Here are some links to data analysis heavy postings for those that are interested in number crunching 8)
1) I downloaded the race results for previous years Eugene marathons and analysed them to see what the average variation in pace from first to second half was.
2) Before my Eugene race I did a comparison of the training cycle with the one for CIM (california international marathon).
3) After Eugene, I did two postings , this and this, comparing my 8 fastest road marathons but especially CIM and Eugene (my two fastest).
4) I found an interesting paper on how to measure LT without blood testing. I will be trying this out on myself with what I call 'calibration runs'.
5) I ran the San Lorenzo River trail marathon in 2011 and 2012 and compared my performances
6) I extracted some actual Garmin elevation profiles for various marathons I have run to help others figure out if that race is for them.