After setting a race goal, the first questions runners ask are how many miles per week they should run and what pace they should be running them. There are complicated ways to determine these, but there are also some easy rough estimates. As I'm training for a mile race, the average training pace should be 1 1/2 times the race pace, with easy runs 30 seconds per mile slower than that. For marathoners running under 3:30, the average pace should be a minute per mile slower than race pace, again with easy runs 30 sec./mile slower.
The number of miles per week is trickier. I've compiled records for people who trained and raced without much outside influence (almost impossible today) and found that, regardless of finishing time, racers at any one distance tend to run about the same number of minutes per day: 40 min./day for milers, 60 for 10K runners, 75 for marathoners. Using total minutes and pace, one can determine mileage.
So, for me looking at 5:18 for a mile goal as an example, 1.5 x 5:18 = 7:57 (call it 8)/mile. Add 30 seconds and easy pace = 8:30. For comparison, Jack Daniels has an easy pace of 7:48 for a 5:16 mile. For mileage, 280 min./week divided by 8 min/mile = 35 miles/week. [For a 3:50 miler, this would be only 50 miles/week, but at that level, I'd have two-a-day workouts, bringing the total to 100/week, about what is commonly done.]
The next question comes automatically. If you're supposed to run 35miles/week at 8:00/mile, should you run the miles and try to improve the pace, or should you do as many miles as possible at pace and build up mileage? My answer - though there are too many exceptions to list - is to focus on time run. So, if I were running 9.5 min/mile (which was the case recently), to get to 280 minutes/week, I'd be running 29-30 miles/week (far less than I was running at the time). As I improve, both the pace quickens and the mileage increases.
When a day off is like a long run
The next question to address is the distribution of mileage over the week. One of the most common mistakes is to run the same amount every day, both for convenience and to be able to constantly compare one run to the next. Running the same distance every day invariably leads to injury when, after weeks of training, one makes a change, such as doing a race. I have very complicated procedures for deciding how many miles to do on various days, but there is once again a simple guideline - you can run the same distance five days and take two non-consecutive days off each week, or take one day off and have a long run of about twice the typical day's run's mileage. This seems to be just enough variation to avoid "staleness."
Why long runs are problematic for milers
Just as the total minutes per week is fairly constant for each race distance regardless of pace, the longest run for each is also a fairly constant amount of time. Faster marathoners typically have long runs of 2 1/2 hours, with a longest run of 2:45; beyond three hours, these runs become counter-productive, as they begin to use muscles and energy systems in ways not used in the race and require too long of recovery to maximize utilization of training time. 10K runners typically have long runs of 2 hours, which is about the limit of what they can sustain at a reasonable training pace and remain completely aerobic.
Milers face a conundrum when planning long runs: their long runs are not much greater than their average. Since the 1960's, the typical long run for a 4:00 miler has been 10 miles in 60 minutes (note that the pace is 1.5x race pace), partly because of the nice round numbers, but also because 60 minutes is about the limit of what one can run at "anaerobic threshold" pace. Some miler's workouts, however, take more than an hour to do, which necessitates standing rests (rather than jogging recoveries) or the very careful mixture of paces that does not do too much at any one pace - and which still amounts to under 80 minutes. Anything beyond 75 minutes is counter-productive for a "pure" miler.
So, if a miler is running an average of 40 min/day for the week, with two days off, this becomes 5 runs of about 56 minutes each: in other words, making every single run a long run! In the earliest phase of training, when training just for (short-distance) endurance, this is not a terrible idea.
Base training progress measurement
In this preliminary phase, improvement is tracked by mileage and pace. When these no longer improve, one is ready to move on to the next phase. There is one flaw to this procedure, when one has an athlete that is highly motivated and competitive - improvement in mileage and pace at the cost of much greater effort. I've made this mistake - repeatedly - in my career, as it is sometimes possible to keep pushing in "easy" training runs until one becomes exhausted, over-trained and one crashes.
A way to prevent this is to track effort levels for each run, either by perceived effort (e.g. the 20 point Borg scale), or with a heart rate monitor. With a monitor, one's heart rate should stay roughly the same as one improves and should be lower for any one given run as one improves. It is not necessary to force oneself to run below a certain heart rate (such as in systems like Maffetone's), just to note trends so one doesn't develop what's called "wind-up," the drive to make every run harder than the previous one. When one's mileage and pace at a constant effort plateaus, it's time to move to the next phase. Here one should check to see current race ability, taking the weekly minutes, dividing by weekly miles and then dividing by 1.5 to get a mile time one could run at that point.