Possible mathematical win
 

After a short hike at a local park I decided to compare my speedometer to my gps. While driving the two seemed close. Once I got home the odometer and scangauge said 10.8 miles, but the gps said 10.98 miles. If my math is right that's a 1.66667% difference.

My last fill up was 9.202 gallons after driving 429.5 miles which comes out to 46.67mpg. If I add that 1.66667%, I get 436.66 miles and 47.45mpg.

I think I'll keep the gps in the car the entire length of my next tank. But at the rate I'm driving it will probably be a few weeks before my next tank starts.

I used Google Earth to map a 24 mile route here in the bay area. I used the "path" feature and zoomed in close enough so I could put over 1000 dots in the individual lane of the highway I use. This way when I went around a corner I made the distances between my points only a few feet. (Remember "piece-wise linear" in math class?) Anyway I'm in the same boat Google Earth is showing 1-2 percent longer then ODO and SG. If I get a GPS that would be four different measurements for the same route.

If you are an engineer, that is one you learn the hard way.
If your odometer says 10.8 you know 3 decimal places. You may be able to read between the lines some, and guess to the next half, but in my car the "tenths" place is not smooth, but tends to jump at the end of that tenth of a mile. Which puts a +/- on the readings. 10.8 miles isn't a long ways. I calibrated my car over 70.0 miles (mile markers that were down to the tenth of a mile) vs my odometer, and the tenths place jumped as I passed 70.0 miles. I also did a comparison with GPS over 200 continuous miles, and Google Maps over 400+ miles. All of them showed between 5.9% and 6% off, so for ease of math, I just use 6%. Just think, if your odometer doesn't click the last tenth of a mile, you are off by 1% at 10.0 miles, over 100.0 miles it's only 0.1%.

I'll definitely take the gps with me for more testing. I won't think about adding the extra miles to my records until I can get a consistent difference over several tests.

My dad put a gps in his Saturn VUE, and the odometer has consistently agreed with the gps. I guess I got a little over excited when I saw a difference.

I guess I'll see what happens over the next few trips.

Update
 

Just a little update.

I don't drive too much these days so testing is going slowly. Since the first post I made two small trips about 30 miles each.

The first trip the odometer said 27.7 and the gps said 28.15 miles(1.62%). On the second trip the odometer said 29.6 and the gps said 30.15 miles(1.86%).

I would like to measure a few hundred miles worth of trips before I come to a conclusion.

Tire wear can account for some odometer error. I just changed the tires on my '97 Escort and with the new tires the odometer and the GPS were dead on on a 500 mile trip and with the old tires I was accumulating a few extra miles. I didn't even think to see if the distance corresponded with the scan gauge or not.

Tread wear might cause a slight error in the opposite direction. My car has the optional 15 inch wheels instead of the standard 14s. The speedometer difference between the standard 175/65-14 and the optional 185/60-15 is about 3.3%. I'm guessing that Toyota may have set the speedometer somewhere in the middle.

I took a hit in indicated MPG when I used my gps to calibrate my MPGuino.

differences in speed is hard to get nailed down as it is hard to look in two places at once but distances are easier. when using a GPS make sure you have a good signal, and start moving as quick as you can on a highway trip as even when stopped the couple feet difference that the GPS calculates every second add as distance traveled up while stopped, but for the most part average out while moving. Go at least 10 miles and compare the GPS trip-o-meter to your car's (or guino's) trip reading.