## Friday, October 25, 2013

### Random thought - Differences between locomotor profile in running and biking, or Power vs Velocity profiling

I am playing with Metabolic Power data estimates and I come to one interesting and confusing "finding" and I want to ask for an opinion and/or direction on this.

Locomotor profile (see slides from Martin Buchheit) is basically MAS and MSS, or Maximal Aerobic Speed and Maximal Sprinting Speed. Based on these two parameters we can predict performance on a lot of distances.  Add there a little bit of CV (Critical Velocity), vLT and we have a full profile.

The same could be done with cycling, but here we use power (see Power Profiling). Certain authors use average power levels over certain durations (CP5s, CP1min, CP5min, MAP, CP20min, FTP, etc) while others use ("real") Critical Power estimation.

Anyway, the rationale behind them are the same. We have a profile and we make training recommendations based on that profile and activity demands, we judge performance improvements, etc.

But, that's not the point of this blog post. The point is their difference. In numbers.

For example, in running for an average athlete MAS is around 16 km/h and MSS is around 30 km/h. The ANR (Anaerobic Reserve) is MSS-MAS or 30 - 16 = 14 km/h. I have calculated various ratios

MAS = 16
MSS = 30
ANR = 14

MAS/MSS = 0,53
MAS/ANR = 1,14
ANR/MSS =  0,46

In cycling the measurement unit is power. CP5min (or average power one can produce for 5min) or MAP (Maximal Aerobic Power) which is equivalent to MAS is around 7 W/kg for very good cyclists. Their CP5sec could be representative of MSS (let's call it Maximal Sprinting Power) and it is around 23 W/kg for very good cyclists (don't get too hang up on numbers, they are just an example). Let's calculate the ratios

MAP = 7
MSP = 23
ANR = 16

MAP/MSP = 0,30
MAP/ANR = 0,43
ANR/MSP = 0,69

We can immediately see that the profiles are totally different when we look at ratios.

Ok - this might be related to differences between cycling and running in general, or using power vs. velocity for profiling.

But here is the thing - I mentioned that I have been looking at Metabolic Power estimates. MAS or in this case MAP is around 20 W/kg approximately 57ml/kg/min O2  for an average player. When accelerating for 5 to 10m, power output might go over 80 to 100 W/kg. That's the ratio MAP/MSP around 0,2 to 0,25. Almost twice lower then when we use velocity (MAS/MSS = 0,53)

The missing piece of the puzzle is the power output in 40-60m zone when we have high speed and zero acceleration. I would be interested to see the difference between Metabolic Power (MP) estimates in 5-10m sprint from standing still, and MP when someone is running at top speed with no acceleration (e.g. zone 40-60m). Another interesting comparison might be straight-line Vam-Eval test results in both velocity and MP (estimated using GPS), and also Vam-Eval done in shuttle mode.

To cut the long story short, here are some of my thoughts:

- The "Metabolic Profile" (to blend Locomotor- in running and Power profile in cycling) might have the tendency to be different depending on the way we estimate it (velocity vs. power). Or depending on the locomotion.

- If we show that MP is lower in top speed than in top acceleration, then the top speed might not be limited by metabolic, but rather mechanical factors (musculoskeletal force applications). Bundle and Weyand wrote interesting article on this.  In other words - supply is there, but the demand is lower than maximal. [squiggle sense alert: supply~demand]

I am not sure if using MP estimates in running (short- medium- long distance) will change much, but they might change a lot in short burst, start-stop, change of direction type sports such as most team sports and tennis. This might mean that metabolic profiling based on power instead of speed might give us more information.

1. I don't really have something you could call an answer/reply to your final questions. But I can't help to think about the practical implications and if there really are any? Do you think that there can be any real implications from this in how you train? Not just how you test/evaluate (if there's such a test?)

2. It might be... Using MP data and using MP based locomotor profile (like shuttle run scores calculated in Power) as opposed to velocity calculations might yield more info for CODs sports. Time will tell. Thanks for chiming in!

I see you have entered the metabolic power questions!
I have no data on cycling, but I have published on my blog some data on metabolic power calculations.
First, there's no only one approach to calculate Pmet: the known approach of Di Prampero, and that of Gray (PhD thesis). Basically, the Di Prampero approach "weights" more on acceleration, and Gray approach does it on speed. More data-time analysis gives very similar averages, but when you apply them to simple actions (as a sprint or similar), you see some big differences.
Both approaches are based on some assumptions that have to be assumed ... so we don't know "real" data, just "calculated" data.
What I have seen till the moment is that using Di Prampero approach, Pmet is higher when performing a maximum acceleration (around 6 m/s/s) than a maximum speed (around 32-33 km/h), all measured by GPS units during training or a game.
Important to know the algorithm used for the calculation, done that each brand uses some modify version of it, and also the original published in Osgnach et. al. 2010 is just "a general idea" of it, that needs to be tuned up.
As Jacob Gudiol has replied, the important question is how we are going to use it in order to modify our training sessions ...
By the way, great job with your blog, thank you!

4. Hi Manuel,

Thank you very much for insightful answer.

One idea of application can be seen in CP/W' model and estimation of real time fatigue if MP gets to be proved valid and reliable.

The validity of those methods should be done. One approach might be to use sub-MAS (sub-VO2max) activities (both constant speed and acc/decc) and see how does MP reading correlated with VO2 reading (taking into account lag time, that can be sorted out with smoothing power).

Next step might me to check validity in supra-max activities (over VO2max) and that might involve looking for Accumulated Oxygen Deficit and MP output.

Having these we can speak further regarding the reliability of the estimates/GPS in general and think of practical applications

Thanks one more time for chiming in!