City Guy

Hello all,

I recently switched to using an Elemnt Bolt on my road bike. Along with the Wahoo Elemnt iPhone app I have been very pleased with the unit and app. On my road bike I have a cadence sensor which paired successfully with the Bolt. I also configured the Bolt ride data to auto upload to Strava. However I have noticed an issue and wonder if any other Elemnt Bolt users have experienced this same anomaly.

Reviewing post ride data on the Elemnt app, I noticed that my average cadence during a recent ride recorded on the Bolt seemed low with a value around 62 rpm. What makes this confounding is that the average cadence shown on Strava for that same ride shows a value of 76 rpm, which based on my experience seemed about right. I reviewed the data from several recent rides comparing the average cadence listed on the Elemnt and Strava. In all cases, the Ement Bolt average cadence rpm value is lower by about 15 rpm.

So, anyone have ideas why the discrepancy exists? I'm not an expert on the data collection and analysis by these sensors, but it seems to me that both the Elemnt app and Strava use the same raw data. Ideas, opinions, thoughts, suggestions welcome and encouraged. I suppose the next question could be, which value is correct?

Thanks in advance...

Marcus_Ti

The problem with any "average" cadence...is exactly how you calculate it. What average? Mean? Median (which median??)? Mode? Zeros excluded or included? Etc.

Strava...I've found, natively, will distort your numbers and make you feel better because they are higher. The exception being if you're using Strava with the "Elevate" AKA "Stravistix extension that gives you detailed readouts and actually tells you what it is doing mathematically. I remember the first time I loaded it, Strava said my "average cadence" was 90---well the actual 75% Median cadence was closer to 60, because I was pedaling and then coasting. Lots.

MonsterFat

I have the same issue. The Bolt shows one avg. cadence and Strava a higher average. Golden Cheetah gives similar numbers as Strava. Whether I set the Wahoo app to include/exclude zeros, I get the same number. Not sure if the calculation is done by the app or the Bolt?

DrIsotope

Sounds to me like the Elemnt companion app is counting zeros. Strava does not count zeros for cadence. Strava also runs literally everything through it's own algorithms, so the end numbers there are often different from the numbers that went in.

With regards to cadence, I've never had the Strava number differ from the device-recorded number by more than 1rpm, including the Bolt.

zacster

It really depends on the calculation method. If the Elemnt excludes only "dead" zeros, then it will have a lower average than something that stops counting below a certain cadence. Sampling interval will also change results, so if it waits for 15 seconds vs. 30 seconds to get to zero one will include more or fewer turns of the crank.

What you need to do is to accept one method and set a baseline and then measure using the same method against this baseline. When it comes to time, power, HR, cadence, calories, you get different results with every device. Miles are miles though, so you can always just use that.

MonsterFat

Don't disagree with your point, however I would expect all the features to work as advertised when paying $250 ish for a device. Since I don't coast on a trainer, the cadence data there is accurate. There-in creates a challenge with your baseline approach. It's impossible to go back and forth between trainer and road and compare. As I am trying to improve my cadence, it is frustrating.

DrIsotope

I guess now we ask what cadence sensor you're using. My cadence comes off of my PMs, and there have never been any issues. My wife uses a Garmin magnetless cadence sensor, never been an issue. I doubt the Wahoo sensors would have a problem, matching the Wahoo computer and all.

Iride01 

On Garmin's, they have a setting to set whether or not you want to include or exclude zeroes from your cadence averaging. Does Wahoo have similar and have you set it up accordingly?

Be wary of any time you compare device data to what you see on a website. Websites may or may not show the accumulated data as the device does and will use other information from the data file to come up with their own idea of what you did. Many times they agree, but for some they don't.

City Guy

Greetings

Thank you all for the info. I'm using a Garmin cadence sensor which magnet-less. Include zeroes option on Bolt is set to off. Todays ride, 2:14:32, Bolt 67 rpm, Strava 80 rpm.

dwing

Theres an option to inoclude/exclude zero in my android bolt app. I dont normally look at the wahoo history but i checked my last few rides and cadence matches strava. I exclude zero cadence. Geez I forgot how nice wahoo ride history layout is... the graphs provide some good info with option to drag the timeline.

zacster

But all of whose features work as advertised? All of the features work on my Bolt. I use either my PM pedals or a Wahoo sensor depending on the bike and they are always accurate. I also can feel cadence without looking at the computer as I've had computers with cadence for over 30 years. I don't need to see it. But what it would say on RidewithGPS or Strava would be different.

As for improving cadence, the only thing that matters is cadence while you are steady pedaling. It doesn't matter what the average is as it includes all the non-zero data where you've let up. On a trainer you never let up, or at least I don't, but on the road there is always stop and go. You can't compare the two, all you can do is get a trainer baseline and improve on that on the trainer, and a road baseline and do the same there. Maybe once you've established both baselines you can track the two in parallel. I use a trainer all winter long and what it tells me doesn't equate to what I do on the road, not power, not cadence, not climbing and certainly not miles. Just treat them differently and you won't get frustrated.

drewguy

Not clear what "feature" isn't working, unless you define as a "feature" consistent treatment of data by other apps to which you have exported data collected by Wahoo. Not sure that's within the scope of what you're getting for your money, although I agree it would be nice.

MonsterFat

I probably wasn't clear. My unit does not seem to exclude zeroes in it's calculation. I have tried turning that feature on and off and yet the average cadence is within 3-5% of each other. As a comparison, 2 other different programs return results that can be as much as 30% different than the Bolt and more consistent with each other and what I would expect. Since average cadence calculated with our without zeroes is a features of the Bolt, I would contend that features does not work.

John_V 

Something that you may be overlooking in your claim that the Bolt's feature isn't working correctly or to your satisfaction is this: The Bolt calculates your average cadence in real time, whereas a website, such as Strava, uses data from an uploaded file and uses their own algorithm to calculate what they display on your ride page. And I really do think that Strava, and some other sites, exaggerate their data. In any case, I would tend to go with the real time data rather than calculated data.

An example: Yesterday's ride my Bolt showed an average cadence of 54 rpm. I'm not a high cadence rider so that about sounds right. I send my rides to Strava and RWGPS to share with my friends. However, I also store all my rides on my phone using Cyclemeter because the database is local to the phone, the statistics are better than any website and I don't have to have a WiFi or cellular connection to see my stats. With the ride above, the average cadence on Strava was 65, on RWGPS it was 64.8 (how they got .8 is beyond me) and on Cyclemeter it was also 54. So maybe the feature on your Bolt is working properly and it's the websites that are off.

MonsterFat

Thanks for your comments. Try this as an experiment. Find a short hill, start at either the top or bottom. Start your Wahoo and pedal to the top at a constant cadence. Then turn around and coast to the bottom(or vice versa). If zeros are excluded, your avg. cadence you be whatever your cadence was climbing. If I do as described with a climbing cadence of 100 and a coasting cadence of zero, my average cadence will show about 50. Either my understanding of exclude zeros is wrong or there is a bug. Wahoo support is looking at this to try and figure it out.

John_V 

Well, here is another test you do, providing that your cadence sensor is dual band (Ant+ and BTLE). Install a cycling app on your phone that allows sensor connections without having to upgrade to a premium version. Connect the cadence sensor to the phone app via BTLE and to the Bolt via Ant+. Run the app along with the Bolt on a ride and see if there is a big difference in the cadence reading between the two.

I have noticed the issue with Strava having different values on ride metrics even before I got my Bolt and regardless of what file format was used. Before I got my Bolt, I used Cyclemeter as my cycling computer along with a Wahoo RFLKT as a display. Cyclemeter sent it's files directly to Strava in KML format and there would always be a difference in something or another. Mileage distance was the biggest difference that I saw. Strava would show less miles than what I actually rode.

On very long and multi-day rides, I keep my wife informed of where I'm at by using live tracking. I don't use the Bolt for that as it runs my phone battery down in no time. Instead, I use my Cyclemeter app and send my location ever 20 miles. I have my Bolt and Cyclemeter connected to my sensors, as stated above, and have only noticed very minor differences between data from the Bolt and Cyclemeter.

Whatever your problem turns out to be, I hope you get it corrected soon. If it is the Bolt, Wahoo is great at customer service and will do right by it.

Iride01 

Try changing the setting on the bolt to the opposite of what you have it now and see what it does. Perhaps their verbage can be read to have opposite meanings to different people.

If that doesn't work, then set it back to what you had it at. I have seen on some rare occasions where one of the settings on my various garmin devices had to be toggled to get it to correctly recognize the setting.

u235

I have the full sized Elemnt. Spot checked my Strava vs Elemnt results and my avg cadence for the last 15 rides is reported 0-3 high and the max cadence is just about the same or +1 on Strava. One exception is on one ride the Elemnt reported a 165 max, Strava has 110. This is a wide variety of rides of various times and miles including single track, mixed gravel, and pure road. I have the option on the Elemnt to ignore zeros. It seems to work for me. Maybe your sensor is dropping out or erratic and Strava and Wahoo app have different error cleaning of the raw data resulting in different results.