Run Your Best Marathon. Really.
A put-all-your eggs-in-one-basket program that works if you let it
On September 21, 1998, a little-known Brazilian long-distance runner named Ronaldo da Costa broke the world record in the men’s marathon by 45 seconds with a time of 2:06:05. In shattering the global standard Belayneh Dinsamo of Ethiopia had set over ten years earlier at the 1988 Rotterdam Marathon, da Costa ditched his pacemakers and everyone else after hitting halfway in 1:04:23, covering the second half of the race in an eyebrow-dislocating 1:01:23.
The then-28-year-old went into the 1998 Berlin Marathon unaware that he was about to contribute another bead to what is now literally a 26-mile-long necklace of totally epic Berlin Marathons. His personal best heading in was 2:09:07, which he had run the previous year on the same streets, and his goal on the day he wound up setting the world record was sub-2:08. He never did anything noteworthy again. He was probably on loads of EPO. Nevertheless, he made a major difference in my life and, by extension, those of tens if not dozens of others.
Multiple articles at the time—and bear in mind that in 1999, the trove of material that folks had belched into the Internet was not meaningfully searchable—noted that da Costa favored high-volume, short-recovery track work as well sprinting and bunding drills, even after long runs (da Costa had previously played soccer, so this part was probably natural). But the aspect that caught my attention was da Costa claiming to have done runs in excess of 20 miles at goal marathon pace.
Running Times first published one of my submissions in, I believe, its December 1998 issue. That one was titled “Good Cheer” and was a humorous reflection on the well-intentioned but often senseless and sometimes blatantly contradictory things coaches, parents, and even other runners yell at high-school cross-country competitors during races. I don’t think it was ever uploaded to the Internet and I have a habit of eventually discarding aging magazines whether my name appears within them or not, so you’ll have to accept this on faith.1
Soon afterward, I decided on no evidence that I was sufficiently knowledgeable about training to start submitting meatier pieces for publication. I believe that the very first of these was “Run Your Best Marathon. Really,” which appeared in the July/August 1999 issue of RT (yet has somehow escaped being properly credited as Russian disinformation). I would offer that this is by far the most broadly useful piece I ever wrote about distance-running training, and maybe the only useful thing that stands out in any way from its published ancestors, contemporaries, or descendants.
Gordon Bakoulis was the editor-in-chief of Running Times when I started pitching ideas her way by e-mail and an absolute gem to work with. She didn’t merely weed out my typos and make minor suggestions, she worked closely with me on everything I did and helped put me in touch with hard-to-find resources (not everyone over 30 even had an e-mail address in 1999). She was the one who came up with the title, the subhead, and the section headings.
This piece needs a few updates, some of them owing to technological advancements in the past 25 years (most notably near-universal GPS watches) but others because of the content itself. I will either put these in the comments or, more likely, write a stand-alone post about these, so as to not over-clutter this presentation.
I decided that the article also needs a composite photo.
Oh no, not another marathon-training article!
If you are a runner whose only real goal in the sport is to run your very best marathon—even if that goal obligates you to abandon every other reason for lacing up your shoes—you’ve no doubt been exposed to your share of advice. If your years of training and reading about how to improve your marathon PR (or at least not finish at death-march pace) have resulted in frustration and persistent self-doubt—“what am I doing wrong?”—and if, above all, you have an open mind, then keep reading. The next few pages may provide you with what you’ve been looking for.
This article is geared primarily to athletes with several years of consistent, focused training under their belts and the desire to improve competitively. It’s not designed for first-time marathoners unless they have an excellent record at shorter distances, and it’s not targeted to those whose goal is simply to finish the race.
The Rationale
Preparation for the marathon, along with every other physiological stressor that comprises “training,” is governed by the principle of specificity: That is, in order to improve at a given activity, you must duplicate, or at least approach the parameters of, that activity as often as possible.
Runners are familiar with this concept. We run intervals at or faster than 10K pace so we can run that speed in a 10K race. We run 26 miles (or close to it) in training so we can accomplish the same feat in a marathon. And so on. Mysteriously enough, however, this principle all but evaporates when it comes to one aspect of marathon training, even among the most ambitious and dedicated. Conventional training wisdom places due emphasis on the need to approximate the marathon distance and pile on the mileage in general. We’re also encouraged to keep up our speedwork while training for marathons, so that our legs don’t become lazy. But most typical schemes blithely ignore the essential need in training for your best marathon—the requirement that training conditions mimic racing conditions.
When Ronaldo da Costa, a relatively unheralded Brazilian, destroyed the men’s world marathon record with a 2:06:05 last fall, he revealed that the key ingredient of his training was performing time trials of 28 to 36 kilometers (about 17.5 to 22.5 miles) at or near marathon pace. The running community was shocked by da Costa’s gusto; I wasn’t. Since taking up the marathon in 1994, I have become convinced that such efforts, thought too taxing by conventional wisdom to fit into most training regimens, are essential to top performances at the 26.2-mile distance.
Alberto Salazar, the former American marathon record holder and a 10K standout before that, once claimed that the basic difference between training for the 10K and training for the marathon was the addition of the long run. Well, that may be true, but I don’t think Salazar had da Costa-like 20-milers in mind. I contend that long, intense training bouts are necessary for marathon success and that all else is secondary. Arguing with a 2:06:05 is difficult. This article spells out the whys and hows of this somewhat unorthodox—but physiologically sound—marathon-training plan.
Why It Works
Pete Pfitzinger, a two-time U.S. Olympic marathoner and an exercise physiologist and coach, agrees with the idea that just adding 20-, 22- or 26-mile runs to a 10K training schedule may not be enough to position most of us for a great marathon. “The argument for [long time trials] is that you are specifically training your muscles to work for a long time at marathon pace, so that fuel utilization and muscle-fiber recruitment are simulating race conditions,” Pfitzinger says. He explains that the ratio of fat to carbohydrate burned as fuel is specific both to pace (a higher percentage of fat is burned at marathon pace than 10K pace) and the duration of the run (the longer the run, the higher the percentage of fat used). Also, as muscles tire the trained athlete is able to call upon still-fresh muscle fibers in order to maintain a given effort level. Faster running trains the body to utilize fuel in certain ways; long runs teach it to recruit additional fibers when muscles tire. But only long, hard efforts maximally train all these systems at once.
The main drawback of such workouts is that they are extremely taxing, both physically and mentally. Who wants to run that far and fast in training? Won’t such exploits bring on days of exhaustion, like that experienced after a marathon? Won’t they increase the risk of injury? And how do you know they will work just because they did for one skinny Brazilian?
Those are valid questions. Let me suggest first of all that any marathon-training program that “guarantees” success is not worth your attention. Sure, there are sound programs out there, and following one is better than having no structure to your training. But there are so many variables involved in marathon preparation and the race itself, it’s folly to “promise” a great race to every competitor. Furthermore, the marathon is not for the meek. If you are training merely to finish, fine; your slow 20- to-26-milers will carry you through. If you want to push beyond the bounds of your current performance standards, though, you’ll need a little more oomph.
Endurance is built over a period of years and is largely limited by our ability and desire to tolerate exhaustive training; as a result, many of us find the marathon perversely gratifying—if we remain healthy (and that’s a big if), we can continue to improve while our performances at shorter events stagnate. To do so, however, we need to narrow our focus and keep in mind what it is we’re really aiming for.
What Kind of Workouts?
The program that has produced the best results for me includes a minimum of three runs of at least 15 miles in the 10 to 12 weeks before my goal race. I do them at or slightly faster than goal pace.
I’m convinced this has made me a better marathoner. When I was training for my first marathon, I averaged 90 to 100 miles a week over a three-month span, logged five runs of 22 to 25 miles at 6:20 to 6:40 pace, and did enough speedwork to crank out a still-standing road 5K PR of 15:25. I tapered significantly over a three-week period. I thought all the elements of a 2:30 marathon were in place. But race day arrived, and although I never “crashed,” those 5:45 miles became impossible for me to hold somewhere around mile 18. I wound up at 2:39—6:05 pace.
Training for the distance the next fall, I added a twist: On three of my six runs over 20 miles, I threw in the last 10 on the track at 5:45 pace. Race day produced a 2:33 this time. In 1997 I again tried something new: incorporating 20K, 25K and 30K track segments at or just below 5:40 pace into longer runs. These workouts were spaced three weeks apart. They took a considerable toll, but I treated them as “races” in their own right and, for three or four days afterward, rested altogether or jogged four to six miles. That fall I ran 2:30 in less-than-perfect conditions.
Last fall, because I was racing more often than usual, I swore off these long, hard workouts while training for the Hartford Marathon. I still did plenty of “typical” long runs and knocked off a 25:17 8K—my first PR at a short distance in years—the week before the marathon. With that sort of speed, sub-2:30 again seemed like a sure bet. Around mile 19 in Hartford, though, right on 2:29 pace, my form faltered. Again, I didn’t crash—my mileage base and 20- to 26-milers probably guarded against that—but I gradually slowed to 6:20 pace and wound up with another 2:33.
It’s worth noting that my average weekly mileage, total number of long runs and the frequency and type of shorter, interval-style workouts were consistent from season to season.
Alex Tilson, a Harvard M.B.A. student, is a disciple of the long-hard run philosophy. [NOTE: After a quarter-century of reflection, I’ve realized that this should have been “hard long-run philosophy.”] Prior to recording a 2:29:10 at the BayState Marathon in 1997, Tilson had built what he calls “a great all-around endurance base.” But his training lacked specific ingredients. While training for the same race in 1998, Tilson, aiming for a sub-2:22, performed five half-marathon time trials at 1:11, often incorporating these into 27- to 30-mile runs. He did one 20-miler on the track at 5:25 pace—exactly the speed his marathon goal would require. He averaged 90 to 95 miles a week.
“I tried to run a 31-flat 10K because I was sure I’d need that sort of speed for a 2:22,” Tilson relates. “In fact, I never broke 32.” But he was exquisitely focused on his goal. “Don’t think you can show up on race day and do something you’ve never done before,” he insists. “A lot of people think they can draw on some special magic because it’s ‘race day,’ but you need to be honest with yourself.” After his long track sessions, Tilson usually took to the pool for a day or two to give his legs a break. Last October, with the temperature rising above 70, he ran a 2:22:45.
One of da Costa’s workouts includes a 30K time trial in which each 10K segment is 30 seconds faster than the last; the final 10K stretch is done at marathon race pace. Another is 5 x 5K, with the first 5K done at marathon pace and each successive one 10 seconds faster. The basic principle is the same: A significant proportion of training is done at or near race pace for the goal distance. As such, it’s really no different than a mile, 5K or 10K training regimen.
There are variations on the theme. Some of Pfitzinger’s athletes run 12 to 15 miles at the pace of a typical training run, then pick it up to race pace in the last five miles; the idea is to shorten recovery time. Other times, they’ll run eight to 10 miles at goal pace (with a warm-up and cool-down, of course). But Pfitzinger agrees that longer, more intense bouts are effective.
“There is a mental and physical recovery involved,” he cautions, but says it would it be useful to do such workouts every second or third week.
Foreign athletes have long had a leg up on Americans when it comes to long, hard workouts, according to Mark Conover, a 1988 Olympic marathoner. “The standard stuff we all learn, like six times a mile or five-mile fast continuous runs, seems ingrained in Americans,” Conover says. “That’s all I did for my marathon preparation, while others like [Gelindo] Bordin and the Africans hammered out some quality 40K-plus runs. They run 2:07 to 2:09 and win medals and prestige, while Americans continue to suffer.”
Boring is Better
Where your long, hard runs should be done is a matter of personal taste; I’ve always done mine on the track. This offers the advantage of close pace monitoring, and a pair of others as well: minimal pounding and the ability to set up aid stations for ample carbohydrate intake. These two factors are vital in training runs of this intensity, because recovery from these bouts has been shown to be strongly dependent on both. Many athletes training for marathons tend to fall into chronic states of low-grade glycogen depletion. (Glycogen is the body’s storage form of carbohydrate.) This clearly limits their ability to do the workouts they aspire to, which can erode confidence and throw them into moods that annoy everyone around them.
“The body stores carbohydrates most readily in the first hour after exercise,” Pfitzinger notes. “Part of my personal secret to being able to train more than other people was that I intuitively sensed that it was important to get the carbs in right after the long runs.” A still better option is to consume as much as you can during your long runs. In addition to the physiological benefits, it prepares you for race-day conditions, when you’ll be taking in fluids and carbohydrates while running at high intensity—a valuable skill to master in itself. The choice of fluids, energy gels or other sources is, in my experience, relatively unimportant. Stick to what you find most palatable; as long as you’re taking in enough, your body will find a way to put it to use.
Of course, some would simply choose to run races in the half-marathon to 30K range for these workouts. Pfitzinger used to sometimes run 30K or 20-mile races for marathon training, and twice ran the first 20 miles of the Boston Marathon at 5:15 pace, about 5% slower than his goal marathon pace. Bill Rodgers did the same. The primary drawbacks to this strategy (besides entry fees) are 1) that you might be jolted by the competitive milieu into running harder than you should, defeating the workout’s purpose; and 2) downhills, particularly late in the race.
That raises the issue of pounding. Ask any beleaguered soul who’s struggled to finish the Boston Marathon course—famed for its uphills but more noteworthy for the placement of the downs—and suffered terrible leg soreness for a week or more afterward, as I did in 1996. It’s the breakdown of muscle fibers that limits our training in the first place; that’s why runners can’t do the routine four- and five-hour workouts put in by swimmers, cyclists and cross-country skiers.
Still, many runners like to do hilly long runs, believing it will “toughen them up” and provide added benefit. But as Pfitzinger notes, “Downhill running causes the most muscle damage because of the large amount of eccentric [braking] forces involved.” If you’re training for a fast marathon, you’re probably going to select a flat course, so it seems reasonable to stick to flat courses in training. Even if you’re training for a course with lots of downhills, like Boston, Pfitzinger says that while it’s a good idea to incorporate hills into your long runs—he cites the consistent success of Rodgers at the Boston Marathon after training extensively on the course—it’s probably best not to tackle really mountainous terrain in training, because the pounding may prove too much.
If you find a track too stultifying, a flat dirt or short-grass surface (which would probably involve lots of “laps,” too) would be OK. A flat road course might work, but the surface, traffic and increased difficulty of getting fluids into your body make this a sub-optimal choice.
The Route to Recovery
Ideally, you should be able to recover from a long, hard workout in about five to seven days. Until then, you won’t be able to do much more than putter along at a pace considerably slower than your typical training speed; pool running is another option. Most runners are paranoid about taking that much “easy” time, but rest assured: Your body is busily repairing itself—if you allow it to.
Da Costa, in contrast to the vast majority of marathoners at his level, incorporates days of full rest into his marathon training. Most every runner I’ve known who has adapted this type of training plan acknowledges the need for total-rest days. I believe that one or two days off from running after any hard run exceeding 15 miles is a good idea in order to avoid injury or burnout. I’ve never been hurt doing this kind of training; my one and only significant injury was a stress fracture incurred when I tried to boost my mileage over 120 a week.
Racing, Mileage and the Taper
When undertaking this training plan, you’ll want to shift emphasis away from short, intense speedwork and short races. The reason is that this type of training, even done wisely, will leave you with a burden of fatigue, even if you don’t always feel it. Your performance in interval workouts and shorter races is likely to suffer, leaving you to wonder if this plan really works. But I’ve always found myself ready to go when the marathon rolls around.
High mileage—whatever that means to you—is important, but its distribution is just as vital. If you can tolerate 90- and 100-mile weeks, go for it. I’d advise scaling back every third week or so to give yourself a break. It probably isn’t useful to do anything long or intense within seven to 10 days of the marathon, though I’ll sometimes do eight or 10 miles at marathon pace 10 to 14 days before the race. This is the time to focus on stretching, hydration, adequate carbohydrate intake and catching up on sleep. I always cut my mileage and intensity drastically in the last two weeks before the race; most fast running I do at this point is at marathon pace.
All Good Things Must End
This program is geared to anyone who’s willing to put all his eggs in the marathon basket. I’ve seen too many people for whom this style of training works not to be convinced it’s valid. But as noted earlier, there are no guarantees in the marathon. Also, the type of workouts suggested here aren’t exactly fun to perform. No one I know enjoys running 50, 60 or 75 laps on a track. You’ll be tired a lot, and you are unlikely to race well at shorter distances.
Still, the benefits are difficult to deny. You’ll develop a keen sense of pace. The way you’ll feel at the end of those sessions provides a great rehearsal for the fatigue that creeps up late in a marathon. On race day, when the finish-line banner is in sight and your goal is well in hand, you might be grateful for those long time trials—or as grateful as anyone can be after maxing out for 26.2 miles.
Kevin Beck is the newsletter editor of the Central Mass Striders. When not training, he’s usually chasing after his yellow Lab puppy, Komen.
The direct inspiration for this column was having recently watched the 1998 New England Cross-Country Championships in Manchester, New Hampshire’s Derryfield Park, at which gathering I heard someone who had to be dad from somewhere in northern Maine yell “HURRY UP, JOHNNY!” at a male-identifying boy who was well within the top twenty with less than 200 meters to go and, perhaps more to the point, already hurrying like the fucking dickens toward the chute with what appeared to be singular focus. As funny as this struck me on its own, had the exhortation not reminded me of a famous (and famously ad-libbed) Jack Nicholson line in 1980’s The Shining, I might not have been moved to write anything about my impressions that afternoon.)