Calculations
From MB Wiki
| Table of contents |
One of the most frequently and passionately asked questions is ... Why is the distance on my Forerunner different than the distance at MotionBased. Here it is:
Track Log Resolution
Your Garmin GPS unit communicates with the satellites in the sky every second to get an understanding of the location. The data collected from this is called a Trackpoint and contains the latitude, longitude, elevation and timestamp. The unit will selectively place each trackpoint in the Track Log. Selectively means it does not place a trackpoint every second, but maybe every 3 or 5 or 10 or 16 seconds. That is what Garmin calls the Auto resolution, or Smart Recording for the Track log.
Auto resolution is an advantage for MotionBased since it doesn't have to store and process all this extra information that may come with a fixed resolution of 1 second. High resolution can produce as much as 3600 trackpoints per hour where in contrast Garmin Smart Recording may report as few as 500 trackpoints per hour. This is similar to image sizes of a JPG vs. a BMP . . . JPG uses smart reduction to make the image smaller without losing most of the quality of the image.
So Smart Recording will actually place more trackpoints along a turn (when you are changing course) than when you are moving in a straight line. The Smart Recording GPS will drop more points during acceleration or other moments of change. In this manner they are able to report the trackpoints that matter and let a calculating system (like MotionBased) assume that no significant changes were made in the time no trackpoint was reported.
Discrepancies
The discrepancies between MB and any device will be a problem that may haunt us for a long time.
Traditional Garmin devices never reported distance, time or any other cumulative calculations of your activity. So MotionBased was required to do all the calculations given a series of known positions and times that you were observed to be there.
Now the Forerunner and Edge series provide this summary information and we are faced with using the information the device reported or the information MotionBased calculated. We could easily show in the Dashboard and Reports the distance your device said you traveled and that may be enough. But MB takes activity analysis another step. The Map Player and Analyzer must do their own calculations. We have to account for your cumulative distance, elevation gain/loss every step along the way. So if there is a big discrepancy between the device summary calculations and MB calculations it would be difficult to account for those differences when exposing the details.
Now the big question: Which is correct? The device or MB? It honestly depends on the situation. The device is talking to the sky during the entire activity and has better resolution from which to calculate. The device reports less resolution to MB than it observes so there could be some "missing information" that could account for the 1% discrepancy between your device and MB. I have had the opportunity to ride and run with multiple GPS units of the exact same brand side-by-side and a 1% discrepancy is quite good even between the devices.
Corrections: GPS is not perfect and the devices can report some rather upsetting information at times. So MB has implemented the corrections algorithm that helps determine your signal quality and clean your GPS data from erroneous reporting. The same problems have been observed with all Heart Rate Monitors.
MotionBased plans to work with all GPS devices as they become suitable for training. If we start reporting the distances based on the device, we now start to lose the standard upon which all devices can be compared . . . the MB calculations and corrections. I once observed my GPS travelling at 2000 miles per hour for 400 miles across the Colorado state line and back into the deep canyon I was rafting (in Utah). I truly hope that never happens to you, but if it does, I would rather not accept erroneous summary information from a GPS device and start providing a standard that you can rely on.
In the future, we will be using your own historical information to make sure your data is consistent . . . and other's information from the TrailNetwork as they have travelled on your same path. We will be correcting your elevation information with known elevation from the USGS and the TrailNetwork.
Many may feel that by exposing the quality concerns with GPS data that that would be shooting ourselves in the foot. Perhaps, but it is hard to argue that any other technology could get you so close to the true distance that GPS can, anywhere in the world doing any outdoor activity . . . and give you a map too. Bike wheel magnets can be offset, running pods on your feet need to be calibrated on the surface you are running. You can't drive your car on the park trail to see the distance you ran.
So at the end of the day, the best advice we can give is to use your GPS unit as a guide, not as the absolute definition of what happened. The same thing should go for your bike computer or any other device that is not a professional survey tool. We hope that we can make training more fun as a bonus!
Just know that your data is being stored at MotionBased and as our technology improves your data will automatically improve too.
GPS Accuracy
GPS is not perfect and it can be incorrect about the location. The GPS unit knows how accurate each trackpoint is since the unit knows how many satellites are connected and how strong of a signal each satellite is broadcasting. The trackpoint could be improved if the accuracy of the location was reported with the trackpoint to help a calculating system understand the confidence of each point.
