Running down to one's potential

It's been five weeks since that somewhat traumatic, unfinished run in the Abebe Bikila Day International Peace Marathon. Whatever had happened to my lower back around Mile 10 of that race has mostly recovered. I've been very diligent in doing my core exercises! My next marathon is Harrisburg in early November, so the current training cycle is a shortened and modified one, with just four long runs. The first one is a 13-mile test run, two weeks after Abebe Bikila, followed by runs of 16, 18, and 20 miles (this past Sunday). This week is the first of a modified 3-week taper,

Starting with the 16-mile run, I made one important change. In previous long runs, I'd run most of the second half miles at or slightly faster than my Boston qualifying (BQ) goal marathon pace (GMP), but I'd take a walking water break with about 1-2 miles to go and then run home at ~GMP plus 30 seconds. During this 16-mile run, it occurred to me that finishing long runs at a slower pace is not what my body--and mind--should get used to! So, I delayed the cool-down and finished the last three miles at 11, 23, and 28 seconds faster than GMP, smiling all the way home. :) I similarly ran the 18 and 20 miles, with generally negative splits from start to finish. As shown in the pace chart, in all three runs, I took planned walking water and Gu breaks at Miles 5, 11, and 15 (except for the 16-miler). I finished the 18- and 20-mile runs at GMP minus 50-55 seconds.

I feel good about how this training cycle has progressed and, in particular, the higher quality of the long runs. I attribute the latter to another key change I made after Abebe Bikila, which is reducing the training volume by about 25%. Previously, I'd labored just to get the second half of my long run pace down to around BQ GMP. Now, I'm running the last miles of long runs at GMP minus 30-60 seconds. Although these end of the run miles were not exactly easy, they're not major struggles either. In fact, I was quite surprised, at the end of the 18-mile run, when I saw 8:05 for the last mile. My main goal for Harrisburg is to just finish (given the DNF at Abebe Bikila). But, feeling strong from these recent long runs, I've been thinking about what is my current realistic marathon potential. Plugging in my half marathon PR in the Runner's World race times predictor results in the following chart, which shows the predicted times of shorter races very close to my actual PRs. For the marathon, however, the predicted time is about 17 minutes faster than my PR of 3:59:23 or about 3:42, which would take me way past my threshold BQ!

Are these 17 minutes currently realizable? The predicted times for marathons, more than those for shorter races, depend on all the stars being aligned on race day. Given my current level of fitness, is it just a matter of running down to my potential? Or, is more training needed to achieve a higher level of fitness, at which, even if the stars are slightly misaligned on race day, the potential could still be realized? Another way to think about the change I've made in recent long runs is I've been training, albeit for an abbreviated period, for a GMP that's faster than my BQ GMP. Will find out in Harrisburg. Oh, wait, I was planning to just finish ...