What my own "tempo runs" typically looked like
Tip: All of this works better if you don't pause to drink like a fish for a week every few months
After declaring two recent articles about tempo runs—one published by Runner’s World, the other by Trail Runner—to be some combination of uninformed and unserious attempts, as well as at least passively asserting knowledge in this area by repeatedly referring to an ancient article on the subject I wrote myself, I decided to review my own history of tempo runs. Runner’s World can electronically burn as much of my work as its mentally flaccid and ethically bereft “editors” choose, but I am busy weaving myself a pink, pudenda-shaped protest-hat and preparing to be nasty as I need to be in fighting for my right to describe things on virtually unknown Internet sites, personal and corporate-run alike.
This post was prompted by an ongoing e-mail exchange initiated by a reader that may develop into a discussion thread. It won’t be the only one of its kind dispersed to the public, although I expect future variants to be both less infectious and less likely to result in symptomatic memories. This one covers the highlights of what I remember in the general area of tempo runs and.
I learned about tempo runs in, I believe, 1990. This had to be from an article in the “popular” press. I was still in college then, but my coach—a former javelin thrower with not one but two ruined shoulders and a porous head between—never implemented these runs or any workouts of proven or even plausible value in the two-plus years I “competed” at the University of It’s Fucking Freezing Here, and it wouldn’t surprise me if he never learned about tempo runs before he retired following a long and wildly unnecessary career wanking and stinking at the helm.
I perhaps regard marathon-pace (MP) runs and tempo runs as closer cousins than other observers, classifiers, nitpickers, and piss-takers. There are two reasons for this: One, I focused primarily on marathons when possible, and two, the faster a runner is, the smaller the arithmetic difference between his or her tempo-run pace (TP) and his or her MP. (Consider the simple case of a runner whose tempo run pace is, by one popular definition, simply his half-marathon race pace—a man with 21.1K and 42.2K bests of 1:00:00 and 2:05:00. This athlete’s TP (4:34.6) and MP (4:46.) differ by only about 11.4 seconds per mile.)
Anyway, I was amused to find after a slight reframing of my personal bests that I once ran for 51.5 minutes at 5:09.2 pace and on a different occasion ran for 68.5 minutes at 5:13.4 pace (not counting a faster half-marathon on a net-downhill Austin Motorola course in 2004 shortly before the 51:32 10-miler). Math teachers can confirm the fact that the numbers 51.5 and 68.5 are equidistant from 60.
So, it stands to reason that at this fitness level, I could have run for one hour on slightly rolling roads at around 5:11.3 pace, the exact midpoint between 5:09.2 and 5:13.4. (This would be closer to 5:11.4 because of the logarithmic shape of the pace-decay curve, but since I actually have no idea what I’m talking about at all anyway, I’m going to leave it alone.)
This would imply that I could have run about 5:06 pace on a track, which is where I almost always did tempo runs. There werte no GPS watches until my "prime" was winding down, and I probably would have done these runs on tracks even had I gotten one. That would mean 5:04 per 1,600 meters or 76 seconds a lap.
What did I actually do? Well, my best year was 2004, when I was 34. And I never did any tempo runs close to that fast, or anything continuous in a workout setting at that pace longer than 6.4K (just shy of four miles).
As for what happened in between: I started running marathons in 1994, when I was 24. My first one was a 2:39:37 on a hilly course in Atlanta. on Thanksgiving Day. I had run 15:25 for 5K on Labor Day averaged over 100 miles a week over an extended period for the first time, my initial foray into the sometimes-dreary jungle of genuine marathon mileage. But I bonked anyway, and just didn't really have the experience with keeping the right amount of pressure on. I was at 1:57-flat at 20 miles and then crawled up the inclines on Peachtree Avenue in the last 10K in 40-plus minutes.
Months before, maybe in July, I had gotten the idea to run 10 miles (actually 16K) on a track at what felt like marathon pace or "long distance race"-pace. I had never done a race longer than 10 kilometers. I remember doing this in 56-something and waiting until 9 p.m. because the summer midday weather in Atlanta is poorly accommodating of such efforts. I might have been able to run a marathon on a track in close to 2:30:00 at that point, but this run was at closer to half-marathon pace (on the roads, anyway).
My second marathon was in 1he fall of 1995 and I remember no MP runs, just high mileage with hilly long runs, VO2 max-type reps and a hilly half-marathon race in 1:12:06. That led to a 2:32:48 with marked negative splits on a two-loop course in Lowell, Massachusetts.
My training for the 1996 Boston Marathon included one 25K run on the track (I remember this one being cold and dark—this was in Lebanon, New Hampshire in January or February—at a little under 5:40 per mile and a 40K on the track three weeks before at a little over 6:00 pace. The second one was unplanned -- I wanted to go 25 miles and I lived practically next to a track, so I started running 94-second laps that gradually became 90s with a few 89s. I ran Boston in 2:36:11, again with a sterling final 10K in around 40 minutes, this time descending rather than climbing.
My next one was in 1997 on the same course I'd run 2:32:48 for 8th place. This time, I did some long, fast, sustained track runs. One of these was a 48:28 15K, split 16:20/16:10/15:58. That's 5:12 pace on the nose, and by the end felt pretty much like a race. I ran 15:27 and 32:02 on the roads during that build-up, though my best effort might have been a 15:59 on a hilly 5K course staged mostly on the dirt roads of Canterbury, New Hampshire, a place littered with unexploded cluster-bombs and about which I tell a disproportionate number of Tales. I also ran a couple 8K and 10K "tempos," or whatever they were, in some order in times of 25-high and 32-low.
I was aiming for sub-2:30:00 this time, but in reality wanted more like 2:28. And I ran a little under 5:40 pace like clockwork through 22 miles, when my decision to stick almost entirely with water paid reverse dividends. I was on 2:27-low pace and bonked my way to a 2:30:52 thanks to a string of miles in the 6:30 to 6:40 range.
Even better, I discovered at halfway that I was leading the race (about 35 runners were ahead of me in the accompanying half, and I was the first one not to make a left turn for the finish when we passed the starting a second time). I led until about 24.5 miles when I was passed by the eventual winner, who thenceforth put 1:40 on me. This was all captured for an ESPN show called Running and Racing, a short-lived production that usually aired at around 2 a.m., right when dedicated runners and running fans were likely to be sitting in front of the idiot box. About the only footage they showed was me getting passed by Alex Tilson, who was wearing a Speedo and had a old-school yellow-and-black Walkman strapped to his arm. Tilson would later qualify for the U.S. Olympic Trials in the marathon and set the official American record for 50K.
In 1998 I entered three marathons, and completed almost 67 percent of them, dropping out of the Vermont City Marathon at 18 miles on a hot Memorial Day. One of the other two was a supposed practice race and the other all-out, but I ran the all-out one in 2:33:26 (Hartford Marathon, October) and the practice one--which did feel "easy"--in 2:33:34 (Fred's Marathon in April, the only marathon I ever won). I didn't send my best that year, for whatever reason.
In 1999 I returned to the Lowell race, the Baystate Marathon, for a third time (and for the third straight odd-numbered year) and ran exactly four minutes faster than in 1997 with a 2:26:52. Almost the only difference between the two races was not blowing up in the final four miles. Also, on the first lap of the race, a seafood restaurant near the 7-mile mark had a tall sign featuring regular items like crabs and lobster, I only remember this because on the second lap, someone had gotten up on a ladder in broad daylight in a city of 100,000 people and changed the letters so words like WE EAT CUNT, FAG! were up there. If you understand Lowell, Massachusetts, this won't even create a ripple in your mind. But it was hard for me not to laugh even from within my shell of "OK, this is a good day" concentration.
Notably, three weeks before that race, I had run 30K on a track at very close to that 5:36 pace, within a second either way. My dog did most of it with me but then started cutting tangents at the corners 20K and kept going back and forth across the infield the rest of the way. he never thought to just wait 84 seconds for me to return to the same spot.
My next 26.2-mile race was the 2001 Boston Marathon. This was both my fastest marathon and the only one in which I stopped to take a dump, a combination that seems to suggest that I was far better fit and prepared on that one day than I was for any of my other marathons. This came six months on the nose after that aforementioned 1:08:29, again on the Baystate course (my first time not running the full marathon there). And this time, I did a 20-miler at 5:31 pace (a road race) three weeks in advance of the marathon, which I ran at 5:30 pace including the minute or so I lost to porta-john stops and attempted stops. But I did a lot of tempo-type work on both the track and a half-mile road loop in Concord I had by then found and infected with kids on area cross-country track teams. These included a 20K on the track in 1:06:09 (5:19 pace).
I also did an 8K in 25:43 and used some longer road races as tempo-like efforts, including a rolling 15K two weeks out in 48:58 (5:15 pace) and Stu’s 30K six weeks in advance.
That stops well short of 2004, when I ran most of my personal bests, but the general pattern is not doing so much tempo runs but allowing for elasticity in pace with increasing or decreasing distance. For example, that 25:43 was at 5:10 pace and the 1:06:09 at 5:19 pace, but they both probably had equal training-stimulus value. And the 10K track runs in 32:15-ish and that 48:28 16K, though these were strewn across different training cycles, fit this same pattern of operating in the range of my true tempo-run pace of 5:15 or so.
Now, I mentioned being capable based on my fastest races of 5:06 pace on the track, at least in 2004. The math makes this all but undeniable, but this seems a lot faster than even my hardest "tempos" on the track. But this does make sense given that, as hard as those tempos were, they were all negative split and all felt comfortable for at least the first two-thirds of each effort.
The last thing I'll add is that I tried to assess my threshold level of effort by my breathing pattern. Like most, I start to inhale and exhale when my feet git the ground, unless running so easily it doesn't matter. If I could maintain a 2:2 inhale-to-exhale pattern, I was running aerobically. But once i needed to switch to a 1:2 pattern, that was racing mode, even if I didn't feel the walls closing in just yet.
If there is a theme in the above rambling, it's that instead of tweaking only the distance of tempo runs, or busting them into segments, one helpful strategy is keep them continuous and play around a little with the pace instead, within sensible limits. The suggested duration is maybe 25 to 75 minutes, allowing for the fact that about 60 minutes, you're not technically doing a threshold run anymore, even if you are.
Next time, I’ll dredge up whatever I did in this area between the spring of 2001 and the spring and summer of 2004, when I set a number of lifetime personal bests. What I write will probably depend on the nature of whatever chatter this generates.