UKC

My whizzy new, over-specified, Garmin seems to have developed a wrist based running power feature following the recent firmware update. Can anyone shed any light on how useful this is, or is it an attempt to subvert runners into data obsessed wankers like cyclists?*

* I know, I know, lots of runners are already there. 

In reply to George Ormerod:

If you're into that sort of stuff power is probably the best metric to train to as it isn't influenced by other factors like HR is. The accuracy of runner's power data on a watch is probably questionable but it doesn't need to be accurate as long as it's consistent.

Personally I've never even bothered with a HRM, I just train by feel.

In reply to George Ormerod:

Absolutely nothing?

Surely it's just an algorithmic combination of all the other metrics? HR/speed/cadence/vertical oscillation etc, it's not actually a measure of "power". I guess if you spend time dialling it in and working out zones for each one it might give a useful measurement, and that's what people use perceived effort/HR/pace for already

In reply to George Ormerod:

There's one simple way of looking at it: If it's meaningful then there will be an error specification (e.g., ±10%) for both calibrated use and uncalibrated. Plus a calibration procedure. Without those you can safely assume it's just marketing fluff and treat it accordingly.

In reply to George Ormerod:

Sounds like what you really want is a Stryd and a Lactate testing kit!

In reply to George Ormerod:

Assuming the readings are accurate and running power behaves the same as cycling you would do a FTP test (functional threshold power) to determine the power you could sustain for an hour. This would then be used to calculate training zones, which you could then target for different training effects. Long slow runs, recovery, intervals etc.

Useful scenarios are potentially making sure you keep efforts light enough in recovery, long slow runs and pacing on harder efforts. 

In reply to hooch87:

I don't think that power-based run training plans can be based on cycling methodologies - The two activities are very different in terms of (a) power-performance correlations, (2) physiological variance and (3) impact tolerance. Cycling is very simple in that regard (low impact, gearing...). Running isn't.

In reply to hooch87:  You'll find that a lot of the methodology that cyclists use is adopted/pinched from what runners have been doing for a long time... the biggest difference that threshold is defined around 15 or 20 minute tests - an hour is a long time running.

I can see why you'd want power, but I think that for those people interested enough to use it properly you won't get much more out of what you can get from heart rate, as the combination of that with tests on an athletics track , where you can monitor speed very accurately, can be very powerful 

In reply to wbo2:

Garmin power and Stryd power are different in absolute value but show similar trends. You have to stick to one or the other. I could not imagine trying to run to a power figure, but I do it all the time on my Wattbike. Today's Wattbike session was warmup/15 mins at 185 watts/warmdown - ran out of effort with 20 seconds to go. Yesterdays base level run was 49 minutes at 6min 9sec pace, hardly breathing, and gave an average power of 214 watts, peak 330, over 5 almost flat miles.  I don't think Garmin running power has anything but a curiosity value, after the event. Take with a pinch of salt..

In reply to petemeads:

Out of curiousity did it do anything over time - flat, trend up? What about heart rate at the same time? Do you regularly train to heart rate limits?

In reply to Marek:

I mean when I was training seriously HR was used for this. Power removes some of the variability + for example, having an espresso then running is a sure fire way to undertrain if using HR, as you artificially bump your HR up. Not the case with power! Agree you can't just steal cycling values and applying them the running, but the principles still hold.

In reply to petemeads:

I immediately wondered whether "6min 9secs" was per mile or per km. Then I saw whose post it was (before seeing the 5 miles and doing the math). Then I knew it was per km 😁. Doing that per mile "hardly breathing" would have been what you used to regularly do 40 years ago!

I too used to be able to do that pace over shortish distances, but I'd have certainly been breathing.

In reply to wbo2:

In running gently, but trying to follow the Garmin pace suggestion, power does seem very variable/spiky - there is an element of "pilot-induced oscillation" as pace corrections are made but the main variability is gradient changes. On the flat, 200ish watts, even gentle uphill nearly 300. This is fair enough and how it should be. Overall HR was pretty constant for the running example. Biking hard was the opposite, steady HR at 125 watts for 15 mins then ever-increasing (to maximum) trying to chase the 185 figure, and failing by a couple of watts. HR and power are intimately related on the bike. I normally run to breathing/perceived effort, with a glance at HR for reassurance that I'm in the appropriate zone.

To Micheal Hood: you flatter me - 6:09 miles were always hard work, I never managed the 10 miles/hour that seems to be de rigeur for real runners...

In reply to ianstevens:

Yes, sorry, I wasn't dissing the concept of training to power, just that measuring power in running - and only using a watch - is unlikely to be remotely accurate (in any sense of the word). In cycling it's easy: you measure the torque and cadence via sensors in the crankset. But how does a running watch do the equivalent? It doesn't actually measure power, it can only access some physiological numbers (HR, weight, cadence via accelerometers) and them guess how those parameters may translate in some typical person into 'power'. Like I said in my first post, if it was tolerably accurate it would be accompanied with error specs and some calibration procedure. Does it?

In reply to Marek:

Garmin's website gives an interesting level of detail on what factors it's trying to take into account:

https://support.garmin.com/en-GB/?faq=QRiQOEq5d09foNiH1DzUt5

Seems like it's not actually using heart rate but focussing more on the actual motion. Though I completely agree with you that it's ultimately a very indirect estimate that seems unlikely to be coming up with anything particularly accurate.

I'm not sure this is entirely true, though. Products marketed at general consumers are very rarely going to prominently feature error specs and will always do their best to workaround or hide any need for a calibration procedure. Some of them will still be quite accurate despite that. My phone and my running watch both do an excellent job of tracking my location, well beyond tolerable accuracy. But neither have any published error specs or calibration processes as far as I'm aware. That's not necessarily because the data's bad. I suspect it's because lots of people won't understand error specs so there'd be no real upside to sharing the data and risk of people just seeing them as a bad thing. And any calibration that requires action by the consumer is just going to be a pain point.

In reply to George Ormerod:

I'm not sure about the Garmin specific power but I have a Stryd pod and its very good.. the new one is even more responsive as well. The only thing is if you want to use power for hills its pretty much useless as it can't account for surface conditions and for steep gradients but if you run on the road primarily power is an excellent metric. 

In reply to Luke90:

My (really quite old) Garmin already has a useless metric for that - heart rate, heart rate variability and pace apparently, mashed together to produce "performance condition".  This thread has me wondering whether anybody has a use for it?

I suppose it might be everso slightly useful to someone training seriously on the flat repeating the same routes often, who has difficulty judging from feel alone erm.. how they feel. 

I run very un-seriously and almost always on a hill, so usually get a score of minus lots gaining altitude (holy crap - you're working really hard but hardly moving!) followed by a score of plus lots coming down again (weeee - this is more like it, flying now!).

In reply to Luke90:

I guess I was comparing it with power meters for bike which all have error specs and calibration and are routinely compared to see if those specs are believable.

Another aspect of 'running power' I find somewhat awkward is that given that running is a factor of 2-3 time less efficient than road cycling, where does most of the power go? Obviously not into moving you forward, so most is 'wasted' (arms and legs pumping?). So how can any running power meter account for the proportion of wasted vs. useful power, given that it's just the 'useful' component that translates into 'performance'? There just seem to be too many variables and error sources provide a meaningful analysis.

In reply to Marek:

And this is why there is a significant difference between Stryd and Garmin power figures- they have different opinions as to how use pace/stride length/ground contact time/elevation and wind data, and there is no way to adjudicate as to the most appropriate calculations. Exactly the opposite of cycling. Unfortunately Garmin need GPS data to provide speed, it effectively stops individuals doing their own calibrations on a treadmill at varying paces and angles, I think Stryd works on a treadmill in the same way a footpod does, so they have no real excuse for not giving a calibration guide.

I use a PowerPod as my cycling power meter- even though it does not have a mechanical connection to the pedals or chainring it does the calculations based on airspeed (pitot pressere), gradient (barometer) and wheel sensor. After a 5 minute calibration ride out and back to learn how it has been mounted it gives figures within 2% of expensive straingauge type meters, and the Wattbike. This is what Garmin should be aiming to do, but the wind component is always going to be a guess based on the local weather forecast at the time of the run (so I disable wind in my Garmin as being one more garbage-in component).

In reply to petemeads:

Interesting that you (and others) are getting good results with this. I thought of making something similar a couple of years back, but couldn't find a suitable low pressure pressure sensor for the pitot so shelved it.