Garmin 405 not measuring correctly?

Did you ever go to grad school and do research? I made a hypothisis and proved it to be true.

The results you cite as skewed are real and unaltered.

Lastly, I clearly acknowlegded the usual difference in arm swing between running and walking, BUT, I also acknowledged that running adds a tangent variable. That is why I stated that you should drive the course to eliminate both variables. That is when you chimmed in and said the arm swing would not make a difference.

Your "hypothisis" did not include walking "very very slowly". Nor did it include covering only a distance that was barely twice the accuracy of the unit itself. Both of those were your additions after the fact, intended to skew the results.

Walk a course, a real course, normally with a Garmin on your wrist. That was what you spoke of originally.

The difference is less than the margin of error of the device.

Brian,

I am by no means complaining and I love my Garmin. I have several. The inquiry is absolutely a simple curiousity and my persistance is only an attempt to point out some variables that exist when using these devices in addition to the inherent degree of error of the device. The better we understand what affect these devices, the better we understand varied results when they occur.

This is all beside the point. So long as the subsequent reading is taken further along the forward route than the previous one, arm swing makes no difference at all - it doesn't need to be cancelled out. The GPS receiver could go to France and back so long as it is ahead of the previous reading along the route by the time it takes the next reading. All that has to happen is that a straight line through all the frozen points would equal the same distance the torso traveled and it would. It doesn't matter what it does between readings or that it is parallel with the torso on the next reading - it is still moving forward along the same route. I will buy that it would make a difference on the very last reading. We know this is true becaus ethe watch is keeping up with the body - you aren't leaving it behind which would have to happen if error accumulated. This is so obvious to me now that I am ashamed I dithered at all. I am just not smart enough to explain it.

So you were running these repeats on a 400m track or a 440y track? If you ran 4-lap repeats on a 400m track then your GPS is measuring .994 mile as 1.00-1.02 miles, actually.

good enough for me wrote:

this stuff is crazy.

i use my garmin for mile repeats on the track, and it consistently reports 1.00-1.02 miles.

OMG! Just do the tests and get back to us.

Brian wrote:

Although, I will say that I call bullshit on your saying you got that kind of accuracy on a track. They are notoriously inaccurate on tracks - even Garmin says so. This is well documented and explained on the Garmin messageboards by both users and Garmin reps. see an earlier post for a GPS tracking map done on a track.

i have read the posts. maybe my tracks are in special areas that the government have more satellites over. you can call bullshit all you want, but i can direct you to my garmin connect pages if you want. why would i lie about that?

they might not be accurate for 100m or 400m repeats, but at a mile they are pretty good. the variance balances out the longer you go.

Oh lord, scanning the Garmin boards, I have discovered another wrench in the works to a real world test - your body blocking access to satellites. A torso mounted unit would have access to fewer satellites, says the Garmin folks. Wrist mounting has that problem as well, but to a lower extent. We need an antenna on a hat.

Either way, I think the Garmin is a pretty damn good tool if you keep in mind that it's accuracy varies with many conditions. I do really think my dog-walking several posts up example proves the point. Picture the accumulation of points on the floor where the dog stops. If the watch is just ahead of your torso on one reading and just behind on the other, there is no cumulative effect of that error so long as the entire system has progressed forward.

.

lost in translation wrote:

So you were running these repeats on a 400m track or a 440y track? If you ran 4-lap repeats on a 400m track then your GPS is measuring .994 mile as 1.00-1.02 miles, actually.

yup, i start behind the line. mile repeats = 1609 meters. thanks for playing.

Well, I'll take your word for it, but I have done 10+ milers on the track and I lose at very least 50m per mile with my Garmin and others have similar experiences to the point where Garmin has a disclaimer about use on tracks. I have never had a mile on the track measure even close to accurate - the inconsistency is always about .025-.03 per mile. And the track, unlike road courses, has lots of variation between miles, whereas all my road courses always measure out the same if I keep a smooth pace without speeding up and slowing down (which always causes overestimation of the distance, for some reason).

I wouldn't accept .02 error for a track workout. That is 32m. 6-7 seconds at tempo effort. If I am doing time sensitive workouts, that is an eternity - that is the difference between being fit and being in marginal shape. An eternity in an interval workout where time matters. You wouldn't see that kind of error on straight roads. It isn't a problem of satellites, it is a problem of trying to draw curves using straight lines and inaccurate left-to-right placement of each point compounding the problem of curvature.

Behind which line?

i start about 9m behind the line. it is actually marked on the track we use.

if you are actually using the gps to accurately pace track workouts, you are a moron. you should obviously use the more accurate measurement, which is the track itself, which is what i do. so, i agree with you, i don't accept a 2% variation for my track workouts. i only mentioned the gps measurements because i thought it was an helpful in terms of commenting on accuracy.

if you need that much accuracy for your tempo workouts on the roads, god bless you. you are running them at 5:00 to 5:52 apparently. being a few seconds off on your tempo pace isn't going to make or break your running career. and remember, as you and others have said, the accuracy improves off the track and when there are fewer curves in the course.

Try telling that to Joe Vigil, buffoon.

good enough for me wrote:

if you need that much accuracy for your tempo workouts on the roads, god bless you. you are running them at 5:00 to 5:52 apparently. being a few seconds off on your tempo pace isn't going to make or break your running career.

Actually Joe Vigil would agree with him.

Of course I don't need that much accuracy - I ran just fine without a GPS for years. But why would I buy a GPS if I didn't want some accuracy. If one worries over how far an easy run is (and the pacae thereof) but doesn't care about 10-15 seconds/mile on their workout days, I'd say they are straining a gnat and swallowing a camel, to use the biblical phrase. I don't even keep a log or keep up wiht mileage, but some ballpark accuracy on workouts are pretty important to me. For me, workouts away from the track are all the GPS is good for. And, thankfully, the accuracy is pretty good. Remember, I am arguing on the side of defending the accuracy, not slamming it, at least as it regards arm swing.

I would bet on a straight road wiht no turns and a good view of the sky, you wouldn't lose more than a few meters per mile over what a jones counter gives you.

What surprises me is the uncanny consistency of the GPS from run to run on the same course. I swear I have had it not deviate by more than .01 over severel measurements of the same run on several courses. You'd think just taking turns differently would cause that much.

CTZ wrote:

Try telling that to Joe Vigil, buffoon.

clearly you have raised the level of debate with your epithet. thank you for your enlightened and superior input.

Yeah, because this is such an elucidating discussion of high-brow proportions. And whomever believes that Vigil would agree that a 50" pace range for LT runs is fine and dandy doesn't know shit.

Anyone who thinks he actually meant a 50 second variation needs to work on their reading comprehension. He was talking about a 2% variation. That's, at most 7 seconds. The 5:00 was OBVIOUSLY a typo, genius.

And anyone who thinks that Joe Vigil would believe "being a few seconds off on your tempo pace" could "make or break your running career," needs to learn a little about Joe Vigil.

Um, what exactly does 5:00-5:52 mean then? Because I comprehend that to mean a 52" pace range. If it was a typo, then correct it. I'd have to be a genius to read minds and know when a numerical value is or is not a typo. I'm not going to sit here and try to presume what people mean versus what they sloppily spew across a message board.

Oh, and what I said about Vigil? I got directly from his mouth, it's the basis of his entire training program. Maybe you just need to learn more than just "a little about Joe Vigil."