20 seconds, or 20 meters?

---

---

"Anyone can work hard when they want to; Champions do it when they don't."

When the GPS indicated I had ran 1 mile, it still took me 20 seconds to reach the end of my 4th lap.

The track might be a bit off...

Were you running the inside lane, middle, outside?

can you benchmark it against your bike?

Anything with short radius turns can be off IE: the track, trials with switchbacks, out and backs, etc.

---

---

"your horse is too high" - tigerchik

GPS is accurate. The military have been using it for years to blow shit up. The only time it might be sketchy would be if you were trail running under tree cover, but on an open track, and on a clear day, if it is saying that your track is short, then I would bet that your track is short.

Having said that, an earlier comment asked which lane you were running in. Only lane 1 is 400M.

Cheers.

I don't think GPS is accurate enough to zero in on a piece of shit - perhaps if it was a piece of elephant shit (but we'll leave that for another conversation) ;)

---

________________
Adrian in Vancouver

Explain this one: my Forerunner will register a little long on a registered 400m track. How the hell does that happen? I thought it would only register short due to calculating based on a straight line between waypoints.



Don't get me wrong...I love my Forerunner but for tight corners like tracks is will not be accurate. Most of my runs are long and somewhat straight (certainly not the turns of a 400m track) and the accuracy is spot on as a result.

---

________________
Adrian in Vancouver

I believe GPS checks the Satelites every few seconds and does calculations. If it checks at the beginning of a curve and at the end of a curve then it wil calculate a straight line. A straight line is shorter than a curve. This is why GPS wont cant be counted on to be completely accurate. These systems are still the best way to go though.

Here is a good explanation of why GPS is inaccurate.

http://www.peterwiltonjones.co.uk/gps.php#GPSInaccurate

---

---

"your horse is too high" - tigerchik

Used to use a Timex GPS system and on a 400 m track, the thing was never more than 5m off an usually within 2 m (based on hundreds of laps). There's a smoothing feature on the Timex that makes it smoother.

May be our guy ran his mile on a 400 yards track. That would explain the 20 seconds.

I alos used a Garmin60CS (bigger GPS unit) fixed on the baby jogger and it also worked fine on the track.

Francois in Montreal

It's seems that having the Forerunner defeats the purpose of going to the track and running circles.......

If you check out the manual, or Garmin's website, I think they explicitly state that the thing is only accurate to within fifty meters for altitude. Maybe that means it's only accurate within fifty meters horizontally as well. 20 seconds at 6:00 pace is 89 meters which sounds like more than we'd like to see it be off by. However, I've found a few trends with my devices. They tend to track more accurately on my bike. I believe, though don't know, this has something to do with the fact that I'm generally going faster on my bike. It's not as accurate when you are going super slow. To prove this, go outside and turn your device on, once it gets it's signal hit "start" and don't move an inch. You'll see the display shows that you are moving, and maybe after a minute it will think you've moved 20-100 feet though you havn't moved at all.

If you are wearing the device on your wrist while running, your device may have actually moved through a space of one mile, even though your core did not. If you run one mile, your wrists and ankles have actually moved through space farther than one mile because they are swinging back and forth. Maybe this has something to do with it. Sometimes when I ride, I'll put the unit on my helmet if I'm in the mood to just feel a workout through from start to finish and don't want numbers in my mind. As I usually find myself working out in the forest, I've found that running with the GPS on my helmet tends to get me better reception. The only drawback is that most people will be very much self-conscious of running with a helmet on, especially if you're doing a brick and just run with your cycling bibs/jersey on. People have actually stopped and asked if I was OK, which makes me just smile real big and ask them "How do you mean?" It's actually quite comfortable though, and it keeps my head warm when it's cold out.

The idea about the satellite and the GPS communicating every few seconds and that makes the corners of the track end up off of a true running line sounds like that could be the case. I've got the run d/l'd on Motionbased.com, but I'm not sure how to share it with you guys. When you look at route on the Motionbased maps...it looks like i'm all over the place. I will inquire at the high school to see what the exact distance is on the track. Oh yeah, I was running the inside lane. It's a pretty modern track and they regularly have regional track meets there, so I was guessing that 4 laps would equal a mile.

4 laps of a 400M track is 9 meters or 30 feet short of a mile. That might explain part of your error. I would guess the corner rounding is the other part.

The accelerometer based designs are more accurate under the right circumstances.

The horizontal accuracy of a GPS is within 15 meters 95% of the time. If it is WAAS enabled (which is the case for most modern units and assuming you are in NorthAmerica), the accuracy is within 5 meters 95% of the time.

If you stand still, your position will vary more or less by 5 meters.

However, most units (and I assume the Forerunner does that - it would be incredible if it did not) track current speed and calculate current position based on the GPS reading and position estimated on preceding position and current speed. This smoothes everything and results in a tracking much better than a simple interpolation based on multiple GPS readings.

The 20 seconds error over a mile makes no sense to me unless the track is a 400 yards (and not 400 meters) track.

Francois in Montreal

http://trail.motionbased.com/...sodePk.pkValue=12056

The above link is the run at the High School track I was talking about. You should have to log in and then it should bring up my run. I did it as a test and ran 4 laps. I would think that there would be one line around the track and that all four laps, the line would just overlap. Obviously, that's not the case. I think the most deviation is around the corners, so who knows.

I also did a run today down by the river and when I downloaded the run and looked at it through motionbased, it looked perfect. No zigzags off of course. This run is a pretty straight run with a few turns. I ran out and back and the two lines overlapped each other.

Hello,

I am pretty certain that up unitl this years model of the Timex BodyLink, Garmin, the people who make the Forerunner, were the suppliers of the GPS system for Timex. So chances of your Timex being more accurate are slim.

I think that the combo of your track/ route being short AND the Forerunner is not being 100% accurate is probaly your situation.

Both system are pretty DAMN cool when you think that they are only a few years old.

Brian

Could this be the rotation of the earth? In that case the GPS device was actually moving.

---

----------------------------------------------------
Striving to have sex more than 66 times per year

I have to disagree androgynotopia about your comment discussing the extra motion of the wrist due to the back and forth motion. Imagine a person with an invisible suit on with only their arm exposed running across ground. All you would see would be an arm accelerating forward and then decelerating, but the arm would move continuously forward and never back track since the body it is attached to is in a continuous forward motion. It is kind of analagous to swimming in which the hand of champion swimmers actually move foward rather than move backwards during a stroke when they are grabbing the water.

I have no explanation. This makes little sense. It's not even that good in the straights. Tomorrow, I'll go around the block a few times with my unit and I'll upload the data.

Francois in Montreal

Timbuktu, you may be right, but unless your arms are completely stiff they aren't moving directly forward, but instead in an arcing motion which again would mean they are moving through a space of more than one mile. Bernard Lagat is about as smooth a runner as I've seen, but I'm still thinking his wrists move through more than one mile of space for every 1609.xxx around the track. Either way, the visuals I got from mentally employing the invisible body suit that only exposed the arm really made my night! I dig your neuronal pathways man.

(See the pictures from right to left, bottom to top, unless Mr. Lagat really was running backwards this fluidly.)

Damn you are totally reaching. Don't forget that GPS doesn't update every millisecond which neglates your creative BS of arm movement - nice try. Also, since you seem to be stuck on creating wayts to make your non-senese right, please provide a way to measure his arm stroke distance compared to his stride distance because I'm willing to bet his legs traveled a greater overall distance.

My arms move one-for-one with my legs, pivot on the same axis, but are shorter than my legs. Unless you have some freeky way of running, I bet you do the same, and your legs cover more distance over time.

---

---

"your horse is too high" - tigerchik

GPS satellites are geo-stationary. So the answer to your question is no.

---

________________
Adrian in Vancouver

I was being sarcastic by saying that the GPS is so hyper-accurate that it's actually measuring the distance the wrist was moving as opposed to the amount of ground being covered. Going back and reading that I can see the tone wasn't very clear.

I believe the Forerunner 201 samples once per second, though I'm not too sure because it only holds up to 3,000 points, which would only be 50' worth at once per second.

When somebody commented that the whole premise of that facetious statement was wrong, as opposed to just the detail of the GPS tracking being wrong, I thought I'd justify it with some pictures. I think it's pretty clear that if the distance the wrists moved through space is equal to X, then we can safely assume X > 5,280' in this example; how much greater we'll never know.

It doesn't matter if his ankles traveled farther overall through space than his wrists because I don't believe the fellow was wearing the device on his ankles. You are right that the legs/feet cover more distance in space over time than the arms, that's one of the reasons why I'm an advocate of barefoot running for elites that wish to move up to the next level, say from 13:15 for 5,000 to sub 13:00, at least in track and field.

As the picture of Bob Kennedy below clearly shows, the core moves in a sine wave motion, (as does nearly all energy it seems) and therefore it is also moving through a space of more than one mile for every 5,280' of land covered.

I believe the four main reasons the original poster was getting less than 100% accurate data is because he was moving too slow, had his device on a less than optimal point of his body (the head being the best spot), was moving for too short of time at just 7:30, and was using a product that clearly states it's not in fact 100% accurate to begin with.





Photography by Howard Schatz, Athlete