Cat Eye Question

[BleeckerSt_Girl]

Wow BSG, seems like a waste of gasoline to me, but to each her own.

I've found rollouts to be extremely accurate. I usually roll the bike out over 3 times the circumference and then divide the measurement by 3.

Actually, you're right- there is a much better gas-free way I forgot about- map your favorite ride out on MapMyRide or any other online mapping program, and see how many miles it is. Make sure to check the "follow the roads" box to be most accurate for the mileage. I have found their program to be pretty darned accurate concerning milage of my favorite rides.
Then calibrate your computer to match that. You might have to ride the route a few times and make small changes to your computer's tire circumf. measurement until it matches the mileage for that ride. The longer the ride you use as a test, the more accurate your results will be. I did stop using my car and switched recently to the online mapping mileage method, but forgot to mention it, since I don't have to calibrate new bike computers very often.

I used Sheldon Brown's method the first time a couple of years ago, but found it to be not as accurate as I'd like, and the table of tire circumferences that came with my CatEye computer was worse.
Any little inaccuracy on a rollout of only a couple dozen yards is going to be vastly multiplied over a 20 mile ride. Plus, you need two people to even come close to getting a good reading on a rollout- one on the bike (you have to be on the bike with feet off the ground for any tire circumference measurement to have any meaning at all, since your weight flattens the tire and thus lessens its circumf.), one in front holding the handlebars so you don't fall over while you s-l-o-w-l-y roll down a slight decline slowly enough to actually count the rotations.
I've also read that the reading you get when rolling slowly will be slightly different from what you'd get if riding 15mph, due to fast forward momentum lessening the weight bearing straight down on your tires, thus slightly altering your circumference. Don't know if this is true, but it sounds logical to me. All I know is I wasn't getting rollout readings that matched up with distances I knew to be accurate on long rides of 10 miles or more. On short rides the inaccuracies were not as apparent.
The online mapping program and using your favorite ride's known mileage (of at least 5-10 miles) I have found to be more accurate AND way easier. At least that is my own experience, after having carefully used both methods.

Jobob- if you get accurate rollouts from 3 rotations then that's great. I was not able to get accurate readouts from that.

Veronica- I do like to know whether a ride was 42 miles or actually 45 miles. I agree that if one doesn't care about that then it's fine and not a problem.
When I didn't have a bike computer at all, I kept wondering how many miles my various routes were, it was hard to tell which ones were longer or shorter, they all varied so much. I wanted to know, especially when I would have a certain time slot to fit a ride in my day but didn't want to chose a ride that would be too long to complete before I had to be home, etc. That's why I bought the computer. I'm such a slow rider anyway, and I don't care about stuff like heart rates so I didn't buy a fancy computer for that kind of stuff. Now I have a little log of distances of various rides I like to do, and I can tell which ones will fit in my day on any particular day....I like that! I also like to see how my average speed slowly improves after winter ends, as Spring turns into Summer and then into Fall and my fitness improves.....