Sign up Latest Topics

  Author   Comment  

Development Team Member
Posts: 36
I did a comparison of pedaling technique and SmO2/tHb between different riding positions at FTP using an improvised 150s interval - 100s rest protocol. The aim was more to look at general trends, to determine what kind of more detailed assessments to do and to set goals for this year, than to analyze details.

I am not a physiologist, so I would be very interested in hearing thoughts of others here about this experiment, how to interpret the results, suggestions on what/how to improve this kind of experiments as well as what to study next.

The SmO2/tHb data for right VL (for downstroke), right Tibialis Anterior (for upstroke), right Deltoid (as less involved muscle). These are averages from the two sets intended for looking at general trends. 

Subjectively, sitting with hands on top of handle bar was easiest (and probably able to continue for an hour as the load was FTP), followed by free hand, then drops, then standing with hands on hoods. Drops was not continuable, and standing was close to limit at the end of the 150s interval.

The pedaling data is interesting as well. Note that - as I understand it - the Pioneer efficiency metric includes radial force and is more "difficult" than efficiency scores that only take tangential forces into account.

There is a big difference between sitting and standing which should be visible in the SmO2 data. Are the more nuanced differences between hand positions noticeable in the SmO2 data?

The timeline for sitting also shows that the subject has two quite different pedaling styles - a very round but tiring pedaling style and one that is sustainable but less "efficient" - between which he is switching in 10 to 20 second cycles. In more detail:

My current thoughts:

- Improving sustainable power in more aerodynamic riding positions by body flexibility and muscle coordination (for flats) and pedaling technique when standing (for climbs) might benefit outdoor cycling performance more than attempting to further improve FTP (from the current 4W/kg) in an upright riding position.

- While sitting upright and being able to breathe well is good for assessments of best performance, I am no longer sure that that has been wise for physiological assessments including 5-1-5 with NIRS and VO2 / lactate tests if the aim is to make informed decisions in order to improve real world performance. 

- It might be interesting to do 5-1-5 and compare limiters with hands on top / hoods / drops.

- Not sure yet what to make of and how to "improve" those two pedaling styles and how NIRS could be useful for that.

pdf Graphs2.pdf     xlsx 20170122215424 v2 wasp with graphs 3.xlsx     


Development Team Member
Posts: 1,501
These  are great  feedback's  for any  cyclists  and great to see as we have hundreds of  this datas  collected many years back to  be used  for different cycling  situation in the road  and on MTB as well as on the track  during 15 years  of  having the privilege  to be a part of  a huge  cycling camp organisation in mallorca  and  Girona  ( Giverola )  which was as well the  spring training center  for  Swiss  cycling. So thanks and I  enjoy to see , what     cyclist  come up  this  days  with the  same, but much cheaper  equipment.
here  juts  from the  archive a  short  shot  from an even closer view  we  had  to lock at  whether position actually would change blood flow  from compression outflow  and  decompression inflow  to occlusion  and occlusion outflow.

super close look geoff at start after 1 min break.JPG 

Than a  question. You write  after a  20 min " warm up "  we you did some first loads. . Than you added  "warm up" was  not  over. based  to me as  it seems  on tHb reaction ?
 Where you sweating  after the 20 min. ?
How  warm was  you  core temperature at the start of  the warm up  and how  warm  was it at the end u off the 20 min. ?

What is  the average  physiological system integration   physiologists  suggest it may need  to see a possible  homeostasis  if  the   performance  or intensity  would allow a homeostasis. Here is one of  many  10 min  same load  data collection. 
Now each athlete  p or better person has a unique time frame  of   adjustment  and all depends on how  the different systems  work together.
So even the  top 3 minute  guys   today agree that  5 - 8 min is the optimal length we need  to see possible  reactions. So  3 min is  on the very short  side  on that  information option.
calib 5.jpg 
Summary  : great  work  and I hope  we see mhc more of this  finally  from the  cycling community.

Last  question. Did  you as well use  a  BSX  as you have one  to see the reactions there in comparison  with  for  example MOXY  and    even portamon ?  Thanks  for this great information


Development Team Member
Posts: 36
Juerg, thanks for your feedback.

Concerning "warm up":

I was - very simplifiedly - looking at VL SmO2 in the rest periods as a very simple trend. In the rest before and after the first interval it only gets up to 80%, in the rest periods after the 2nd interval it tops out at 90% and is still above 85% when getting slowly tired in the final intervals. (After the standing intervals it's even higher but that's I believe due to the very different load when standing.) The first two intervals also subjectively felt harder than from the third one on - which in total led me to think that I still wasn't properly "warmed up" after the 20mins.

With your feedback, I am thinking now that this was a bit too simplified and I should probably stick to a proper definition of the term warm up. 

I think I was starting to sweat after around 15 min. Didn't look at core temperature - how do you measure that during such test?

Concerning length of intervals:

My plan here was to first look at general trends - for which I didn't think I needed to reach homeostasis in the intervals - before narrowing the conditions down and re-doing the experiment. So the priority here was to test the conditions in a as similar as possible physiological state - and be able to check that the physiological state indeed had been more or less comparable. For the latter, I thought I needed at least two sets which I could compare against each other in order to be able to see whether effects would have come from not being warmed up or already getting tired. For the former, I thought I needed a good-enough warm up before starting the intervals, and the whole duration should be short enough so that the final intervals wouldn't be too compromised. With four experiment conditions and two sets, 150 seconds was the maximum that I felt was doable.

And the results make me think that I should care more about what riding position I do the testing in, if I want the results to be applicable to real world riding. (A limiter that I identify when riding on the top of the handle bar might not be relevant if I am in the drops.) 

That said, with your feedback, I now think that reducing the number of experiment conditions to three (say, sitting-hoods (as the most general position that also does not limit respiration), sitting-drops (for sprints), standing-hoods (for climbing)) and increasing the length to 4 minutes each could be a "better" experiment design - and doable if I shorten the warmup a bit. 5 minutes would probably be only possible with just two experiment conditions (unless I reduce the load). (I actually had enough power left to do some Zwift riding afterwards... :-)  )

Concerning other sensors:

I had one BSX on the calf, the data is in the xlsx and some of the powerpoint slides. Just saw again how different the scale of its data is compared with MOXY, so I left it out of the data I was posting as images. I guess it's not just the scale that's different and it may be measuring something else or at different depths so there might be some other usage for it ...?

Would you - or someone else - have any suggestions which muscles to use the MOXY on when comparing riding positions? Are the current locations - VL, Tibialis Anterior and Deltoid - reasonable? (GM or BF could be nice to complete the crank rotation, but if I remember correctly their SmO2 wouldn't change so much and I have just 3 MOXYs anyway...)

The only other sensor I had was a fingertip SpO2 sensor that doesn't record data so it's in a movie file and I haven't extracted that to CSV/Excel. Had never tried that before so I was just looking how SpO2 behaves. Would love to get hands on a VO2Master but maybe not yet.

Graphs3-complete timeline.jpg


Development Team Member
Posts: 369
Takura, thanks for sharing this interesting data and the excellent presentation of it.

Development Team Member
Posts: 1,501
Takura  first of  all did  you  mentioned  somewhere  :
 I am not a physiologist, 

No you are not  that's  why  you do this incredible  feedbakcs  with loud  thinking  and a  super open mind. Wowwww 
Remember the  article.  Where  do we  place  luckily  for  all readers  Takura ?
Thanks so  much.

The practice of science at the edge of knowledge.
Grinnell, Frederick

Postmodernism;; Science -­-­ Philosophy;; Theory of knowledge
The writer describes the postmodern view of science and the way in
which science should be taught.

IN RECENT DECADES, postmodernists and sociologists of science have argued that science is just one of many human activities with social and political aims-­-­comparable to, say, religion or art. They have questioned the objectivity of science, and whether it has any unique ability to find the truth. Not surprisingly, such claims have evoked a negative response from proponents of the traditional view of science;; the debate between the two sides has been called the science wars. In the debate, scientists have made few attempts to meet the postmodern critique on its own grounds, through serious reflection on the everyday practice of science. Yet that is the only way to understand the nature of science and
the features that distinguish science from other activities.
The behavior of baseball umpires helps define the issues. There are three types of umpires. The first type says: "I call balls and strikes as they are." The second says: "I call them as I see them." And the third says: "What I call them is what they become."
What distinguishes the types of umpires is not the situations in which they find themselves, but the attitudes that they bring to their work. As a result of those attitudes, they practice umpiring differently.
The first type claims truth;; the second, perspective;; and the third, power.

So  first  Warm up.

There is a  clear limitation of  NIRS  and tHb  to look at " warm up" The tHb is an indication of blood volume /  flow in the tested area.
But an increase in tHb  does NOT  indicate  by itself, whether there is a chnage in   warm up.
 In fact the term warm up  as in temperature is not a good term  for any activity preparation at all..
Core temperature tends  to  try to  stay in a very  stable range +-. How  to test  we use  actual  capsule  you swallow ( look at the red bull studies )  to actually  see that  core temperature relative   okay  it is  still not real  core as it  sits in the stomach. 
I use a  infrared bio harness sensor, where I get  at the same time  RF  respiratory  quality, HR  HRV  and much more . The Infrared sensor is  placed  just in the xiphoid area, so mediastinum.
Than we  can compare   temperature trend  which is  closely linked  to RF  and respiratory  depth ( CO2 ) as well as HR  . 
Will try to dig out an old  studies we  compared  glucose  reaction and  temperature reaction from an Australian study to  classical  and our  metrics.
Here a  fun print out  for m an interval   and the  temperature  reactions. the short  sections are the load section.

skin temperatur.jpg 

This is  from an ice hockey workout with short loads  and  longer rests. Below a  basic  experiment  we did  to see the difference in TSI  %  and SmO2 in MOXY  when influenced  by temperature as seen above. You can see  the  red and blue  as usual as O2Hb red and HHb blue and the  thickness as the  depth of  the feedback  tick is deep  thin is  more towards the surface. What can you see.  ?

Now we  could go into much more details    to show  where NIRS  can be used  for  " warm up , but temperature  per see is not a great indicator  of readiness  for a  muscular activity. In fact " cooling down prior  to a certain activity ( depends  what you cool down my  actually improve performance. The  fact  that you  started  to sweat after 15 min indicated  ????

The  reaction to actually avoid warming up more ( sweating could mean  you have to  control core temperature  and you try to get rid of heat  so you  try to at least keep  temp stable or  cool  down creates  an unfortunate  shift of  blood  which contains  O2  from where to where. ? More later  to this great feedback.


Development Team Member
Posts: 227

Hi Takura,

Impressive data, big thanks for sharing, I’m in the process of doing similar on TT bike with different cranks. Your excel sheet is gold!!! Big thanks for posting (I’m a number nut [smile]


Is “pedaling metric time line” from Pioneer app? I have Garmin vectors and had to use excel to obtain similar data, major pain. I’m working on my “pedaling style” I have noticed I literally pedal squares few months back. I slowly see improvement but it seems like long way.  I think it is all very individual, in my case after working on hip abductors and tensor fasciae latae I feel like I stabilized my knee and improved my off the saddle by 100% but as well pedaling at higher power, I feel my leg is much more solid and stable.



For real world racing, I think longer intervals in extreme positions would be more in line with racing. Eg. 10min in droops, followed by 10 min on the bars could simulate breakaways and rests. You work on drops in front for 10min and switch with team mate falling behind with hands on the bars and resting.

It would be interesting to see how deep Smo2 goes in droops and how quickly you can recover on the bars (the list and most comfortable positions, the rest would fall in between). I’m my case on the trainer it is about 8-10% of Smo2 on crappy day and 5-6% on good “respiratory” day.


Since now I have second Moxy I will be testing Moxy on my back, Trapezius and lower back as well) I think in my case drop in SmO2 is caused by back muscles taking more O2 as well hamstrings. Moxy is truly amazing tool to look what is happening inside. Again thanks for sharing analyzing someone’s data speeds up the process, since there is not enough days to test everything on myself [smile]


Development Team Member
Posts: 49
This really is great data made possible by great equipment.

The short stint I dabbled in triathlon I struggled at first to find a bike fit that worked - only when I finally swapped out the stock 175mm cranks for 170's did I feel good on the TT bike.   It enabled me to raise my saddle, open my hip angle up, and the raising of the saddle also flatted by back because i left my bars in the same position.

Again... this data and the use of the moxy is great.

Development Team Member
Posts: 36
Thank you to all who have replied.

Crank length is indeed something that I am planning to look into too. Not sure though yet whether to simply test on different days on different bikes or to try to test in one session and also what other position factors to change as a change in crank length will most likely make sense only if other parameters like saddle height are adjusted ... but then it's obviously no longer clear what is causing what...

Sebo, yes, this is from the Pioneer power meter. Just looked into Garmin Connect and agree that it'd be more convenient if they had graphs for efficiency and smoothness. That said, and while the force vector visualization on the Pioneer is neat, Pioneer doesn't allow to extract the raw data which is a big minus.(You'd have to either reverse engineer their proprietary file format or have an NDA with them in place.)

I am actually half way of making an ConnectIQ app that saves efficiency and smoothness from Vector as (duplicate) custom data which would make it possible to graph them in Garmin Connect similar to the SmO2 graph from the Moxy data field. The only issue there seems that there still seem to be bugs on Garmin side which make it difficult and inefficient to debug this kind of ConnectIQ apps.

Agree that something like a 10 min drops, 10 min bars test would be useful. Could probably use Moxy also to find an optimum pose/cadence/power for the rest periods?
Previous Topic | Next Topic

Quick Navigation:

Easily create a Forum Website with Website Toolbox.

HTML hit counter -