This question is directed to people who use power meters indoors and outdoors.
Specifically mountain biking.
We see people take advantage of the manner by which power meters measure power by engaging in certain pedal techniques.
I would like to put a power meter on my MTB but I am concerned that the manner of riding on my local trails would tend to be a poor environment for accuracy.
I am up and down the whole time.
Hammer and recover.
Here is my HR graph from a 10 mile 1 hr ride
I use this as my “hard” interval day but have no idea about power.
That is more of a Zwift problem than a power meter problem. Power meters are a good purchase for any kind of outdoor riding if you want the data. I put cheap left-side crank arm power meters on two of my bikes and that’s good enough for me. It’s not as accurate as dual sided and doesn’t show L/R but it tells me if I’m better or worse or if perceived exertion matched output.
The issue people have is folks that stand and pedal hard for <5-10 seconds and then sit back down and do nothing, and repeat… which is “an abnormal riding style”, and not something anyone would do outdoors… Spin class maybe??? but outdoors, no.
Either way, unless you’re doing that and then also producing world-class kind of numbers and taking wins left and right until A+, most people won’t bat an eye.
As said before, your use case wouldn’t be doing what some folks are doing for 30-60min, such as
600 watts for 10 seconds > 0 watts coast for short time > 600 watts for 10 seconds > repeat
I had a video recording of one guy doing exactly that some time ago - it was very clear that he was using this loophole to maintain a higher than normal speed.
Power meters for outside riding are very useful. I have a Quarq on my Cervelo S5 (the quark PM chain ring). Indoors I have Kickr Bike and Favero Assioma Duo shimano pedals - the latter only really to track my power balance.
Thanks for all the replies
I’m still not convinced my MTB power will be accurate.
I have Assioma Duo Shi on my road bike and switch to my trainer in the Winter.
If I accelerated, the power reading always seems to shoot up then stabilize at a lower reading.
The folks that are cheating on Zwift are taking advantage of the momentary overshoot.
This probably won’t affect real life rides because I can’t start and stop until I get to the top.
I am going to do a MTB ride with my Assiomas but I was waiting until I got my MTB legs back.
Zwift is good for fitness but not so much for skills, especially when HR is near max.
I’m still unexpectedly having to put my foot down too much and the Shimano road cleats are kind of clunky on trails.
Do the crank based seem to fluctuate less than pedals?
I like playing with data.
Here’s 2 rides from the same week.
One is a 1 hr 20 mile spirited group ride taking pulls at the front and the other is the 1 hr 10 mile MTB ride where I get dropped.
I will try to structure workouts this winter to emulate some of these efforts.
Sticky Watts in Zwift comes from how the game continues to record the last known power for a few seconds after the rider stops pedaling. You can compare Zwift vs your head unit by using that pedaling style and dual recording. In Zwift you should see a short flat top on your power reading when you go from pedaling to coasting. Do you see that in the dual recording on the head unit? You should not.
I’m not really concerned about sticky watts rather exaggerated watts.
When I stand up and push hard on the pedals, it is common for the power to momentarily spike up to 400-500 + then come down.
I think the initial force to overcome the inertia of the wheel seems exaggerated.
I think you’re overly concerned over still something unlikely to be an issue.
Increased watts when you stand is normal. Standing and sitting every 4 seconds for an hour or more is… what’s not normal; particularly indoors / on a roadbike, on the road.
Riding an MTB on trails is not comparable; you’re constantly shifting weight around. This is not the case indoors/on the road.
Dual recording the Assiomas on your MTB will be a good telltale of where things are though; in the meantime, there’s no real concern / anything to be done.
Until you do that test, it’s impossible to know how accurate or not your power from the MTB will be, and all this discussion would be for naught.
If it turns out it’s way off, then sure, it’d be worth diving into to resolve for your own use.
Power meter readings are generally stochastic meaning that values can vary dramatically from second to second. This is completely normal and represents what is actually happening. If you think of a pedal stroke you have more power in the main part of the down stroke, and it varies instant by instant through that stroke. A powermeter is sampling this data and reporting it once per second (or 10 times per second in some cases) and this intrinsically adds variation from sample to sample reflecting the nature of pedaling. This raw data is made more useful by tracking over a long period of time and by averaging. Erg mode helps to smooth out the power since the smart trainer itself is holding you to the power target.
Despite the variation, power data has proved to be extremely useful over many cycling disciplines. Mountain biking is more variable than road cycling. Mountain biking tends to be on/off and rapid variation in cadence, while road cycling is generally smoother. Mountain bike racing adds even more variation because of hard efforts on any climbs. Still the power data reflects what is going on.
I would consider what you are going to use the power data for. IMHO using it to track your time in power zones, your FTP and your kJ/TSS are valuable - these focus on the power over the whole ride, not second by second. You can also use for intervals or as part of a structured training program. Interpreting a power chart for a ride is not easy - it’s going to look very squiggly and you have to look at averages over a period of time and metrics like normalized power to get something out of it. In my experience you need tools like intervals.icu or TrainerRoad to get value from the raw data.
Another good use case is to pace yourself according to the power on the head unit. You can use 3s or 30s or lap average power for this. It is a real eye opener when you see how much your numbers vary, particularly when mountain biking. You can use this feedback to become smoother and burn less matches so you are less likely to get dropped. You will probably also see that everyone goes hard on the climbs, but then backs off dramatically on the crests and false flats. You can use that to your advantage to close gaps with less overall effort.
Hope this helps (apologies if all this is obvious).
For these events it was basically unstructured training. I was using Zwift races for intensity, and long weekend bike rides for endurance. There was a series of 1 hour XC races that I used as prep for the long Sea Otter race. I also did a lot of course pre-rides at 50-70% race length. Volume varied a lot and wasn’t ideal. I think I could easily take 30 minutes off that 6.5 hour race time if I was to do it again.
None of the above is optimal by the way - if I was properly serious about this, I would be following a structured training plan. If you are interested in structured training take a look at the TrainerRoad podcasts on YouTube.
Yeah, no drafting in the 1 hour race - I was holding a gap of 5-10 seconds ahead of my main rival/friend for almost the whole race. There were opportunities for drafting in the Sea Otter race, but not as much as you would think due to field spread. With my “slow start” strategy I ended up catching and passing a lot of riders, but of course you only catch the ones going slower than you and then drafting a slower rider doesn’t help so much. And the benefits of drafting on single-track are not as significant.
