Disclaimer: THIS IS NOT A SERIOUS SCIENTIFIC EXPERIMENT! Take this post with a grain of salt ;)
Someone on Running Mania posed the question: how well do our half marathon times predict our full marathon times, and how does experience come into play? So, we posted our race times. As some people started hemming and hawing over what the lists meant, someone asked what it would look like on a graph. Well, I have too much time on my hands, apparently, so I threw the numbers into a spreadsheet:
According to the 47 people who posted their times, the formula for predicting marathon finish time was roughly:
Marathon time (in minutes)= Half marathon time (in minutes)*2.08 + 16min
Now, this cannot be considered as a true marathon time predictor like MacMillan or Riegel of course, because the data is not controlled. In many cases, the half marathon times were not representative of the same relative fitness level as when the marathon was run (the times may have been a few years apart), or the person may have only run one marathon. However, there didn't appear to be any correlation between number of marathons completed and predicted finish time, either.
All I can concluded from this is, if I ever want to run a 4 hour marathon, I had better get a heck of a lot faster in the half marathon ;)