I don t like to apply precise, limiting criteria....whether percentage or time based....to the relationship of the long run to total mileage on a week by week basis. For instance, some people, including some experts, suggest that the long run should be limited to a certain percentage of weekly mileage. Some say that it shouldn’t be longer than 20-25%. Others say 30-35%. Still others say that it is OK to go as high as 40%. Bob Glover suggests that it can go as high as 50%. OTOH, Jack Daniels says it should be limited to 25-30%. The recommendations are all over the spectrum.

Applying the criteria of somewhere in the 25-35% range, which is where most of the recommendations that I have seen fall, would limit the 40 mile/week runner to a long run of 10-14 miles. I think that is inadequate to prepare to run a marathon seriously. I think it is perfectly OK for a 40 mile/week runner to go 18-20 miles or even higher, although not very often. Of course, the 80 mile/week runner will only expend 20-25% of his/her weekly mileage on a 18-20 mile long run and can comfortably go even longer within a 25-30% criteria....and s/he can do it a lot more often than can the 40-mile/week runner. However, s/he doesn’t have as much need to push that element of his/her training as the 40-mile/week runner does.

The problem I have with using this criteria is that it leads one to look at one week at a time. However, a week is a snapshot within a training program. There are typically 16-18 of these “snapshots” in a marathon program and they can vary in mileage significantly for any number of reasons having to do with other aspects of one’s life. The result is often to choose between either adjusting the long run to be consistent with a percentage criteria or to deviate significantly from normal long run/total mileage percentage criteria from week to week.

I think that it is better to step back and look at a broader training picture. I think the number of miles spent on long runs, compared to total miles, throughout a training program is far more important than the relationship of long runs to individual weekly mileage. That permits averaging the percentage of long runs vs. overall mileage over a longer base. Another way to say it is that it enables one to better manage the role that long runs play in the total work effort involved in a marathon program. And I think that role depends on one’s conditioning and race goals. Let me offer two examples from my personal experience:

1) For the 1989 MCM when I ran my all time PR of 3:22:27 at age 51, my very balanced 16-week training program totaled 730 miles for a weekly average of 45.6 miles. My individual weeks ranged from 25.7 miles to 60.2 miles. My long runs which peaked at 24.6 miles, totaled 223 miles, or 30.5% of my overall 16-week mileage. However, my individual long runs ranged from 22.8% to 50.2% of their respective weekly mileage, which was a typical mix of long run mileage vs. total program mileage during my first running life from 1983-90.

2) To prepare for the 1997 Philly Marathon, which was the third race and first marathon of what I call my second running life, I focused solely on mileage and long runs because my goal was to lose weight and get back into shape following a 7-year hiatus from serious running.....in other words, it was base building. My 16-week training program included no speedwork. I ran 766 miles (an average of 48 miles/week) of which 320 miles (41.7%) were spent in long runs...that’s over 10% more than I typically ran in my first running life. My 16 long runs ranged from 14 to 26 miles....with 10 of them 20 miles or longer, plus two 18-milers. My subsequent marathon was the strongest (not fastest) marathon I have ever run. I think the strength I had in that marathon was due to running the highest mileage and the most 18-26 mile long runs, by far, than I have ever run. The following year, I sacrificed a little (about 2-4%) in total mileage and long run mileage, added speedwork to my regimen and ran a faster marathon. That’s one form of macro-cycle periodization.

Now to your specific questions.

1) Does the 40 mile a week runner really benefit more than a 80 mile a week runner with a 20mile long run? Isn't the physiological effects the same for both runners?

Yes, I think the 40 mile/week runner benefits more. And, no the physiological effects aren’t the same. I think there are a couple of reasons for this. Firstly, the 80-mile/week runner is probably more advanced than the 40 mile/week runner. S/he has probably been running longer and is better conditioned; thus derives less development benefit from long runs. Secondly, the 80 mile/week runner is doing more total work than the person running half that mileage. Said another way, the 40 mile/week runner is running a less balanced program and has to rely more on the long run to develop endurance and running strength.

That isn’t to say that long runs aren’t important to the 80 mile/week runner. They are. However, their relative value is less within the context of a balanced program than for a 40 mile/week runner. The 40 mile/week runner might not be able to do as many of them or extend them as far, but they play a more important role in his/her total training regimen than they do for the 80 mile/week runner.

Taken to an extreme, since Salazar never had to run longer than 18 miles to compete at the elite level, then why should any runner have to go longer than that to just run good marathons? The difference is in the total program. The long run is just one element....and I think it is second most important after total mileage for runners running 50-60 miles/week or more. However, when average weekly mileage is less than about 50-60 miles/week, then they take on a primary importance because they constitute a major element of development work.  Generally, the greater one’s total mileage, the less important the long runs become in relation to total mileage for endurance, strength, and running efficiency development, which is the purpose of both long runs and total mileage.

2) Wouldn't our 40 mile guy likely need more recovery than our 80 mile guy after the 20miler (all other things being equal)?

Probably. But s/he also has more opportunity to recover. For instance, at 40 miles/week, s/he can afford a day off after a long run, if necessary. S/he is also probably doing less of other forms of quality running in the form of speedwork and hillwork than the 80 mile/week runner, who is doing more total work and probably running every day....maybe two-a-days many days....with lots of quality work and less room for down time in his/her schedule.

Remember, it all comes down to the total amount of work being exerted in the overall program. Don’t get overly focused on a specific element of a training regimen. Try to keep each element in perspective relative to the overall program.

3) Based on your experience, what is the sweet spot for most runners in terms of time of the longest run for marathoners? I personally now think it is right around the 2.00 - 2.30 mark.

Just as I have a problem with imposing a percentage of weekly mileage limit on the long run, I also have a problem with the imposition of time limits. I have seen recommendations, including from experts, that the long run should be limited to 3 hours, or 3 1/2  hours, depending on who you listen to. Now, you suggest 2 – 2 1/2   hours.

However, most marathoners run slower than 4-hours today. I suspect that somewhere around 4:15 is about average. Thus, imposed time limits of 2:30-3:30, would mean that many, many marathoners would not run longer than 10-15 miles in training. I think that would be very unwise.

I think that limits of 2:15 to 3:30 are OK for 2:15 to 3:45 marathoners....maybe even for a 4:00 marathoner....who are running relatively high mileage in training. However, I think it is better to follow a few other criteria concerning long runs:

1) Run a minimum of 18 miles at least once in training for a marathon.

2) Run at least one long run in the time that you plan to run in the marathon.

3) Run as many 18+ long runs as you can handle without compromising other elements of your training program, especially total mileage and speedwork.

All runners from beginner to advanced who plan to run a quality marathon should do 1 and/or 2 above, regardless of weekly mileage. All marathoners who plan to run a quality marathon in less than 4:30 or so should do 2 and 3 above. None of the above applies to those who plan to just finish in whatever time they can using any necessary combination of running and walking. :-)

Some of the thoughts and opinions that I have offered above are probably controversial. Certainly, some contradict what some of the experts say. Others on this forum will disagree with them. That's OK. None of this is an exact science. As the old saying goes, each runner is an experiment of one. There are very, very few "absolutes" in this sport. But, one is that both long runs and total mileage count a lot in marathon training. However, the combination and composition of them that is best for a particular runner can vary widely.

Jim2