So are you saying that the gps and your computer are not even close in information? Speed, distance, time moving, stopped, etc?

Well, if you live in a flat desert, then they're probably pretty close, but in a forested, mountainous area or in an area with multi-storey buildings, the GPS is likely to be way off.

GPS is fine for navigation, but it's far from a precision tool relative to a bike computer.

I've found when I got the cyclometer calibrated, it's within 1 percent of the GPS and I live in South Carolina on mountain bike tires. If I get lazy and let the air pressure down in the tires, the cyclometer will be off up to 5 percent on the same rides. It's not the GPS that has changed, it's the accuracy of the cyclometer due to lower air pressure and tire wear. Once I calibrate the cyclometer again, it's within 1 percent of the GPS.

Two notes.

1) You got to property calibrate your cyclometer. That means measuring the tire distance and not just using a chart. And especially on thick tires, you'll have to calibrate every so often.

2) If you are in the dense woods, a canyon, or in a major downtown area, a GPS will not be nearly as accurate as a properly calibrated cyclometer. I wouldn't say way off, in downtown Baltimore this year, I estimated it was off by 3 percent after I plotted the course using mapping software.

There's no reason not to have both considering how cheap a cyclometer is. They compliment each other, not contradict each other. The one thing a GPS is when you follow the same ride is the consistancy. I always know when I am going to pass 3 miles on my ride (stop sign)... 5 miles on my ride (last mailbox before turning right to Lake) etc.

I consider 1% way off! 3% is not even in the ballpark. My bike computers are at ~0.3% (compared to true distances from a digital map released by our national mapping service) at worst, and usually better. And even then, you're getting good numbers for a GPS. Last time I walked through a built-up area with a GPS unit, it regularly "teleported" me 200-300 ft in a random direction, simply because GPS doesn't work well when you don't have an open sky around you.

Garmin thought of this with the Edge series of their GPS enhanced training computers. The cadence/speed sensor option on the Edge 305 (Edge 305 + CAD or the Edge 305 Bundle) enhances accuracy by filling in data missed by the GPS. The Edge 305 also uses a barometric altimeter.

I've never been teleported 300 feet in another direction in a GPS in six years of using unless I'm in a deep bridge or tunnel. However, do you know what a 1 percent difference is? If your real speed is 10MPH a one percent difference is 10.1 MPH. Or on a 100 mile ride, it's 1 mile off. My GPS is well within 1 percent because I know almost exactly when it's going to reach 3.00 miles right around the stop sign. Considering that I don't ride in a perfect straight line, it varies by about 30 to 50 feet.

And what national map do you use? I own NGS maps, Delorme maps, Delorme Topo, and Mapsource national maps.

All of the Garmin handheld units will fill in any data missing by the GPS, it's not just unique to the Edge series.

I'll upload later today 100 rides I have taken along the same road that is covered in forest land since I can merge them all together. Out of the 100 miles I have ridden, not once has it varied by more than 50 feet.

Current GPSRs that are waas enabled with SiRF III are 99% accurate. I've been using a Garmin Forerunner 305 5-6x week since december and the data is always reliable. My previous Forrunner 201 without SiRF III chipset had nowhere near this accuracy as it consistantly lost signal in tree cover. Same deal for my old magellan 315 handheld.

For mapping (after the trip with a computer), the Suunto X9i is the only choice I know of. It collects and displays GPS waypoint data. The problem is, it's difficult to pack enough battery capacity in such a small space for the GPS. From what I've heard, in GPS use one needs to recharge the X9i at least once a day to keep it running. I believe Timex used to have similar "all inclusive" product, but I'm not sure if it's still in production.

Suunto, Polar and other manufacturers also have wristop watches that get GPS data wirelessly from a separate GPS receiver. The receiver has its own batteries, making the capacity problem that much easier to deal with. None of these watches (that I've seen) actually display the GPS position data, they just use it to calculate speed and distance. I don't know if one can download the full data from the receiver to a PC to map it as with the X9i.

If I needed HRM as well as speed/distance info, I might be tempted to look at these more carefully. As it is, I have a separate mapping GPS unit (Magellan) and a bike comp.

--J

You can go on and on how much you want about GPS accuracy, but the fact still is that it is nowhere near the accuracy of a calibrated bike computer. No matter how many acronyms or initialisms you throw at it.

But then, I'm pretty anal about having my computers as accurate as possible.

gpsblake, as my profile will tell you, I'm not in the US. What national map I use is irrelevant to you.

This has been a very helpful thread. I'm getting a new bike and was trying to decide whether I should hop on the GPS bandwagon or stick with a traditional computer. I was thinking that going with something like the Forerunner 305 would be useful since I also run a bit. However, after reading about issues with accuracy, reception in forests (which is where I do most of my running), and shorter battery life, I've decided that a traditional computer, with heart rate monitor, will suit me just fine for now.

I have the Edge 305 with the cadence/speed sensor and HRM. When I have the cadence/speed sensor working properly, the speed seems very accurate compared with my bike computer (a cheap one I still have mounted). A few weeks ago, I had a problem with the cadence/speed sensor, and the first thing I notice was inaccurate speeds when under heavy tree cover. Once I adjusted the speed magnet to where it was reading again, the accuracy is MUCH improved again. I guess the speed sensor fills in data when GPS coverage is poor, thus improving accuracy somewhat.