Thanks to a recent story in the New York Times, GPS watch accuracy has been in the news a bit and has caused quite a stir in the blogosphere. Any experienced runner who has used a GPS watch can tell you that they're not perfectly accurate. Granted, as the NY Times article points out (but fails to get into any real details) some of these watches are more accurate than others. I used to have a Timex watch with a GPS receiver on an armband that is put to shame by my Garmin Forerunner 305.
But the question remains, why spend $200 on a GPS watch if it's not even 100% accurate?
Of course, you're not going to realize the inaccuracies of your GPS unless you wear it during a race. As DC Rainmaker has pointed out, races are measured by running the tangents. The actual race distance is the shortest distance you can legally run the course. This is pretty much impossible to do when running in a crowd of seven or eight thousand of your closest friends. So when you run a marathon, you might get a distance of 26.5, 26.8 or even over 27 miles. Even in smaller races you might find it tough to run the tangents perfectly, because...well, it's a tough discipline to master, mostly because you can't really safely practice it.
On the other hand, if you are running trails you might find that your watch is giving you a shorter distance than the race claims. At the Finger Lakes Fifties 25K in 2010, my Garmin told me I had run 16.2 miles - but the actual race distance is 16.5 (which is still greater than 25K - it's one of the liberties that comes with putting on a trail race). You see, GPS devices measure distance by marking your location at a given point in time, then marking it again a second later. So a course with a lot of curves will have corners cut and if you're in a thickly wooded area you might lose reception altogether which could cut a quarter-mile section down by half!
In the image above, the black line is a portion of the course for a hypothetical trail race. The red line is the course that a GPS is recording. You can notice several corners being cut, while others seem to land dead-on. The course-cutting is a very small percentage of the distance, but this might just be a few tenths of a mile of a given course. Multiply this out into a ten, twenty or fifty mile trail and you're looking at quite a discrepancy!
So with these inaccuracies, why would you spend $200 or more on a GPS watch? Well, let's have a look at your options:
1. Drive your car the entire route and write down the total. This isn't especially accurate since you're not necessarily running on the same side of the road as the car, you are crossing streets and your car's odometer is accurate to a tenth of a mile. This is also impossible to do on a trail.
2. Get a Measuring Wheel and measure your route that way. This could get hazardous on a road course since you'll want to measure the tangents just like a race. It's also just as expensive as a GPS watch - more so depending on the watch.
3. Buy a GPS watch and turn it on for every run. You might be off by a few tenths, depending on the route and the weather, but for the most part this is best option.
You see, GPS units are light years better than any other method the average runner has available to them. There are the old standbys that say they're just fine going by time and guessing the distance, which is fine if you're comfortable knowing that your guess could be wildly off. Would you rather think that the course you just ran was "between eight and ten miles" or that it's 8.4 miles - accurate to probably within a tenth of a mile (at that distance)?
That establishes that there is value in using a GPS in training, but what about a race? Doesn't a GPS totally throw off your pacing? Well, yes...and no. If you know what to expect from your GPS you can adjust accordingly. If you're running a trail race, you know that you have a little leeway in your "average pace" because your GPS is going to tell you that the course is shorter than it really is. If you're running a road race, bear in mind that you have to keep a pace slightly faster than your goal pace.
Regardless a GPS is going to be more accurate than your "best guess." Just keep in mind how it will be inaccurate and adjust your training and racing accordingly.