Home General Training Discussions

Measuring OWS with Garmin

I know my Garmin 910xt routinely measures long on my Runs vs.  Most courses and all OWS files.... A recent 5k run test on a track measured 3.3 miles vs. the exact 12.5 laps of the 5k on the track.... Usually my garmin measure an extra 1-2 tenths on any 5k-marathon course....

This got me thinking and into a discussion with Steve West at Quassy... He suggested since I swim the same point to point distance all the time that I measure it with google maps.... This am I did just that....

My go to point to point swim measures 887yds via Google Maps...

A sampling of 10 intervals from this point to point swim averaged out to 943 yds (all between 924-965)

The difference on average is 56 yds or 266 yds over the course of an IM swim.... Very significant indeed !  This is almost 4min for me....

Now I realize I measured point to point direct and I never swim an exact straight line.... If I threw out the high and low from my 10 samples of swim the difference is only 958-927=31yds .... So I believe that my average variability in my ability to swim a straight line on this course is only about 30yds.... But that is the best case scenario... And still shows that Garmin way overstates the distance !  

Comments

  • Tim - this sort of stuff can drive you crazy. But since today is a rainy one in the mountains, I've got some time on my hands, so I reviewed what I already know about Garmins, Google Maps, and Open Water Swims. I too have a go to OWS locale, been doing it since before I had a Garmin or there was a Google Maps. So I have years (maybe 80-90 swims total) of data swimming in this water hole, as well as a couple dozen IM OWS, and my personal sense of distance honed from 55 years of competitive and recreational swimming laps.

    My conclusion: Garmin measures long (duh!) due to the inherent "dither" built into the GPS signals so they can;t be used for precise work by us amateurs. But (this might be news), Google Maps might also a poor source for precise point-to-point measurements. Look at this recent file from my Garmin in Horseshoe Lake:

    https://connect.garmin.com/modern/activity/787568099 

    Notice at the end of each "length" of the lake, I am shown as crawling out of the water and "swimming"  for maybe 10-15 meters, as if I am some sort of lungfish trying out my fins as feet. Either the Garmin is whack, or Google Maps is distorted in some way, or a little of both.

    I could go into the arithmetic here, but my conclusions are that, measuring point-to-point Google Maps thinks the lake is 595-600 meters long; Garmin thinks it's 710-720m; comparing my times in it to my IM swim times (there also the question of how accurately those courses are measured) and my own innate sense of distance, I'm swimming 620-640 meters. Part of that last range is due to the natural shrinkage of the lake from spring to fall. If my "wandering" is worth 20-40 meters/length, then maybe Google Maps does produce an accurate measurement of the point to point distance.

    The most important thing, IMO, is that I have a consistent course I can compare from OWS to OWS, and that I can use to compare to actual racing. I know my time for six lengths of the lake, including resting at the "wall" each length for about 30", is a pretty good proxy for my IM swim split. In this way, I am able to bypass the inherent inaccuracies of Google and Garmin in figuring out the meaning of the workouts I do in that lake.

    There's another option if you know your distance per stroke. I count every stroke I swim, in pools or OWS. It takes me 560-600 strokes (280-300 the way Garmin counts) to do one length of the lake, another means I used to triangulate the actual distance.

     [BTW, I'm guessing 4.66 lengths of your lake = 2.4 miles.]

  • I have always been hoping that the garmin swim distance was a little off.
    Last year at Lake Placid it had my swim distance as 2.68 miles...but the overall time was around what I expected. https://connect.garmin.com/modern/activity/551274799
    I could have swam straighter and plan to in the future but was a little all over the lake but not as bad as this mapping shows.

    Even yesterday swimming in a local lake after work it had me all over the place, when I thought I was swimming in a somewhat straight line. ON the west side of the lake there is a floating channel marker that I kept swimming back to the EXACT same spot right next to it, but the garmin had me in a few places in the general area of it. And a couple of times when just floating in-between laps in the lake it would loose signal if I did not have my hand close enough to the water surface.
    Could it be with the loosing of signal,? or how often it records a measuring point from the GPS signal? or a combination of both?. https://connect.garmin.com/modern/activity/799399892

    This time interval between recording data points changes from mode to mode in the programming, or at least that is what I have thought. Which explains why it does not know what to do when running around smooth curves on a track, it just measures a bunch of tangents and even assumptions on where the next data point is. I am just guessing but suspect that while our gps survey equipment is very precise, the instrument is also set up in a recording point very still .. not moving while recording positional points.

    Just my guess, but would love to see what others have on this.
  • Al - your weath of knowledge and incredible insight has been invaluable to my learning process as a triathlete.

    Thank you.
  • Would the watch location make a difference - wrist vs. back of swim cap? Maybe going underwater even for a moment may cause some signal loss. I seem to recall with signal loss it tries to guess where you went by connecting known way points.

    I also spent 2 minutes on the Garmin forms and searched on gps accuracy.  Looks like some models are better than others.  Most of their watches have gps and some also have  "GLONASS", From the website:  "Global Navigation Satellite System,  is a Russian satellite-based navigation system that works alongside GPS (Global Positioning System) to provide position information to compatible devices. With an additional 24 satellites to utilize, GLONASS compatible receivers can acquire satellites up to 20% faster than devices that rely on GPS alone."  The Garmin 510 has it, the 910XT does not.  Most of the users said using both systems made the computers way more accurate.

    Anyway, swim straight and don't lose too much sleep over it image

Sign In or Register to comment.