Interesting study

gvanwagner said:

Im always skeptical of these training studies. If your using trained cyclist then you won't see an 8%increase in power in the short term unless you serious increase overload or they had a terrible plan before. Then your only testing what the difference is between an increased training load and a normal one. OTOH if your numbers are a more realistic 2% increase then that's too low to be considered statistically sig in most cases. So it's a catch 22. Plus 6 weeks is a long time outside of a controlled setting- a lot of things can change.

I have seen a 12% increase in power climbing a local 4km hill just from a return after a bad flu. This was a new 20min MMP record as well. All from training to improve my 20-60min power rather than clocking up huge miles.

Andy has seen some pretty huge improvements in training subjects. I assume big enough to conclude that 8% in 6 weeks isn't anything to get excited about.

I would say there isn't anything special in what gear you do your aerobic intervals in, just that you do them!

HF

acoggan said:

I am not aware of any scientific research supporting the notion of "strength endurance" training. Moreover, from a first principles perspective there is little reason to expect much benefit:

home.earthlink.netsetrainingOpen ↗

Nice study! I will try and incorporate more big gear standing starts and big gear "seated stomps" when training. How do you explain the improvement I make after a block of SE training though? After all the goal isn't too increase muscle its just to ride quicker.

If you were coaching someone would you recommend SE efforts and or what type of efforts would you recommend besides big sprints?

Personally SE works for me (atleast uphill over 2-5 minutes) and I have the proof but I can see how too much SE could be dangerous and unbeneficial.

home.earthlink.netshow image.htmlOpen ↗

ALSO: can you make sense of this graph? What does this mean and what does it tell us about SE training.

fergie said:

I have seen a 12% increase in power climbing a local 4km hill just from a return after a bad flu. This was a new 20min MMP record as well. All from training to improve my 20-60min power rather than clocking up huge miles. ..

Exactly- you were't trained at the beginning. Id even include from a study perspective people who at that time a year haven't worked themselves from all angles. Ex. after doing the traditional coffee ride winter then any higher intensity training will magically increase your power by bounds.

Afterall, we already know the general principles that work so the studies are designed to test emerging possibilities for improvement. If you have all the first principles down then the improvement will be smaller then sig. But if it's larger then your not really trained and could have improved that much just by focusing on the basic proven issues.

Plus the placebo effect because there is no way to blind a training study. If your told just to do what you did before then you'll do your training just fine. But if your told to do these experimental exercises then that's something to get excited.

Id be much more excited about a study that said that this workout raised enzyme levels by X and this experimental exercise that's completely different raised levels by X*120% for the same workload. Then extrapolate to chronic effects because otherwise the noise makes the study useless.

This is why you can prove anything you'd like by setting up the study in a certain way. For example- does anybody really think that the Tabata protocol marks a new way for TTers to train.

You mentioned that SE training times are always faster up the hill. Then you must be putting out more power in these intervals so do you think that that might be the cause of your improvement.

acoggan said:

Yeah, except that most coaches that I've encountered are insufficiently skeptical, and at best have a limited understanding of scientific methodology (e.g., appropriate experimental design, proper use of statistics). The result is that most are far too prone to jump to conclusions based on incorrect, ambiguous, and/or anecdotal observations.

Smart comment. And it remains true the other way around as well. Look what I got with only few notepad find/replaces :

Quoted post said:

Yeah, except that most scientists that I've encountered are too skeptical, and at best have a limited understanding of training methodology (e.g., appropriate plan design, proper use of statistics). The result is that most are far too relunctant to jump to conclusions based on training, racing, and/or anecdotal observations.

(Boy! I've never sounded so smart 😄)

[quote="SolarEnergy"]Smart comment. And it remains true the other way around as well. Look what I got with only few notepad find/replaces :/QUOTE]

Except that doesn't sound like very many scientists that I know. By nature or nurture, they tend to be a very circumspect lot, plus they tend to have extensive training in research design, statistical methods, etc. You therefore don't find very many scientists who believe in magic.

dm69 said:

Nice study!

Not a study, just a cogent analysis illustrated w/ actual data. IOW, I didn't actually have to do any "strength endurance" training sessions to know, e.g., the even halving the cadence wouldn't increase pedal force sufficiently for the latter to really represent an overload.

dm69 said:

How do you explain the improvement I make after a block of SE training though?

All forms of training are likely to lead to some improvement - the issue is whether pedaling more slowly than usual while keeping everything else constant (something that Mason failed to do) results in greater and/or more rapid improvement. As I said, I am not aware of any data showing this to be true, nor any real reason to expect it to be true.

dm69 said:

If you were coaching someone would you recommend SE efforts

Not unless their racing entailed grinding along at 50-60 rpm for extended periods of time (and even then I might instead suggest that they just mount lower gears).

dm69 said:

what type of efforts would you recommend besides big sprints?

That would depend on one's personal goals, abilities, etc.

acoggan said:

Not unless their racing entailed grinding along at 50-60 rpm for extended periods of time (and even then I might instead suggest that they just mount lower gears).

This came up in the discussion that followed Andy's article. On my windtrainer with high resistance or on the hills I do produce more watts at a lower cadence. Catch 22 is do I train for more watts when I never plan to race at 40-60rpm. If I were to race that low I would never get up to speed in pursuits and in scratch races on track or road races would be exploited by other riders.

I repeat it's not the gear you do your aerobic intervals in it's that you do them!

acoggan said:

You therefore don't find very many scientists who believe in magic.

Pfft 🙄
You scientists...

Before you came up with this performance chart, peaking was all about magic...

... and easier to understand too. You want to perform? Just RAP (Rest And Peak) 😄

SolarEnergy said:

I'm web interviewing this lady at the moment. An article will be written at the end of the interview. In fact, that'll begin in a couple of days from now.

Hopefully I'll be able to find even more evidence (ground or anecdotal, whatever) than that resulting from this study. Mainly because there has been some controversy around it.

OK, first off, let me preface this by saying that the author of the study may be a fine person and all...and I don't mean this to sound overly harsh (which means, of course, that this is going to sound overly harsh...).

What the study shows me, more than anything, is that a degreed excercise physiologist is just as likely to exibit the cognitive biases that the most belief-based coach is likely to exibit. However, since it's wrapped in pseudo-science, it's much more dangerous.

I can see it now: the 'coach' prescribing big-gear workouts to 'increase strength' throws this study out to his young charges to confirm his believe in big gear works--"see, even science says so!"... It's misleading at best and counterproductive at worst.

Here's the part that really frustrates me: everything was done correctly to actually set the stage for a useful study--but then, the control protocol is totally blown?? If she had the control group doing similar intervals in intensity (measured in watts), length and frequency, except with self-selected cadence, and she had the same results, well, I probably would have been out doing big gear intervals this afternoon. OK, maybe not, because I'm suspicious anyway of the mechanism by which big gear work improves endurance peformance--but you can bet I'd be taking a very critical look at the study to try to figure out why the gains were made.

Instead, the 'control' group has no controls, and the study is obviously designed to elicit a response that confirms Mason's preconceived bias, and here I am, wondering how the hell people get through that much schooling and training without understanding some really basic tenents of cognitive thinking?

Seriously, reading throught the 'practical inference' part of the table reads like the very definition of confirmation bias. Understanding biology and physiology is great, but how about teaching folks to recognize their own cognitive biases so they don't waste their time constructing a study that is clearly designed to elicit a specific response (whether intentional or not)??

You made a good point RVC.

We will probably dig a bit deeper into the training protocol used during the study (for both groups). Though I can already say that no power data is available unfortunatelly.

Thanks

RipVanCommittee said:

What the study shows me, more than anything, is that a degreed excercise physiologist is just as likely to exibit the cognitive biases that the most belief-based coach is likely to exibit. However, since it's wrapped in pseudo-science, it's much more dangerous.

Here's the part that really frustrates me: everything was done correctly to actually set the stage for a useful study--but then, the control protocol is totally blown?? If she had the control group doing similar intervals in intensity (measured in watts), length and frequency, except with self-selected cadence, and she had the same results, well, I probably would have been out doing big gear intervals this afternoon. OK, maybe not, because I'm suspicious anyway of the mechanism by which big gear work improves endurance peformance--but you can bet I'd be taking a very critical look at the study to try to figure out why the gains were made.

Instead, the 'control' group has no controls, and the study is obviously designed to elicit a response that confirms Mason's preconceived bias, and here I am, wondering how the hell people get through that much schooling and training without understanding some really basic tenents of cognitive thinking?

Seriously, reading throught the 'practical inference' part of the table reads like the very definition of confirmation bias. Understanding biology and physiology is great, but how about teaching folks to recognize their own cognitive biases so they don't waste their time constructing a study that is clearly designed to elicit a specific response (whether intentional or not)??

Welcome to the world of scientific research. Most studies start off with a preconceived bias, and it is difficult to find studies where the researcher does not ultimately draw conclusions from their study that provides some support for their initial hypothesis. I mean, when have you honestly seen a study where the researcher concludes that trial disproves his or her hypothesis. The closest you will see is that it "requires further study" or is "inconclusive." You will hear about problems with the controls if the trial is inconclusive.

In the medical context, it is more troubling. Aside from potential biases involved in the funders and sponsors of the study, you frequently have problems with a compromised population sample. People willing to participate in medical research or drug studies are not the typical population. I am not saying all research is bs, but absent good reproducible population studies with large samples, it is difficult to draw broad conclusions.

Again, I am not indicting science. But the promulgation of pseudo science based upon small trials makes people skeptical of real science based upon properly controlled studies subject to rigorous peer review.

kopride said:

I mean, when have you honestly seen a study where the researcher concludes that trial disproves his or her hypothesis. The closest you will see is that it "requires further study" or is "inconclusive." You will hear about problems with the controls if the trial is inconclusive.

How frequently would you expect that to be the case? Assuming that hypotheses are constructed based on observations and some understanding of the physical principles at work, is it really that unreasonable to think that people could guess the most likely causes of a particular phenomenon with a decent degree of accuracy (or at least not to propose hypotheses that make it all the way to formal trial only to be completely refuted by study results)?

If the hypotheses are logically constructed and even informally tested before being put to a formal trial, then the study is really an attempt to 'confirm' the hypothesis rather than 'test' it, as you say.

Edit: I'm not sure of the protocol here, but if one conducts a study which actually refutes the initial hypothesis, is one still required to write up the results and have peers review them? I could see the value contained in the negative test result, but OTOH, if the hypothesis were due to a misconception or error in the first place then I'd basically be demonstrating for all my peers what a dumbass I am. 😄

kopride said:

Welcome to the world of scientific research. Most studies start off with a preconceived bias

If it's your synonym for "hypothesis", then I agree. And that's the way it should be.

A study is a formal test, over some more or less anticipated "hypothesis".

** edit ** oupss sorry Frenchy, I hit quick reply before having read your post. I agree with you (to say the least).

SolarEnergy said:

If it's your synonym for "hypothesis", then I agree. And that's the way it should be.

A study is a formal test, over some more or less anticipated "hypothesis".

** edit ** oupss sorry Frenchy, I hit quick reply before having read your post. I agree with you (to say the least).

My thoughts exactly. The scientific process requires formulating a hypothesis prior to testing. By definition you have some preconceived result you're testing. That doesn't undermine the results if you carefully structure the test including the control and are rigorous and honest in your interpretation of the results. If you aren't you can be sure an unbiased peer review will set you straight. If you hand pick your "peers" well then you can get away with a lot but I'd argue that stretches the term "science".

BTW there are thousands of published studies that produced negative results. Sometimes they were structured to challenge an untested belief(in which case they're again proving their hypothesis) but in my field (RF engineering, signal processing and wireless systems) I've made use of many research studies that failed to prove the original hypothesis but led to other important findings and spawned future studies. Lot's of scientists publish results of negative findings. Heck classic science is full of them, take Michelson and Morly who set out to find the "ether" believed to be necessary for light waves to exist. They failed miserably because it doesn't exist and set the stage for Max Planck to create Quantum Theory and eventually for Einstein to develop the Special Theory Of Relativity.

-Dave

Personally, I believe that an increase of 2% whilst being statistically questionable may be actually worth while.

Little bit of a rant, shoot me down if you like. 😉

I work in a field where calculations and the mountains of litrature available enable design of biological systems that generally always work, however at the end of the day it comes down to designers and operators who develop a feel for things being the ones who make them work very well. Every design parameter can be met in operation however a truely experieced operator can know a problem without even seeing the results of Labratory analysis. They may change the smallest parameter by only a couple of percent and see vast improvements in final quality. I have seen people infinitly more intelligent than myself fail in the field because they cannot let go of their precious calculations at the time of fine tuning. They are great in the office in front of their computer, just dont let them on site.

The same can be said about good coaches. They get a feel for how an athlete reacts and there comes a time that a "statistically irrelevant" change to a training reigeme can result in measurable performance at the other end. Unfortunatly if they try to conduct testing and publish findings regarding these principles people then try and shoot them down and rip their findings to bits to prove "statistical irrelevance". They would be better off keeping their findings quiet or mabye putting them in a book and letting people make their own minds up, publishing data etc is useless, as the people who have differing opinons will go out of their way to discredit them.

Personally I have never been coached for cycling, and do what feels good for me. I have however been extensivly coached by coaches in other sports (who have had international success at Worlds and Olympics) and as much as they know the numbers etc they also go by "feel".

The best thing people can do is try small changes to existing training with substitution and see if they see an improvement, then try and determine how much that improvement is the result of specific training or general increase in fitness.

Hmm over typing..... but I have undertake strength endurance training and seen results that made me feel it was worth it, and made competitors wonder what I had started doing.

Remeber that the BS stops when you hit the track. If someone beats you (especially in an individual TT) using statistically irrelevant methods they still beat you. Simple as that. 😄

Paulie-AU said:

If someone beats you (especially in an individual TT) using statistically irrelevant methods they still beat you. Simple as that. 😄

I hear Powertap are introducing a new CPU with an inbuilt "Infinite Improbability Drive"

Anyone with better splits will have either a whale or a bowl of petunias land on their head.

..... apologies to the late Douglas Adams

frenchyge said:

How frequently would you expect that to be the case? Assuming that hypotheses are constructed based on observations and some understanding of the physical principles at work, is it really that unreasonable to think that people could guess the most likely causes of a particular phenomenon with a decent degree of accuracy (or at least not to propose hypotheses that make it all the way to formal trial only to be completely refuted by study results)?

If the hypotheses are logically constructed and even informally tested before being put to a formal trial, then the study is really an attempt to 'confirm' the hypothesis rather than 'test' it, as you say.

Edit: I'm not sure of the protocol here, but if one conducts a study which actually refutes the initial hypothesis, is one still required to write up the results and have peers review them? I could see the value contained in the negative test result, but OTOH, if the hypothesis were due to a misconception or error in the first place then I'd basically be demonstrating for all my peers what a dumbass I am. 😄

I'm not aware of anyone who has argued that there is an ethical obligation to publish negative results (although it is unethical to withhold data that contradicts your interpretation of data that you do present). Moreover, negative results, although theoretically just as enlightening as positive findings, are very hard to get published. Consequently, people either don't bother to try to publish the results of studies that aren't definitive, or they reexamine the data to see what in fact it does say, and then write the Introduction as if that were the hypothesis they were testing all along. The end result is that although many scientific studies read as if they merely confirm the investigators' preconceived notions, this should not be viewed as "confirmation bias".

*I've conducted one such study myself...never really trusted the data, so decided not to publish it.

Alex Simmons said:

I hear Powertap are introducing a new CPU with an inbuilt "Infinite Improbability Drive"

Anyone with better splits will have either a whale or a bowl of petunias land on their head.

😄 😄 😄 😄