Primo Numero Uno wrote:
I would guess most in the OP time range are easily spread across 40-60mpw.
A lot can do it with far less although then you're getting into the lifetime mileage debate. A college teammate of mine in his early 30s just ran 2:40 off about 40mpw after a years of casual running.
I'd guess if I was to stop my normal training and run 15mpw for 6 months I would have no problem being well under 3. But I also have around 50K miles in the bank over the past 13ish years which matters. So when you ask this question there are two parts. What does their current build up look like and what does their lifetime history look like.
So, shifting focus from averages and percentages to predictions for individuals, there's personal training history (lifetime miles, or perhaps going back a year or two), historical bests, and current 5K speed.
For instance, a 35-year-old has been fitness running for two years -- 3 miles three times a week.
In college, he ran 15:30 5K, averaged 60 mpw, but took a dozen years off from any training. Right now, he could run a 5K in 18:30.
With his current base of 9 mpw, he gradually builds mileage for five months, then averages X mpw for 16 weeks, and runs 2:59. What would be the minimum value of X?
I'd guess about 30. Same situation, but substitute 17:00 for college 5K, 20:00 current 5K, and I'd guess 40 mpw.
So many variables, but I bet a pretty accurate algorithm could be devised. The only point would be, "I'd like to break 3 hours -- how many miles would I need to train?"