For the "pedaling technique doesn't matter crowd"

[CoachFergie]

Where is that change documented as being statistically significant (the only change that counts in a scientific paper)?

Ummmm, I said they matched the training loads, I assume to remove any confounding effect (any change due to the difference in training rather than the difference in crank).

Luttrell found a statistically significant change in efficiency. If you don't believe that is an important performance metric, so be it.

And three other studies have found no difference in efficiency.

In the big picture the importance of this study is similar to a poster presentation except it was felt to be more meaningful as it was selected to be presented orally before the group. You may not like it but the design was such that the subjects acted as their own controls.

More nonsense.

This study was considered worthy for this recognition by the Canadian Society of Exercise Physiologists. Contact them if you don't like it. Every year major journals publish a supplement of usually hundreds of studies not quite worthy of publication but worthy of mention. You might be surprised but people actually read these and sometimes they contain gems.

Yes, I do look at them, several more Gimmickcrank studies showing no benefit to performance.

 

[CoachFergie]

Snake oil
From Wikipedia, the free encyclopedia

Snake oil is an expression that originally referred to fraudulent health products or unproven medicine but has come to refer to any product with questionable or unverifiable quality or benefit. By extension, a snake oil salesman is someone who knowingly sells fraudulent goods or who is himself or herself a fraud, quack, charlatan, and the like.

Ditto

 

[CoachFergie]

The Dixon study discussion from Slowtwitch...

http://forum.slowtwitch.com/gforum.cgi?post=1054379;

inc this little gem...

Mr. Polarbear.

Your results are anecdotal. These are part of a systematic study with reported statistical significance to the results. There is no comparison.
--------------
Frank,
An original Ironman and the Inventor of PowerCranks

 

[JayKosta]

I've read the full text of the Burns article and it mentions several concerns that have been discussed in this thread -

1) The 5 week test period - The article mentions that the 5 week test was an examination of initial adaptation to uncoupled crank training.

2) Comparison with regular crank training - The article mentions that there were no negative effects from training with the uncoupled cranks, in the 5 week test.
-----------------------------
My understanding from the article is that a test of longer duration is needed, and that the testing procedure needs to incorporate measurements of actual 'cycling performance' - perhaps a time trial.
It would also be interesting to have before and after 'spinscan' or 'WattBike' displays of the the testers' pedaling style to see if they show any differences.

Jay Kosta
Endwell NY USA

 

[CoachFergie]

It would also be interesting to have before and after 'spinscan' or 'WattBike' displays of the the testers' pedaling style to see if they show any differences.

Neither are a valid test of pedalling. Both make an estimate in the similar fashion to the Cateye Cyclesimulator estimation of power.

 

[hiero2]

Of course they would not make the observation that 5 weeks was insufficient data. That would be like simply saying to the publication: "THIS STUDY IS TERRIBLE, PLEASE DO NOT PUBLISH IT" . . .

Not the case at all. Studies, where the conclusion is "this needs more study time" are published on a regular basis.

I've read the full text of the Burns article and it mentions several concerns that have been discussed in this thread -

1) The 5 week test period - The article mentions that the 5 week test was an examination of initial adaptation to uncoupled crank training.

2) Comparison with regular crank training - The article mentions that there were no negative effects from training with the uncoupled cranks, in the 5 week test.
-----------------------------
My understanding from the article is that a test of longer duration is needed, and that the testing procedure needs to incorporate measurements of actual 'cycling performance' - perhaps a time trial.
It would also be interesting to have before and after 'spinscan' or 'WattBike' displays of the the testers' pedaling style to see if they show any differences.

Jay Kosta
Endwell NY USA

Jay;
In this instance, direct quotes to back your conclusions would be useful. Nay, more than useful. Given the heat of this "conversation", I would say essential.

I know I'm not paying good money to read the full text.

 

[JayKosta]

...
Jay;
In this instance, direct quotes to back your conclusions would be useful. Nay, more than useful. Given the heat of this "conversation", I would say essential.
...

=========================================
hiero2,

My copy contains this copyright statement, and I will not do any other 'direct quotes' (even if they are 'legal' by fair-use law). And I don't plan to ask for permission.

"Copyright of International Journal of Sports Physiology & Performance is the property of Human Kinetics Publishers, Inc. and its content may not be copied or emailed to multiple sites or posted to a listserv without the
copyright holder's express written permission."

And YES, the article does say that more study is needed to determine results of training for a longer period.
Also I don't think additional details from the Burns study will supply any more 'light' to the 'heat' of this discussion.

Jay Kosta
Endwell NY USA

 

[hiero2]

=========================================
hiero2,

My copy contains this copyright statement, and I will not do any other 'direct quotes' (even if they are 'legal' by fair-use law). And I don't plan to ask for permission.

"Copyright of International Journal of Sports Physiology & Performance is the property of Human Kinetics Publishers, Inc. and its content may not be copied or emailed to multiple sites or posted to a listserv without the
copyright holder's express written permission."

And YES, the article does say that more study is needed to determine results of training for a longer period.
Also I don't think additional details from the Burns study will supply any more 'light' to the 'heat' of this discussion.

Jay Kosta
Endwell NY USA

Fair enough. Thanks.

 

[FrankDay]

Not the case at all. Studies, where the conclusion is "this needs more study time" are published on a regular basis.

But, that is not the only conclusion of the study and this statement is usually found in the discussion portion of the study, not the findings or conclusions area of the paper. So, this statement, is usually preceded by the main conclusion that "what we did was not sufficient to see positive results". Such a main conclusion can be easily misinterpreted by those who wish to misinterpret it to support a particular bias that this result "proved" their contention. Further, it doesn't matter whether the authors make that statement or not (in this case they did) because, in science, method matters (it is why there is also a "method and materials section of studies) and it is always the case that doing something in a different fashion may result in a different outcome.

 

[FrankDay]

I know I'm not paying good money to read the full text.

No need to pay good money. The full text should be available on-line at Researchgate. It is free to sign up then search for uncoupled cranks and two articles will show up, the Burns study and another "negative study" by Williams et al. If one also searches for PowerCranks the Luttrell study becomes available. The full text is usually available to registered users if one then clicks on the download or View full text link if present.

One thing I like to do when reading these "negative" studies is to look at the raw data to see if I can see trends that might be uncovered if either more people were included or more time were involved. Take the Williams study. Table 2 shows the raw data differences compiled by group. (Remember, scientific papers usually require a P of <0.05 before something is noted to have reached statistical significance. This means the odds of the change being due to randomness of is less than 1 in 20.)

The VO2 max of the uncoupled group went up while the VO2max of the coupled group went down but the P of these differences was only .420, meaning the odds of this not being random was only about 2.5 to 1. On a ml/kg basis the odds of this change not being random improved to about 2.9 to 1.

Gross efficiency of the uncoupled group improved from 19.7 to 20.9% (1.2%) whereas the coupled group only improved from 19.8 to 20.3% (0.5%). The P of this change was 0.250 meaning the odds of it not being random was 4 to 1. There was a similar change in GE at 200 watts.

Changes in cadence almost made the statistically significant level. The uncoupled group went from 91 to 87 whereas the coupled group went from 90 to 89. The P for this change was 0.057 making the odds this not being random 17.5 to 1. Remember 20:1 is the standard for statistically significant (although some journals allow 10:1).

Time trial power increased in the uncoupled group from 284 to 298 (14 watts) whereas the coupled group increased from 274 to 281 (7 watts). P for this difference was 0.125 meaning the odds were 8 to 1.

All of these changes suggest that there were differences between these groups but that the study design simply wasn't powerful enough to pick them up. But, scientific rigor prevents authors from stating there were differences such that those who only read the conclusion (like Fergie) conclude the study is definitive when it most certainly is not.

 

[CoachFergie]

All of these changes suggest that there were differences between these groups but that the study design simply wasn't powerful enough to pick them up. But, scientific rigor prevents authors from stating there were differences such that those who only read the conclusion (like Fergie) conclude the study is definitive when it most certainly is not.

Cherry picking the studies to fuel your confirmation bias.

Here is the link to Burn's Masters Thesis on the topic.

http://ro.ecu.edu.au/cgi/viewcontent.cgi?article=1017&context=theses

 

[coapman]

It would also be interesting to have before and after 'spinscan' or 'WattBike' displays of the the testers' pedaling style to see if they show any differences.

Frank has comparison graphs which compare powercrank pedaling and standard crank natural pedaling. They clearly showed powercrank style pedaling was no different from circular pedaling and everyone knows research by Coyle et al confirmed mashing is more effective than circular.

 

[CoachFergie]

Have you cherry picked the data from Sperlich? A non significant decrease in most performance and physiological measures from 30 training sessions using a Gimmickcrank.

http://link.springer.com/article/10.1007/s00421-010-1785-z?LI=true

 

[FrankDay]

Frank has comparison graphs which compare powercrank pedaling and standard crank natural pedaling. They clearly showed powercrank style pedaling was no different from circular pedaling and everyone knows research by Coyle et al confirmed mashing is more effective than circular.

While "everyone" knows this "fact" it is an over interpretation of his study. In fact, even Coyle does not come to this conclusion. This comes from the discussion portion of that paper: "Therefore, the superior characteristics of group 1 compared with group 2 could reflect adaptations derived from more years of endurance training; yet genetic predisposition cannot be discounted. … It is clear that any future research attempting to quantify effectiveness must consider more than the orientation of the applied pedal forces. This finding makes it unclear whether feedback devices that allow a cyclist to improve pedaling effectiveness (2,3,21) will lead to improved performance. … However, the present study found less negative torque during the upstroke than the previous studies. In some instances this reduction in negative torque was produced by pulling up on the pedal. Data from our laboratory (Kautz et aI., submitted for publication) indicate that the torque production patterns exhibited by a cyclist were similar with increasing work rates, with the major difference at higher power outputs being increased peak torque during the downstroke and, to a much lesser extent, reduced negative torque during the upstroke. … Therefore, the exact factors that allowed the subjects of group 1 to generate more power and to produce higher peak torques during the cycling downstroke are unclear. … it should be recognized that biomechanical factors which alter the distribution of work also have potential to reduce fatiguability and improve performance."

Interpretation of scientific studies are hardly ever black and white.

 

[FrankDay]

Have you cherry picked the data from Sperlich? A non significant decrease in most performance and physiological measures from 30 training sessions using a Gimmickcrank.

http://link.springer.com/article/10.1007/s00421-010-1785-z?LI=true

All that can be said is that what they did, (about 50% of training time for 6 weeks done on uncoupled cranks) in this group of highly trained athletes was not sufficient to show any difference between groups.

I contend that it is much more difficult to show improvement in highly trained cyclists compared to average cyclists. Selection criteria for participants required that they had more than 40 races in the prior year and their training be more than 10 hr per week with minimum volume of >300 km per week. This sets them well apart from most cyclists and the uncoupled intervention was only about 50% of their training time, at best. We have almost always advocated for immersion training for the best effect. And, this study did not check to see if there was any change in style.

I, again, find this study inadequate to make any determination regarding the long-term benefits of proper use, i.e., according to our recommendations. You should reach the same conclusion if you were intellectually honest.

For those who are interested the entire paper can be viewed at researchgate.

 

[CoachFergie]

I contend that it is much more difficult to show improvement in highly trained cyclists compared to average cyclists.

Frank Day: Master of the obvious!

 

[CoachFergie]

We have almost always advocated for immersion training for the best effect. And, this study did not check to see if there was any change in style.

I, again, find this study inadequate to make any determination regarding the long-term benefits of proper use, i.e., according to our recommendations. You should reach the same conclusion if you were intellectually honest.

Yes Frank, but in the real world we race on normal cranks and most riders don't have nine months to devote to immersion training. Your claims about the ideal study or proper use are just as ludicrous as your claims that at some point longer than six weeks that Gimmicranking suddenly makes you 40% more powerful.

An intellectually honest person wouldn't cherry pick the data to suit nor interpret the data beyond what is presented.

 

[FrankDay]

Yes Frank, but in the real world we race on normal cranks and most riders don't have nine months to devote to immersion training. Your claims about the ideal study or proper use are just as ludicrous as your claims that at some point longer than six weeks that Gimmicranking suddenly makes you 40% more powerful.

It doesn't take 9 months to discern a difference. Dixon, with a similar cohort to Sperlach, was able to discern statistically significant results in only 6 weeks. What was the difference? Dixon's subjects used immersion training vs Sperlach's part-time intervention. Study design makes a difference.

Oh, and I don't claim that at some time after 6 weeks people suddenly become 40% more power. Some time starting at about 4-6 weeks most people (with immersion training) start seeing more power which gradually increases to become 40%, on average, in 9 months. And, while you may not have the time or inclination to devote 9 months to immersion training on them, many do. In fact, we have many users who are 8-10 years out on immersion training with them. Now, many users also see improvement with part-time training on them but the results usually take longer and are much more variable, which is why I recommend immersion training. The results will come faster and usually be larger. I think most people who decide to try them would like to know that.

 

[CoachFergie]

It doesn't take 9 months to discern a difference. Dixon, with a similar cohort to Sperlach, was able to discern statistically significant results in only 6 weeks. What was the difference? Dixon's subjects used immersion training vs Sperlach's part-time intervention. Study design makes a difference.

No control in the Dixon study so all that proves is that training works.

Oh, and I don't claim that at some time after 6 weeks people suddenly become 40% more power. Some time starting at about 4-6 weeks most people (with immersion training) start seeing more power which gradually increases to become 40%, on average, in 9 months. And, while you may not have the time or inclination to devote 9 months to immersion training on them, many do. In fact, we have many users who are 8-10 years out on immersion training with them. Now, many users also see improvement with part-time training on them but the results usually take longer and are much more variable, which is why I recommend immersion training. The results will come faster and usually be larger. I think most people who decide to try them would like to know that.

Anecdotes are not evidence!

 

[FrankDay]

No control in the Dixon study so all that proves is that training works.

Sure there were, you just don't understand the study design.

Anecdotes are not evidence!

Sure they are. They just don't count as scientific evidence.

 

[CoachFergie]

Sure there were, you just don't understand the study design.

Imaginary control group's don't count

All Dixon shows is that a period of training works. We already knew that.

 

[FrankDay]

Imaginary control group's don't count

Enough of a control group to satisfy the Canadian Society of Exercise Physiologists (wonder what they know that you don't? - plenty I bet)

 

[CoachFergie]

Enough of a control group to satisfy the Canadian Society of Exercise Physiologists (wonder what they know that you don't? - plenty I bet)

There was no control group. All Dixon proves is that training works. If they had compared the Gimmickcrank group to a group performing a similar training load using normal cranks like Burns, Williams, Bohm and Sperlich studies we would have something to discuss.

 

[CoachFergie]

Enough of a control group to satisfy the Canadian Society of Exercise Physiologists (wonder what they know that you don't? - plenty I bet)

Appeal to authority. Watch the Snake Oil Salesman in action.

 

[acoggan]

No one has ever demonstrated that pedaling efficiency improves with more training

Patently untrue. In fact, the very 1st study reporting that cycling efficiency improves with training was published in the 1930s.