Battle of the Predictions

As you know, I've been testing the George Parrott marathon prediction theory. However, I've also always loved playing with the McMillan Running Calculator. For the Buckeye Half Marathon, I plugged my Parrott predictions into the McMillan site to arrive at my goal finish time and pacing parameters.

With the Akron Marathon just a few weeks away, I'm trying to formulate a pacing strategy, using the two prediction systems as a basis. Problem is they're giving me mixed signals.

Last week's Parrott prediction, which includes my race and the brutal run home from work, gave me a marathon finish time of 4:50:53, my slowest forecast yet.

With my Buckeye time as a base, McMillan predicts a finish of 4:38:11, paradoxically my fastest forecasted race time.

And all along, I've set 4:30 as my A-plus goal.

This feels similar to my first marathon, when my goal was 4:00 and I finished with a 4:26, only add a half hour.  Of course, I can't rely on the past week's runs as my sole criteria for guessing my race finish time. I still have two more big training weeks ahead, with an 18-miler this weekend and my only 20-miler next week.

I might as well add some more predictions to the mix. When do you think I'll finish the Akron Marathon?