proof bike fitting works??? [IJ] [ In reply to ]
Quote | Reply
I am going back to school to peruse a kinesiology degree and master's degree.

I have been doing bicycle fitting as a "professional" since 2005. I am a Level 2 or "Master" Body Geometry Fitter going for Certification exam this next winter.

I believe the days of bike shops charging $100 to $150 a hour after they have a employee attend a 3 day class are going to be coming to a end soon.

I am working with a woman who is trying to get to Kona this year, as part of her Sponsorship from my shop we are giving her all the bike fitting attention she wants. I have seen her pretty much every month. I am using her as a test subject for some of my theories.

I can not believe how much her flexibility has changed from January to now with her training ramping up. I am able to change her position a little more aggressive every time i see her.

I am still at a loss how to show her that she is getting faster. but with her and her coach looking at her times and data she is faster, and after last weeks "tweaks" she feels the best she has ever felt on the bike.

The only "Proof" that i have been able to find while doing bicycle fitting is Customer's perceived value in the bicycle fit and if they enjoy there bike more.

I believe bicycle fitting is about making our customers enjoy the sport we all love more. If they enjoy riding more after the fit then before, I believe that is all the "Proof" I need.

TLDR: If customer is happy at the end of the fitting and 6 months later does "proof" matter?
Quote Reply
Re: Proof bike fitting works??? [s5100e] [ In reply to ]
Quote | Reply
It is my understanding, that you do not need to use multivariable analysis when the variables are independent. Then you can just run multiple single variable tests, which cuts down on the size of the tests. With additional input.and assumptions for example a continuous and or monotone or even simplified to a linear behaviour we can cut down on Test samples again. Of course tests would need to be included that question these assumptions.
I think the only way to arrive at a useful test is to gather all we know about the topic and use this as an input to design a specialized test protocol. That's why I wanted to gather some opinions and ideas here.
Quote Reply
Re: Proof bike fitting works??? [ErikSchmidt] [ In reply to ]
Quote | Reply
I think its great, that you have one athlete with whom you can try different things with, I also have two people that end up being the test subjects for my ideas. I also test nearly everything on myself as well.
Also I agree that if you want to get the most out of fitting, you have to see it as a process not just a one time event.
If your athletes goal is to qualify for Hawaii then she has a clear performance related goal. If she ends up not getting the ticket because she felt great but actually didn't get any faster, then I think the fit did not reach its goal.
Also if you end up going to a windtunnel you will get quantified results of the aerodynamic improvements. In my opinion, to balance this with the physiological factors, both sides need to bring numbers.
Quote Reply
Re: Proof bike fitting works??? [IJ] [ In reply to ]
Quote | Reply
Does bike fit work
"Does your dog Bite.
No.
That is not my dog."
the Pink Panther

It all depends on the goals of the cyclist.the skill of the Fitter. A search for truth. testing. Re testing the ability to reproduse results.
Defining what you are calling a bike fit before you start and what you think you can accomplish during a fit. I like to have a baseline before we start. then we can measure changes in the bio markers . the goal is first do no harm. then try to improve comfort and efficiency, Bike handling speed and power.
Can you quantify your results? some of us can.

http://www.medicineofcycling.com/conference/

as to the document you found it is a by product of Medicine of cycling .The next meeting is August 22-24, 2014

We are planning on day dedicated tp bike fit.



[/url]
Happy Freedman
Orthotic Consultant
Bike Fitting Specialist
Prosthetics and Orthotics/DME
Hospital for Special Surgery
510 East 73rd Street, Suite 201
New York, NY 10021


Happy
Quote Reply
Re: Proof bike fitting works??? [Happy] [ In reply to ]
Quote | Reply
Happy wrote:
http://www.medicineofcycling.com/conference/

as to the document you found it is a by product of Medicine of cycling .The next meeting is August 22-24, 2014

We are planning on day dedicated tp bike fit.

If we are interested in going to conference what do we need to do to attend?
Quote Reply
Re: Proof bike fitting works??? [ErikSchmidt] [ In reply to ]
Quote | Reply
Click on the link below to get to medicine of cycling web page


http://www.medicineofcycling.com/conference/

Happy
Quote Reply
Re: Proof bike fitting works??? [Happy] [ In reply to ]
Quote | Reply
Do you need to have a medical background or can the run of the mill bicycle shop fitter attend is more of what i am asking.
Quote Reply
Re: Proof bike fitting works??? [ErikSchmidt] [ In reply to ]
Quote | Reply
Its open to all who are interested. MD,PT, OT ,DPM, Trainers, Coaches, Bike fitters and others.If you have questions contact me at Freedmanha@hss.edu .I have been to all 4 meetings. and am planning my 5th.

Happy
Quote Reply
Re: Proof bike fitting works??? [IJ] [ In reply to ]
Quote | Reply
For what its worth:
I am not a fitter, but part of my job as a physiologist help Specialized validate their fitting. We've done a fair amount of research on the BG fit concept with positive results for the system. What's difficult is the time and money associated with finding a large group of like enough subjects to find statistical significance.

The above poster is a physiologist employed FastTalkLabs. However, statements are not made on behalf of nor reflective of FastTalkLabs in any manner... unless they're good, then they count.
Quote Reply
Re: Proof bike fitting works??? [IJ] [ In reply to ]
Quote | Reply
J Biomech. 2014 Jun 3;47(8):1894-8. doi: 10.1016/j.jbiomech.2014.02.029. Epub 2014 Mar 4.
Optimal cycling time trial position models: aerodynamics versus power output and metabolic energy.
Fintelman DM1, Sterling M2, Hemida H2, Li FX3.

http://www.ncbi.nlm.nih.gov/pubmed/24726654


What do you think?

Damon Rinard
Engineering Manager,
CSG Road Engineering Department
Cannondale & GT Bicycles
(ex-Cervelo, ex-Trek, ex-Velomax, ex-Kestrel)
Quote Reply
Re: Proof bike fitting works??? [damon_rinard] [ In reply to ]
Quote | Reply
Thanks for your answer,

I read the paper and it is definitely an inspiration in which direction to go. What I was missing in the text, were retests on how repeatable the Spiro-measurements where and how they interpolated the values between the measurements. To my knowledge, its hard to get a repeatability above 3-4% with most spiroergeometers (and test subjects for that matter).

When I started, I thought it would be hard to get the aerodynamics side predicted well enough, now I think the problem is more on the physiological side.
Quote Reply
Re: Proof bike fitting works??? [IJ] [ In reply to ]
Quote | Reply
Hi IJ,

Yes, it's nice to see some published science.

I know that some pretty good studies are not published. The knowledge is expensive to get, and can be a competitive advantage if it's kept secret.

Cheers,

Damon Rinard
Engineering Manager,
CSG Road Engineering Department
Cannondale & GT Bicycles
(ex-Cervelo, ex-Trek, ex-Velomax, ex-Kestrel)
Quote Reply