David Epstein

Regular readers of this blog will know that I recently read and greatly enjoyed David Epstein’s book Range: Why Generalists Triumph in a Specialized World. It inspired my post about why foxes make better coaches than hedgehogs, and it inspires the present post about why a power meter is like a thesaurus.

There’s an interesting section in Range on the topic of analogical thinking. Here’s a bit of it:

Deep analogical thinking is the practice of recognizing conceptual similarities in multiple domains or scenarios that may seem to have little in common on the surface. . . Analogical thinking takes the new and makes it familiar, or takes the familiar and puts it in a new light, and allows humans to reason through problems they have never seen in unfamiliar contexts.

Learning from Epstein’s book that smart people tend to use analogical thinking to solve difficult problems and generate novel insights boosted my ego, because I am prone to analogical thinking, and have been for as long as I can remember. For proof, look no further than past posts to this blog, such as the one where I analogize pacing to leaping across a ditch, or the one titled The Chased-by-a-Bear Analogy of Mental Fitness, or the one titled The “Price Is Right” Analogy of Marathon Pacing.

For my next trick, I will use an analogy to explain why the best use of power meters and other training devices is to minimize dependency on these very devices for optimal execution of workouts and races. But first let’s talk about thesauruses.

A thesaurus is a handy tool that writers can use to find synonyms of any given word. The synonym-lookup functionality built into word processing applications is another version of the same thing. Whenever you need another word for a given word, the thesaurus is there to serve. Here are some beliefs I have about this tool:

A thesaurus cannot make up for a poor vocabulary.

In principle, the thesaurus equalizes vocabulary across the writing population. With this resource at hand, the writer with the smallest working vocabulary has access to just as many words as the writer with the largest working vocabulary. In practice, however, no amount of reliance on a thesaurus can elevate the quality of a limited writer’s writing to a meaningful degree.

There are many reasons for this. One is that words aren’t just words; they’re concepts. Take the word anodyne, for example. A writer who does not know this word, which is subtle in meaning and has no exact synonym, is unable to think the thought “anodyne,” and is therefore unlikely to go looking for the word even if it’s the mot juste (there’s another one!) in a particular sentence.

Another reason why Bad Writer + Thesaurus ≠ Good Writer is that working vocabulary is only a part of language mastery. Even when they’re not using fancy words, skilled writers are able to craft more artful sentences than less skilled writers, and there’s nothing a thesaurus can do about it. And to that point, there’s a difference between vocabulary and working vocabulary. You can often tell when a writer has used a word they just discovered in a thesaurus. It has the feel of a mad lib—an awkward fit.

A thesaurus can be used to expand one’s vocabulary.

None of the foregoing is meant to suggest that the thesaurus has no value. Of course it does. While it lacks the power to turn a bad piece of writing into a good piece of writing by artificially propping up the writer’s vocabulary, it can be a tool in the process by which a writer improves their vocabulary, gains greater mastery of language usage, and becomes a more skillful writer. Routine writing practice, supplemented by voluminous reading, is certainly more effective in this regard, but a writer who consults the thesaurus whenever they get stuck on a word in the process of reading and writing will accelerate their development.

The best use of a thesaurus is to reduce one’s dependency on the thesaurus.

It follows from the first two points—that a thesaurus has little power to improve the quality of any given piece of writing yet can play a positive role in a writer’s long-term development—that the best use of a thesaurus is to reduce one’s dependency on this very tool to find the right words to use when writing. If you’re having a hard time wrapping your head around the notion that the best use of a certain tool might be to render itself useless, consider the following analogy (I told you I love analogies!).

A few months ago my elderly father fell and broke his leg. For many weeks afterward, he was confined to a wheelchair, but through physical therapy he was eventually able to advance to walking with a walker. This in turn led to further gains in strength that allowed him to graduate to walking with a cane. And it is our hope that walking with a cane will improve his balance to the point where he can walk unassisted.

In this analogy, the purpose of both the walker and the cane is to render their user nondependent on these very tools, just like the thesaurus.

The less one needs the thesaurus, the more one benefits from its selective use.

The difference between using a thesaurus in writing and using walkers and canes for rehabilitation is that the thesaurus remains useful even when it is no longer needed by a writer who has developed a large working vocabulary. In fact, the most advanced writers are able to make the best use of the thesaurus because their overall mastery of language enables them to do so in sophisticated ways.

I have more than forty years of writing experience and I still consult the thesaurus every now and again. When I do, I very rarely (almost never, in fact) encounter a word that’s not already part of my working vocabulary. I use the thesaurus instead to weigh my options when I plan to use a word that has multiple cognates and I wish to pick the one that fits best tonally, or when I’m concerned about overusing a particular word in a given passage and I wish to explore my options for mixing things up, or when I’m aware that a word I’ve used is likely to be unfamiliar to most readers and I wish to explore my options for selecting a more demotic substitute.

There are probably some writers out there, more skilled than I, who take pride in never consulting the thesaurus. But I believe that no writer is truly “too good” to benefit from this tool, and again, the best writers are able to make the best use of it.

Now Back to Power Meters

Everything I’ve just written about the thesaurus is also true of power meters and other endurance training devices. To succeed in the objectives of completing races in the least time possible and performing workouts at just the right intensities to maximize their benefit, endurance athletes must be skilled at perceiving, interpreting, and controlling their effort. Power meters and other devices can aid both objectives, but they do not have the power to guarantee successful race or workout execution when used by athletes who aren’t very good at perceiving, interpreting, and controlling their effort.

The reasons are twofold. First, endurance performance is limited by perceived effort, which power meters can neither measure nor regulate. Second, each race and workout is unique. Athletes begin each workout and race in an overall state that is slightly different than any past state, and the situation they encounter is at least slightly different from any they’ve encountered in the past. This makes it impossible to predict what optimal execution of the race or workout will look like, requiring the athlete to be reactive and adaptive. Consider the common scenario of an athlete who experiences a rough patch during a long race. Athletes who make the best of these moments and thereby salvage their race do so by backing off, gathering themselves, controlling their thoughts and emotions, and perhaps taking practical measures such as consuming extra nutrition or dousing themselves with cold water. Then, when they feel better, they pounce. Can a device do all of this for an athlete? Absolutely not.

In workouts, the stakes are lower, but an athlete who is able to meet their body where it’s at each day and execute every workout optimally will gain significantly more fitness over time than an athlete who’s a slave to their device. Consider the following workout:

15:00 warm-up
5 x 5:00 @ Critical Power/3:00 active recovery
15:00 cooldown

Critical power is defined as the highest output an athlete can sustain in a relatively stable metabolic state, which is equal to the highest output an athlete can sustain for 20 to 30 minutes. Sounds very scientific, but it’s impossible to pinpoint this threshold with exactitude outside of an exercise laboratory, and in any case it changes slightly from day to day depending on fitness, fatigue, and other factors. What’s more, there is no evidence or reason to believe that hitting critical power with absolute precision in workouts offers any more benefit than getting close to it. From my perspective as a coach, the athlete for whom the example workout is prescribed will get the most out of it if they maintain a very steady output within and across the five repetitions, with a general trend toward increased output as the workout unfolds, and if the athlete’s perceived effort rating is above 9 and below 10 on a 1-10 scale during the final rep. They will pay attention to their device, but in a peripheral way, using it merely to track their performance and make small adjustments. Only an athlete who is skilled at perceiving, interpreting, and controlling their effort can pull this off. An athlete who lacks these skills but “knows” their critical power and tries their best to stick to it throughout the workout is unlikely to get the same result.

As with the thesaurus, though, power meters and other devices can be helpful in developing the ability to self-regulate effort and pacing. My book On Pace: Discover How to Run Every Race At Your Real Limit offers guidelines on how to use training devices to overcome dependency on these very devices for effective pace and effort regulation. If you want a specific tool you can put into use immediately, try this workout.

Again, the endpoint of this process is not throwing your power meter into the ocean but using your device in sophisticated, nondependent ways that are akin to how skillful writers use the thesaurus. I mentioned above that I have more than forty years of writing experience. Well, by the time long covid took away my ability to run in 2020, I had more than 30 years of running experience, and had long since arrived at this endpoint. Here’s an example of the sort of use I made of my Garmin Forerunner before my health imploded:

One of my favorite workout designs that I came up with as a self-coached athlete is what I call a relaxed time trial. It entails running a standard race distance (usually 5K or 10K) at a 95% effort, or 5% slower than I feel I could do in a race. It is important to go completely by feel in a relaxed time trial, ignoring your device, because the idea is to discover your current performance capacity rather than predetermine it. In June 2020, when I was feeling fit but there were no races to run, thanks to the pandemic, I decided to run a relaxed 10K time trial followed later by a solo 10K time trial. I completed the former in 35:04, going completely by feel as the rules require.

Based on this result, I concluded I had a chance to improve upon my old 10K PB of 33:34. So, nine days later, I wheel-measured an exact distance of 10 kilometers and then covered it as fast as I could. In contrast to the relaxed time trial, this time I glanced at my Garmin fairly often, not so much for guidance as to keep a fire lit under me. Whenever my current pace reading slipped above 5:25 per mile, I nudged my tempo upward just a mite. The final kilometer was exquisitely painful, as I was deep in oxygen debt but still on track to achieve my goal. I stopped the clock at 33:25, 4.8 percent faster than my relaxed time trial, having racked up 1K splits of 3:25, 3:21, 3:19, 3:19, 3:20, 3:21, 3:21, 3:21, 3:19, and 3:13. A work of art!

This is the sort of thing an athlete can do when they’ve achieved a degree level of mastery of effort management and pacing. Devices play a role, but they’re not calling the shots; rather, they are being used and exploited in selective ways by the athlete. To reach this level of mastery, less experienced athletes must not blindly indulge their dependency on their training devices, as so many do, but instead use their device with the specific intention of overcoming their dependency.

This concludes my analogy between power meters and thesauruses.

Few reading experiences have been more intellectually validating for me than the one that’s being supplied to me currently through David Epstein’s Range: Why Generalists Triumph in a Specialized World. I chose it in the hope that it would inform my longstanding belief that generalists (i.e., creative problem solvers) make better coaches than specialists (i.e., people who have a lot of specialized knowledge), and in this regard it has exceeded my expectations.

With a veritable avalanche of science and real-world examples, Epstein demonstrates that “foxes,” who see the big picture, perform better in complex environments than “hedgehogs,” who focus on details. “Beneath complexity,” he writes, “hedgehogs tend to see simple, deterministic rules of cause and effect framed by their area of expertise, like repeating patterns on a chessboard. Foxes see complexity in what others mistake for simple cause and effect. They understand that most cause-and-effect relationships are probabilistic, not deterministic. There are unknowns, and luck, and even when history repeats, it does not do so precisely.”

Among the most compelling examples Epstein cites is a prediction tournament organized by the Intelligence Advanced Research Projects Activity (IARPA). Teams of forecasters were asked to bet on the likelihood of certain geopolitical events happening at particular future times. Most of the teams were made up of highly specialized intelligence experts, but all of them got their asses handed to them by a team of generalists called the Good Judgment Project.

To be clear, Epstein does not contend that foxes are better than hedgehogs at everything. They’re not. His reference to chess in the quote above isn’t arbitrary. Specialists, who through heavy repetition acquire vast stores of information that they use to identify patterns and select optimal responses, tend to perform better in so-called kind learning environments like chess, which are a lot simpler than most real-world situations, which present “wicked” learning environments. So, the crucial question for us is this: Does endurance training represent a kind learning environment, like chess, or a wicked learning environment, like geopolitics?

There are some coaches with specialized knowledge of exercise physiology who seem to believe that specialized knowledge in exercise physiology is necessary and sufficient for effective endurance coaching. In their view, endurance training is simple and straightforward. All you need is a thorough understanding of how the body adapts to various training stimuli and how these adaptations translate to competitive performance, plus a steady supply of relevant data from the athlete as they progress through the training process, and from there it’s paint by numbers.

I disagree. In my experience, endurance training isn’t nearly so predictable, even with the aforementioned knowledge and information. There are far more unknowns than knowns in real-world exercise physiology, and there’s far more to coaching than managing athlete physiology. Like psychology.

Recently, NAZ Elite coach Ben Rosario told me about a workout he designed for Aliphine Tuliamuk ahead of the U.S. 20K championships. It was a fairly complex hodgepodge of hill surges, tempo segments, and whatnot. “No physiology-type coach would ever give an athlete that workout,” Ben said. Why not? Because there was no clear physiological rationale for the format Ben had come up with. What mattered to him was that it was appropriately challenging and reasonably specific to the 20K race distance, and more importantly, its everything-but-the-kitchen-sink structure was sure to keep Aliphine on her toes, simulating the unpredictable dynamics of racing in a way that, assuming she performed well, would leave her feeling confident and prepared.

What Ben understands that hedgehogs don’t is that athletes are more than the sum of their blood lactate levels and heart rates. They are sentient beings who experience training, and how they experience it matters. By treating athletes as contextualized human beings rather than as tidy closed systems of physiological processes, fox-type coaches are able to deploy a broader range of tools to help them succeed.

It so happens that I’m writing this post on the day Emily Sisson set a new American record of 2:18:29 at the Chicago Marathon. In an interview conducted before the race, Emily’s coach, Ray Treacy, described how her training had evolved over time to become less “textbook” and more idiosyncratic as they learned from how she responded to standard methods like exposure to high altitude (which didn’t agree with her). Hedgehog coaches who adhere to a rote method of training athletes based on specialized knowledge and data have a hard time figuring out how to get the most out of athletes like Emily Sisson, and the fact of the matter is that all athletes are like her, which is to say that all athletes are unique individuals in unique situations. Indeed, of the six athletes I coach at this time, all six have at least one unconventional element in their training that was arrived at through experimentation and learning.

Credit: letsrun.com

Foxes are much better than hedgehogs at adapting in the face of unexpected events. As Epstein puts it in reference to the aforementioned IARPA prediction tournament, “When an outcome took them by surprise . . . foxes were much more likely to adjust their ideas. Hedgehogs barely budged.” I’ve seen unfortunate examples of this rigidity in endurance sports. When Hedgehog coaches fail with an athlete, they blame the athlete instead of themselves. As they see it, the only reason their narrow, inflexible, one-size-fits-all methodology didn’t work was that the athlete brought messy complications that didn’t allow the methodology to demonstrate is infallibility. It never crosses their mind that messy complications are the norm, and because of this coaches must be flexible and creative in their approach, treating each athlete as a unique case.

You might be wondering why a coach can’t be part fox and part hedgehog, combining the virtues of big-picture thinking and specialized knowledge to coach more effectively than either pure foxes or pure hedgehogs. This makes sense in principle, but as a practical matter it’s next to impossible to be both a fox and a hedgehog. The reason is that the fundamental difference between foxes and hedgehogs is not that foxes know a little about a lot of things and hedgehogs know a lot about a few things. It’s that they have completely different ways of thinking.

Hedgehogs hate ambiguity and uncertainty. That’s why they try to absorb as much knowledge and information as possible—to eliminate ambiguity and uncertainty so that they always know immediately what to do in any situation. Foxes put reason ahead of information—process ahead of content. They accept that complexity and unpredictability are irreducible realities, so they don’t worry about always knowing immediately what to do in any situation. They draw confidence instead from knowing how to figure out what to do. And for this reason they make better coaches.

Interested in learning more about coaching? An 80/20 Endurance coach certification program is on the way. To be the first to know about updates regarding the course, sign up for our waitlist here. And just between us… we’d recommend signing up for the waitlist before Black Friday (November 25th, 2022) *wink*.

$ubscribe and $ave!

  • Access to over 600 plans
  • Library of 5,000+ workouts
  • TrainingPeaks Premium
  • An 80/20 Endurance Book


30 day money back guarentee

For as little as $2.32 USD per week, 80/20 Endurance Subscribers receive:

  • 30-day Money Back Guarantee