Posted: 13/10/2014 at 13:55
In case an example around similar times helps:
After a year of regular running - mainly 5 to 10K, building up to 4 runs and 20-30 miles per week - I managed a 1:28 for my first half marathon when McMillan would have predicted a 1:25 off of my 5K time.
That was with 3 or 4 long runs of 12-14 miles during the couple of months beforehand. So my legs had some familiarity with the distance, but I was still very much undertrained for racing the distance well.
If my 5K had only predicted me a 1:30 at that time I suspect I'd have ended up with about 1:34.
Or if my long runs had all been a few miles shorter then I suspect I'd have ended up with about 1:34, or injury.
Or if both of the above then I suspect I'd have been looking at about 1:40.
The following year when I did more sensibly structured training - double the weekly mileage above - mostly following Hal Higdon's training plans for Half Marathon ( http://www.halhigdon.com/training/51133/Half-Marathon-Advanced-Training-Program ) and Marathon - that's when I ended up with a 3:04 first marathon and a HM time that bettered the McMillan prediction from my 5K.
Other training plans are available...