I find this data really interesting. I don't remember my exact strategy in '08, although I do remember wanting to keep up with the 4:00 marathon pace group in order to finish at 2:00 or less. That worked for the first mile, but apparently I got to feeling pretty good and took off. Considering that was the first time I'd ever attempted to run more than 10 miles, that was a horrible decision, as I never saw a sub 9:00 mile past the half way point of the run and really lost it at the end when I started to cramp and needed walk breaks over the last three miles.
On the other hand, I went in with a much better strategy in '09. My goal was to maintain an 8:50-8:55/mile pace, and the reason for the fast miles 2 and 3 was to make up for the slowish start caused by crowding out of the gate. After that I was pretty happy with how I did, although I wonder if going a little faster in miles 9 and 10 hurt me during miles 12 and 13.
Also, I believe the split times for mile 6 are off by quite a bit because that's when we went through a tunnel and my Garmin lost the satellite for a while. I was already hurting by that point in '08 and I don't think there's any way I ran an 8:40 then. I also don't think I went anywhere nearly that fast in '09 considering how I was settling in to my 8:50-8:55/mile groove then.
I have no explanation for the gross difference between the '08 Garmin and chip times, other than the fact that I used the lap function of the device trying to get a more accurate time to compensate for satellite loss on the run, but I'd think that would give me a faster time than chip time rather than slower.
Oh well, it's still always fun to look back on things and view them with 20/20 vision.