A couple of weeks before this year’s Iron Mountain Trail Run (IMTR) I ran “the two-hour-loop.” This is a course I scouted back in 2005 and used as a benchmark while training for the 2006 Mountain Masochist Trail Run (MMTR). The route starts near Skulls Gap and contours along the Iron Mountain on “old 84” to Hurricane Gap. It crosses the ridge of the mountain at Barton Gap and then drops to the bottom of the mountain at the trailhead for Jerry’s Creek (near the trailhead for Rowland Creek Falls). It then climbs the mountain again, following Jerry’s Creek, to near Skull’s Gap. The terrain has equal parts single track, double track, and gravel forest service road. The course climbs 2700 feet in nearly 18 miles. It took about 2 hours 20 minutes to run it at first, so I decided to come back to the course at regular intervals and see how close I could come to running it in 2 hours. Treating the run like an extended tempo workout, I began to approach the 2 hour mark after a few weeks.
About that same time, David Horton came down for a training run in preparation for the first IMTR. The middle section of that course is very similar to the two-hour-loop except that IMTR follows Rowland Creek instead of Jerry’s Creek back up the mountain. During that run Horton asked me my goal for MMTR. I told him that I wanted to break 7 hours. He quickly retorted that I couldn’t do it. Only a couple of people had done it, and when the inestimable Eric Clifton had failed to he reportedly said that he didn’t think there was anyplace he could have made up the necessary time.
My response was that I’d let my result speak for itself. I also said that my training included the two-hour-loop, and that I thought when I could break 2 hours on it, I could break 7 hours on the MMTR course. Though we quickly moved on to other topics, it is interesting to note that Horton can recall this story with more clarity than I can. Part of the reason, no doubt, is that I did ultimately break 2 hours on the loop and 7 hours at MMTR. He greeted me at the finish, as he does for all runners, and was able to witness the most special moment of my athletic life.
Horton is one of the few people who could really appreciate what reaching the finish line in Montebello meant to me. That place represented much more to me than winning a race, or meeting a goal, or proving something. That race finally brought into relief the full scale geography that is me. I had first stumbled down that same campground road in 1997 in the middle of my thru-hike of the Appalachian Trail. I was empty – destitute and alone, and slept on the wooden porch of the pavilion. My second approach to the same road was in 1998 after suffering a beating through my first ultra at MMTR. My calves had shredded by mile 37, and I had to hobble on my heels for the next 16 (?) miles to the finish. I had been sunk as low as I have ever been twice – both times across from the small general store that defines that tiny town. The contrast in 2006 could not have been more apparent. I had started to build momentum on the climb up Buck Mountain, and I didn’t let up. I blasted the final mile along the road in under six minutes and charged into David’s embrace. I cried for the only time at a race.
I retired the two-hour-loop for some time after that. It wasn’t so much the weighty feelings surrounding it, but more that it’s a tough run and shouldn’t be overdone. I’ve had one other reason to avoid the two-hour-loop. It would be too good an indicator of my fitness. I was 38 in 2006, and I’ve been getting older ever since (!). I wouldn’t necessarily want to know how my performance had declined.
That is why I was so pleasantly surprised early this spring, when after a minimal flurry of speed workouts, I was able to again go under two hours for the loop. Soon after I ran a PR at the Promise Land 50K, a race I had already ran respectably. As quickly as I had begun to run fast, though, I slowed back down because my goal for this year had been to run 100s. I slowed my training and muddled through Old Dominion and the first 70 miles of Burning River. The misery induced by these races finally caused me to reconcile the space between my aspirations and my talents. So much for 100s.
I like to run fast. I decided to keep my races at or under 62 miles this fall, and to run MMTR in November once again. In preparation I’d confront whatever the two-hour-loop had to say about my fitness. Two weeks before IMTR it said 2:05. That translated into a 7:15 two weeks later – a respectable time on a tough course. Historically runners are able to run similar times at IMTR and MMTR. Of course I’d like to run faster than 7:15 come early November. So this past Wednesday I was back on the two-hour-loop, having carefully considered what I had to do to go under two hours.
Here's an experiment I would like to do: get a bunch of runners to do a hilly 8K time trial. Tell all of them to go for the fastest time they can without killing themselves. Then break them into two random groups. Tell one group to push the uphills, and explain that will help them keep the best overall pace. Tell the other group to slow down on the uphills and pick it up on the downhills, and explain how that will help them to even out their effort and yield the best overall pace. I’ve given contradictory advice – I can’t have been truthful to both groups.
I’d like to do the experiment because I know now how it works for me – and how I’ve been lied to.
There are two distinct models for understanding how fatigue limits performance, and they predict different outcomes for the experiment. A "peripheral fatigue" model says -- and this is oversimplified -- that pushing the uphills will fatigue the muscles and therefore (eventually) cause runners to slow down. It would predict that maintaining a constant threshold effort (by slowing down on uphills and speeding up on downhills) limits fatigue and therefore maximizes exertion across the whole exercise. According to this model the advice is simple – DON’T push the uphills.
In contrast, a "central governor" model says that runners (perhaps subconsciously) anticipate what is ahead and experience fatigue depending on how they interpret their current exertion relative to how much is left to run. This model at least leaves open the possibility that pushing the uphills could result in a better time, because performance depends on how the runner interprets what he or she is doing. So, while not the advice is not as clear, we can at least advise: PUSH the uphills if you can.
When I was in college heart rate monitors were just coming into use. Our team bought one and rotated it among team members during training. The idea was to make sure we were not over-exerting. The theory was built on a peripheral-fatigue model. One inescapable conclusion was that the best pacing strategies tend to even out effort across the run. In other words, all else being equal, slow down on uphills so that your heart rate remains relatively constant. We didn’t apply this to races, because in a race all else is not equal. Relative place matters much more than time, and maintaining place on climbs surely trumped maximizing overall pace.
My tempo runs are generally alone, though, and so I have tended to fall back on the theory that I should slow down on the uphills in order to get the best out of the total distance. That is the lie. And I think I know why.
On Wednesday I ran just under 1:58 for the two-hour-loop. I worked the uphills pretty hard. Part of the reason I have to do that now may be that I can’t run downhills as hard as I used to, but I think there is another reason. When we charge uphill we raise the threshold for perceived effort. We breathe hard and our hearts pound and we go for it knowing that the climb will top out and we’ll continue on an easier course. Because of the raised threshold, however, we tend to pick up the pace on the downs and flats compared to what we would have if we had backed off on the climb.
I don’t really know if this is a personal quirk – perhaps shaped by my own way of thinking about running or just my unique physiology – or a generalizable finding that would help corroborate a central governor model of fatigue. That’s a good reason to run the experiment. In the meantime, I’m going to push the uphills.