Shownotes

John's background

02:36 -

• I'm a coach, a writer, and a scientist—usually in that order. I’ve been coaching for over ten years, starting at the high school level and now working primarily with post-collegiate adult athletes.
• I earned my PhD in Human Performance from Indiana University two years ago, and I’ve been writing about running on my website, Running Writings, for almost 14 years now. I started the site fresh out of college, back in 2011.
• Currently, I split my coaching about 50/50 between athletes based locally in the Twin Cities and others I coach remotely around the world. Most of my athletes are road runners, although not exclusively. I coach distances ranging from 800 meters to the marathon.
• Right now, around 70% of the athletes I work with are focused on the marathon, but I also have a couple of 800-meter runners, some 1500/mile athletes, and several focusing on the 5K and 10K.
• Academically, while my PhD is in Human Performance, my specialization is biomechanics. I’ve always been deeply interested in how people move when they walk and run, and how that movement might relate to injury risk.
• I felt I had a solid understanding of how to train athletes for performance, but I wanted to better understand why so many runners struggle with injury. That’s what motivated me to go to grad school—to dive into the biomechanical causes of injury and how training might be adapted to prevent them.
• My research included work on injury in general, but my dissertation focused specifically on wearable devices—things like Garmin chest straps, Stryd pods, or smart shoes—and the data they generate. I was interested in whether that data could give insight into tissue stress in areas like the Achilles tendon or the knee.
• One of the most useful takeaways from that research is the understanding that the faster you run, the more mechanical stress—or “damage”—you do per mile.
• Most runners intuitively understand this: if today you run eight miles easy, and tomorrow you run two miles easy, four miles at tempo, and then two miles easy, you know the second day is harder on your body. But both runs are technically eight miles.
• So if you’re tracking training load, even if it’s a simple system, it’s important to account for speed—not just distance. Faster running amplifies the strain per mile.
• You might ask what the exact exchange rate is—how much more damage a faster mile does compared to an easy one—but that’s still an open question in the research. There’s likely some nonlinear relationship, a sort of positive curvature, where increases in speed result in disproportionately more strain.
• Another practical insight is that for runners with a history of injury, it can be worth working with a physiotherapist to make small adjustments like increasing cadence or reducing vertical oscillation at a given pace.
• The goal is to reduce the stress placed on the body with each stride, which could help in staying injury-free over time.

Footwear and impact on injury

07:45 -

• One of the more surprising things I learned during graduate school was how little evidence actually supports many common paradigms around footwear.
• For instance, the idea that people with high arches should wear a specific type of shoe, or that pronation dictates the type of shoe you need, doesn’t hold up well under scientific scrutiny.
• During my time in grad school, we witnessed the emergence of super shoes, which helped address questions around footwear and performance. However, the issue of footwear and injury prevention remains largely unresolved.
• Interestingly, one area where there is solid evidence is the importance of comfort. Research has shown that selecting a shoe that feels comfortable is more beneficial than choosing one that does not. Beyond that, though—whether it's heel-to-toe drop, arch support, or other design features—there’s a lack of strong, evidence-based guidelines.
• This gap is partly because academic biomechanics isn’t well positioned to solve such questions in a robust and comprehensive way. So, in the end, my best advice when it comes to footwear is simply this: wear a shoe that feels comfortable.

John's daily work

09:24 -

• A typical day for me starts with checking coaching-related messages that have come in overnight. I often begin the morning by reviewing and updating athlete schedules. Most of my daytime hours are currently dedicated to a large writing project that I’m working hard to complete.
• In the evenings, if time permits, I shift my focus to research—primarily scientific coding. This includes developing tools like the interactive calculators and visualizations on my website. While these features may appear simple on the front end, there’s a significant amount of coding behind them.
• My daily routine is a mix of coaching, writing, and technical development. Occasionally, especially in the warmer months, I meet up with athletes in person for workouts, although that’s less frequent during the winter.
• One aspect of my coaching that might surprise people is how much emphasis I place on effort and correct execution. When I deliver a training plan, it includes not just the workout calendar, but also a detailed set of instructions.
• I explain how the workout should feel, how to set it up, the intended difficulty, and the perceived effort target—whether it's something like 10K pace or marathon pace.
• While my website and tools might give the impression that I’m focused on precision and calibration, the reality is more nuanced. I’d estimate that in around half of the workouts I prescribe, I encourage athletes to focus less on exact paces and more on tuning into the right effort. If they can execute the effort correctly, they’re achieving the goal of the session.

Running Writings website

11:48 -

• My website began as a kind of brain dump—a place to process everything I had learned and experienced during my college track career. I had just finished my final season, was dealing with injuries, and felt uncertain about whether running would continue to be a central part of my life.
• I wanted to capture the knowledge and thoughts I had accumulated, in case I decided to move on to something entirely different.
• What started as a personal project quickly evolved into a deeper exploration of running, coaching, and sports science.
• I began writing extensively about training theory, injury prevention, and the scientific principles behind endurance performance. 
• As I got involved in coaching, much of my content naturally gravitated toward practical methods and analysis of elite coaching philosophies.
• I’ve written in depth about Renato Canova’s approach—his methodology, workouts, and the principles that underpin the success of his athletes. One of my most-read pieces is a detailed breakdown of Marius Bakken’s double threshold system.
• The platform now offers a blend of evidence-based science, training insights, and coaching perspectives.
• I’ve explored topics such as heart rate responses at different physiological thresholds (LT1, LT2), and I’m currently working on a guide to designing a half marathon training plan from scratch.
• In addition to written content, the site also hosts several of the pace calculator tools I developed as part of my post-graduate work. These tools reflect my interest in making training concepts more accessible and actionable.
• I also run an email newsletter where I occasionally share updates and new content. While the frequency can be a bit irregular, I use it to stay connected with readers who want to dive deeper into the art and science of running.

Website calculators

14:25 -

• The genesis of my work in this area dates back to when I was a high school coach. I often found myself standing outside in the rain, hunched over a notebook or my phone, trying to work through calculations—often to solve practical training questions on the spot.
• That experience shaped my motivation to create tools for coaches in the field: mobile-friendly, intuitive, and fast to use without complicated menus. The goal has always been to build something usable on the roadside, under pressure, in real time.
• The first tool I developed was a basic pace percentage calculator. For example, if I needed to calculate 90% of a 3:50 per kilometer pace, I wanted something that didn’t require a spreadsheet or mental math.
• This kind of functionality is especially useful for coaches applying the Renato Canova or Italian-style training methodology, where sessions are typically prescribed as percentages of reference paces—like 90% of marathon pace or 105% of it, or 80% of 5K pace.
• When modifying an athlete’s training plan, I often needed to rapidly calculate percentages of a moving target—say, if I believed an athlete’s 5K pace ranged from 3:25 to 3:30/km.
• After graduate school, while learning web development, I built this percentage calculator as a side project, and it unexpectedly attracted a significant amount of traffic. That led me to consider what other science-based concepts could be made more accessible through simple online tools.
• Another tool I created was a grade-adjusted pace calculator. It was based on research from the early 1990s where runners performed on treadmills at various incline and decline gradients while their oxygen consumption was measured.
• Using those data, I fitted a curve that allowed me to estimate the flat-ground pace equivalent for a given uphill or downhill gradient, based on energy demand. 
• This became useful, particularly for marathoners racing on undulating courses—for example, helping them understand that if they sustain a 4:00/km effort on the flats, it might translate to 4:25/km while climbing a significant hill at the same exertion level.
• Another favorite tool of mine is the wind calculator. After Eliud Kipchoge’s Breaking2 project, there was renewed interest in how wind and drafting affect performance.
• Building on studies conducted in wind tunnels or using pulley systems to simulate drag, I realized I could merge this data with well-known aerodynamic principles.
• Using inputs like wind speed and running velocity, it became straightforward to calculate the wind’s effect on pace.
• Furthermore, I integrated terrain-based modifiers. Wind forecasts are typically given at a height of 10 meters, whereas a runner’s torso is closer to 1.5 meters off the ground, where wind speed can be significantly lower depending on the environment.
• By referencing civil engineering research on pedestrian-level wind exposure, I implemented coefficients for different environments: countryside, suburbs, and urban areas.
• This allows for practical adjustments based on whether an athlete is running in an open rural area or a city shielded by tall buildings.
• I personally used this calculator during lactate testing on a windy day at the track. It helped normalize paces to a calmer-weather equivalent for more accurate analysis. One version I’m still developing is tailored specifically to 400-meter tracks, where wind direction and speed vary as the runner moves around the loop.
• It requires segment-by-segment calculations but follows the same principles—I just haven’t coded it yet.
• In addition to these, I created a critical speed calculator that also integrates VO₂max pace and threshold pace estimation. Unlike traditional critical speed calculators—which require multiple maximal efforts across a range of distances—I designed this one to work with just a single race performance, such as a 5K or 8K time.
• It provides a best estimate of critical speed, along with upper and lower bounds (CS+ and CS−) based on population-level data.
• My motivation here was pragmatic. Outside of track athletes, it’s rare to have athletes perform several all-out efforts over varying distances. So I built a model to estimate not just a critical speed from one result, but also its uncertainty—expressed as a confidence interval.
• That approach helps reflect real-world physiological variability between athletes with identical performances but different underlying metabolic profiles.
• In terms of terminology: VO₂max pace corresponds to the slowest pace that will eventually elevate oxygen uptake to VO₂max—typically represented by CS+.
• Conversely, threshold pace is linked to CS−, the fastest speed that maintains metabolic steady-state, often aligned with LT2 or maximal lactate steady state. The calculator uses these bounds to give practical estimates for training zones without relying on laboratory testing.
• To build the underlying model, I compiled a dataset of approximately 3,000 athletes, primarily from high school, collegiate, and masters-level track and road racing.
• These athletes had performed multiple races within a consistent time window, enabling robust curve-fitting of the critical speed model across 800m, 1500m, 3K, and 5K distances.
• For each athlete, I recorded the CS, CS+, and CS− values derived from real data, then applied modern statistical methods to estimate percentile bounds for any given performance time.
• This is a more rigorous alternative to traditional threshold calculators, which often provide only a mean value based on population averages.
• The problem with such methods is that they implicitly assume a 50% chance of the athlete being in the correct training zone. My approach instead provides ranges that reflect 90–95% likelihoods—offering more confidence in the prescribed paces aligning with the intended metabolic stimulus.

Coaching style and methodology

34:01 -

• A robust aerobic fitness base is, in my view, a critical foundation for achieving race-specific fitness. To illustrate how I’ve arrived at this perspective, it may be helpful to outline my coaching influences in chronological order.
• Early in my career, I was heavily influenced by Arthur Lydiard’s approach, which emphasized high mileage, a substantial aerobic base, and strict periodization. Over time, as I engaged more deeply with scientific literature and coaching philosophies, I became particularly influenced by Renato Canova’s methodology.
• Canova advocates for a race-oriented framework: starting from the performance goal and working backward. For example, if an athlete aims to run a 40-minute 10k, they need to sustain a pace of four minutes per kilometer.
• From there, we identify the workouts necessary to support that performance and determine the foundational capabilities required to execute those sessions effectively.
• If the athlete must run 5 x 2000 meters in eight minutes each (i.e., 4:00/km pace), we must ensure they’re capable of completing longer sessions—such as 15 to 20 km at an easy pace or 12 to 16 km at moderate intensity—to underpin that workload.
• In practical terms, much of the training I prescribe may resemble traditional threshold work and long intervals. However, one distinctive element in my coaching—something many of my athletes have not previously encountered—is the incorporation of long, fast runs.
• These may range from 15 to 25 km or longer, especially for marathoners, performed at 90–95% of marathon pace or similar percentages of 5k or 10k pace. The aim is to develop a resilient aerobic engine that can support race-specific work.
• This focus on long, fast runs is one of the key differentiators between Canova’s philosophy and Lydiard’s. While Lydiard’s athletes might have averaged six-minute miles, these were not necessarily run at intensities close to race pace. In contrast, Canova-style long runs are executed at intensities that more closely mimic race demands.
• Another core principle I emphasize is the balance between high stress and high recovery. Over the course of a training cycle—say, six months—I begin with more traditional base work: high-volume training with moderate efforts. As the athlete progresses, I concentrate training stress into fewer but more demanding sessions.
• For example, a 32 km run at 95% of marathon pace is an extremely taxing session. This type of workout is not introduced early, and when performed, it is followed by multiple days of recovery—often three to four easy days—before resuming intensity. This approach fosters both physiological adaptation and long-term durability.
• I also strongly discourage the repetition of identical workouts week after week. Many athletes fall into routines where they repeat the same threshold, VO2 max, and long run sessions every week. I believe this limits progression. Instead, I vary the training stimulus constantly.
• This could involve increasing volume (e.g., from 8 x 1k to 10 x 1k), changing the structure (e.g., 4 x 2k), reducing recovery, or increasing pace. The key is to avoid stagnation and continue challenging the body to adapt.

John's training cycle

40:29 -

• In practice, training plans often need to be adapted to an athlete's weekly schedule. For most people, it's unrealistic to expect them to complete a 30 km run on a Tuesday, for instance.
• Therefore, I typically organize training in two-week blocks. If a long session is scheduled, it's usually planned for the weekend—either Saturday or Sunday—since this tends to be the most convenient time for athletes.
• From there, I structure the surrounding days to support recovery and balance. For example, a long session might be followed by two easy days, then another workout, and so on. In some cases, athletes may have three or four easy days between key workouts.
• Earlier in the training cycle, when the sessions are not particularly demanding, it's common to schedule runs on Monday, Thursday, and Saturday—or even Monday, Wednesday, and Friday/Saturday—with the understanding that none of these workouts are particularly intense.
• Rather than adhering to a rigid weekly cycle, I operate with the mindset that each type of workout follows its own frequency. A long, fast run might only appear once every three to four weeks, while progression or threshold-based workouts may occur every five to ten days.
• I approach planning by considering which training paces need to be addressed, how frequently they should be revisited, when the last session of that type occurred, and how best to fit the next one into the schedule.
• Regarding easy days, their structure often depends on the phase of the training cycle. In base-building periods, most days can be categorized as easy to moderate. However, as the sessions become more intense later in the cycle, the need for deeper recovery increases.
• For instance, after a particularly demanding workout, the following day might involve running at a pace 20–30 seconds per kilometer slower than usual. Then, training volume and intensity can gradually increase again until the next key session.
• If I follow a "micro-cycle," it's likely found in the recovery stretch between workouts. For example, after a Saturday session, I might prescribe a very easy run on Sunday, followed by an easy run on Monday, and then a slightly more moderate or progressive session on Tuesday—potentially including two runs: one easy and one steady.
• Another core principle I emphasize is expanding the athlete’s range of training paces in both directions. A common mistake I see is athletes running at the same pace every day—regardless of whether they feel fresh or fatigued.
• For example, always running at 4:30/km or 7:00/mile, regardless of context. To correct this, I encourage athletes to vary their pace and include workouts that intentionally span a wide range, such as the Kenyan-style progression run.
• In this workout, the athlete starts very slowly—perhaps at 6:00/km even if capable of a 2:30 marathon—and gradually increases the pace throughout the run.
• This helps the athlete become familiar with different intensities, monitor fatigue, and develop a better sense of pacing and internal feedback across various effort levels.

Balance between volume and intensity

45:40 -

• When considering the total training volume required, it often becomes a question of injury risk and practicality. For example, while an optimal training volume for a marathon might be around 200 kilometers per week, time constraints and the risk of injury make this level of training unrealistic for most athletes.
• Instead of prescribing a rigid volume, I work backwards from the demands of the race. For the marathon, the goal is to ensure that the athlete can handle the race distance—42 kilometers—without it being an unmanageable leap from their training.
• This doesn't mean one must run 42 kilometers in training, but rather that running that distance in a race should not feel vastly different from prior training loads.
• In contrast, for shorter races like the 5K, the focus shifts to the early-season volume needed to support longer, faster workouts later on. For instance, if an athlete plans to run 10 kilometers at 95% of 5K pace, that workload should be supported by sessions such as 12 kilometers at 90%, or 15–16 kilometers at 85%.
• To sustain this, the athlete needs to routinely run 15 to 20 kilometers at an easy pace without excessive fatigue. This establishes the necessary easy-run volume during the foundational period of training.
• Regarding periodization, my approach tends to follow a flat distribution of volume and intensity in the early and mid phases of training.
• As the athlete approaches peak race specificity, I shift toward concentrating more intensity and volume into specific, isolated sessions. This is paired with a reduction in easy volume on subsequent days to allow recovery.
• One example of this would be performing a key session like 30–32 kilometers at 95% of marathon pace, followed by only 8 kilometers at an extremely easy pace the next day. Even if the athlete had been doing 20–25 kilometers per day earlier in the block, the recovery day is kept deliberately light.
• For athletes on lower total volume, I may even recommend taking the day off entirely. The rationale is that once the key session is complete, the priority becomes recovery and adaptation—not adding more volume.

Training common mistakes

48:42 -

• In my experience, one recurring issue I observe among athletes is their tendency to repeat what they consider a "magic week" or rely heavily on a set of favorite workouts. For example, some runners might consistently repeat a specific block or insist on sessions such as 6 x 1 km or 8 x 1 km at threshold pace. These routines often become habitual rather than strategic.
• When I begin working with a new athlete, I have them complete a comprehensive questionnaire that covers their training history, health background, injury record, and personal motivations for running.
• One consistent pattern I see is that most athletes are not incorporating certain critical training components. It's rare for someone to begin coaching with me already implementing these elements effectively.
• In particular, I often find a lack of long interval sessions at 8K to 10K race pace. Workouts such as 4 x 2,000 meters or 3 x 3,000 meters are seldom present in their training.
• Similarly, extended fast runs are uncommon. While many athletes include tempo efforts of 20 to 30 minutes, it's rare to see sessions extending to 45, 60, 70, or even 90 minutes at a sustained, fast pace. These longer efforts are essential, especially for distances ranging from 5K to the marathon.
• Another area that’s often neglected is the principle of applying significant stress followed by adequate recovery. Some athletes will schedule long runs on both Saturday and Sunday, which may increase total mileage but does not effectively build the specific physiological adaptations required for racing—particularly for the marathon.
• Simply accumulating easy mileage is not the most effective way to prepare the body for the demands encountered in the latter stages of a marathon.

Alternation between different training models

51:43 -

• During what is commonly referred to as the base phase—but which I prefer to call the general phase—I still calculate workout paces using percentages.
• However, I tend to rely more on physiological parameters during this period. I have developed, though not yet published, a set of percentage ranges that link 5K race pace to key physiological thresholds, including critical speed (CS) and LT1.
• These ranges are outlined in a blog post I have released, and I often use them to design workouts that target specific zones—for instance, sessions that are above LT1 but still below the upper end of the steady-state threshold.
• In these cases, I may prescribe training in a range that corresponds to a certain percentage of 5K pace, tailored to elicit the desired physiological response.
• For workouts intended to target the higher end of the steady state—what I refer to as "steady state max"—I typically use intensities around 95–96% of 5K pace. At this level, there is a high probability (over 90%) that the athlete will be operating in a metabolically unstable state.
• These physiological capabilities are viewed as foundational support for race-specific fitness. For example, during the competitive season, an athlete may need to sustain a metabolic steady state at higher intensities in order to meet the demands of their target event.
• To achieve this, we manipulate training intensities across various zones to develop the necessary physiological traits. Sometimes the focus is on preparatory work that supports adaptations two phases ahead of the key race-specific training.
• In the final six to eight weeks before a race, however, my focus shifts more explicitly toward workouts that are aligned with race pace. For instance, if an athlete aims to run a 40-minute 10K, it becomes imperative—regardless of individual characteristics such as height, weight, sex, or anaerobic work capacity—that they can sustain a pace of four minutes per kilometer. Therefore, sessions during this final period are increasingly structured around that specific pace.
• Regarding terminology, "steady state max" refers to what is also called the maximal metabolic steady state—a physiologically defined threshold representing the highest intensity at which a steady internal state can be maintained.
• This is analogous to a measurable yet indirectly assessed parameter, like core body temperature. Just as different tools (e.g., oral thermometers, forehead scanners, ingestible sensors) offer varying levels of accuracy in estimating core temperature, various tests such as LT2, critical speed, and maximal lactate steady state (MLSS) serve as methods for estimating an individual's steady state max.
• Each method comes with its own trade-offs in terms of invasiveness, accuracy, and practicality. While LT2 testing is relatively simple to conduct, it is less precise than either critical speed or MLSS testing, which generally require multiple laboratory sessions.
• Despite yielding slightly different results, all three methods aim to estimate the same physiological threshold.
• Finally, in addition to using percentages of pace, I also incorporate subjective effort descriptors—such as "10K effort"—when prescribing workouts.

HR training

56:20 -

• I personally do not use heart rate in my coaching practice. While I understand why some coaches and athletes incorporate it into their training, I find there are too many limitations—particularly the phenomenon known as cardiac or heart rate drift.
• This refers to the progressive increase in heart rate during steady-state exercise, even when the physiological intensity remains unchanged.
• A particularly insightful study on this topic was done by a researcher who completed a dissertation around 2018 exploring cardiac drift. In the study, participants were instructed to self-pace their efforts while maintaining a fixed oxygen consumption (VO2).
• They were given real-time feedback, similar to how cyclists monitor power output. The same protocol was repeated using heart rate as the pacing guide. The results showed that when athletes tried to maintain the same heart rate, their oxygen consumption dropped significantly after several minutes.
• This demonstrated that heart rate alone could be misleading, even in controlled lab conditions and cool environments. While heat and hydration issues can worsen the drift, it occurs regardless.
• My concern is that if athletes rely too heavily on heart rate, they may unintentionally slow down and reduce the metabolic intensity of their sessions. The variability in heart rate can also come from factors like mild inclines, accumulated fatigue, or even biomechanical changes late in a session.
• This makes it difficult for me to confidently interpret what heart rate is signaling. Is the heart rate increasing due to rising effort, a barely perceptible incline, or simply fatigue affecting efficiency? The signal is too noisy for my approach.
• In contrast, since most of the athletes I work with are road or track runners, I have the luxury of using pace as a primary metric. With tools like wind and gradient-adjusted pace calculators, I can make fairly precise estimations of true effort. This makes pace a more reliable anchor for training and race execution.
• Moreover, my coaching philosophy centers on developing an internalized sense of effort. I want athletes to build a strong connection between how a pace feels and the actual output it represents. Ideally, I want an athlete to be able to run 10 kilometers at 90% of their 5K pace purely by feel—even if I were to take away their watch.
• That kind of intuitive pacing is what leads to consistent and effective race performance. For that reason, I deliberately avoid fostering a dependency on heart rate and instead focus on cultivating a direct connection between perceived effort and running pace.

Running power

1:00:30 -

• While I own one and a few of my athletes also use them, I rarely incorporate running power into coaching. Unlike cycling, where power meters directly measure mechanical output through torque and crank velocity, running power is not a direct measurement.
• Instead, it serves as a grade- and wind-adjusted pace proxy, derived from data like speed, incline, and wind conditions. Because of this, it doesn’t reflect physical force application—such as push-off ground force—and requires a mental shift away from traditional pace-based metrics. Personally, I find greater utility in pace or grade-adjusted pace rather than this proxy of “power.”
• Despite using Stryd data in my dissertation and being familiar with the technical details, I still don’t use it for coaching. Its limitations become apparent on treadmills, where readings are often unreliable, and on uneven or unpaved surfaces where power tends to be underestimated.
• One potential benefit, however, would be for athletes training predominantly indoors, as it can help capture data on time spent at various paces, which is valuable for injury-prone runners. Tracking how much mileage is spent under specific paces (e.g., below 4:00, 5:00, or 6:00 per km) could help manage musculoskeletal stress more precisely.
• The idea is to build models linking pace zones to injury occurrence using chronic vs. acute load principles. There’s active research in this area working on this concept.
• My own graduate work aimed to create a more sophisticated version of this, estimating biomechanical stress (like Achilles tendon or knee joint loading) using speed, cadence, vertical oscillation, and related metrics.
• Although this approach holds promise, there are significant practical challenges: acquiring reliable data, identifying true injury events, and ensuring complete training logs.
• Still, it remains a promising area of research, and such models could eventually offer valuable insights. As noted, these observational data studies are far easier to get through ethics committees than controlled trials involving tracking injuries in real time.

Durability

1:06:29 -

• Cycling races tend to be longer, so the effects of durability are often more pronounced. However, it's just as relevant in running.
• My current understanding of durability in running is largely informed by a 2023 paper by Andy Jones and his team at the University of Exeter in the UK. Jones was one of the lead physiologists behind the Nike Breaking2 marathon project.
• In addition to the well-known insights about drafting that emerged from that work, they also contributed to our understanding of what Jones refers to as "physiological resilience."
• There’s no perfect term for this concept. What we’re really talking about is how much an athlete's body breaks down over the course of a long effort. In practical terms: does their effective fitness degrade significantly, or only slightly, during a long run or ride?
• In both cycling and trail running, people often use the term “durability” to describe this, though in running, the word can also refer to injury resistance—how well someone tolerates high training volumes without getting hurt. While that's certainly valuable, it’s a separate concept.
• In this context, “durability” refers to physiological performance degradation. Resilience isn’t a perfect term either, because it sometimes carries psychological connotations. What I’m referring to is: does VO₂max decline over time? Does running economy deteriorate? Does lactate threshold (LT2) drop as a percentage of VO₂max after extended efforts like a 25 to 40 km run?
• From my perspective, running economy is the most significant component. Let’s say a runner, when fresh, can sustain a steady-state pace of 4:00/km.
• Over the course of a marathon, due to factors like muscle damage, glycogen depletion, and the accumulation of metabolic byproducts—even without going anaerobic—efficiency drops. By 30 or 35 kilometers, the same metabolic output might only yield 4:25/km pace. That’s a 10% slowdown, which is on the higher end of observed deterioration.
• On average, most runners experience about a 5% decline in running economy over a marathon. Some runners are remarkably stable, finishing nearly as efficient as they started. Others can lose up to 10% of their efficiency.
• Beyond running economy, there's also likely a decline in VO₂max and steady-state threshold relative to VO₂max during prolonged exertion.
• However, these are harder to measure. It's relatively easy to have someone run 30 km, then assess their running economy on a treadmill. But it's much more difficult to conduct a full VO₂max test immediately after such a long run.
• When athletes come to me disappointed with their marathon performance, one pattern I often see—especially if their 5K and 10K times are much stronger—is a progressive breakdown in performance. They might feel good at 15 km, okay at 20, and then things begin to fall apart at 25 km.
• This could be pacing, but often, it's a physiological issue. Their body is gradually deteriorating, and what was once a comfortable steady-state pace becomes unsustainable—effectively turning into 10K pace by the final third of the race. At that point, the body can no longer cope, and performance collapses.
• This framework—viewing durability as physiological resilience—has become one of the most valuable tools in my coaching over the past few years.
• It offers a clear explanation for why some athletes excel at shorter distances but underperform in the marathon. Understanding and training this aspect of fitness has made a significant difference in helping athletes close that gap.

Factors that impact durability

1:12:02 -

• The way I’ve been thinking about this topic ties back to what I mentioned earlier regarding the key contributors to the deterioration in running economy. I suspect that VO₂max and steady-state max also play a role, although I’m less certain about the exact mechanisms involved there. In contrast, I have a clearer understanding—at least in part—of the mechanisms behind the decline in running economy.
• From what I’ve observed and studied, the three main contributors are: muscle damage, glycogen depletion (both general and local within the muscle), and central fatigue.
• Central fatigue is particularly fascinating—it refers to the brain and spinal cord gradually reducing the capacity of the muscles to contract as high-intensity efforts continue.
• Neurological studies on cyclists show that when one leg is used to pedal hard, the brain's signaling to that leg becomes inhibited, whereas the unused leg doesn't experience this inhibition. This highlights the specificity and complexity of central fatigue.
• Addressing each of these three factors requires targeted strategies:
• For muscle damage, long runs are a classic intervention, but I’ve also been incorporating more long runs through rolling hills. Laboratory studies often simulate muscle damage by having subjects run downhill on a treadmill, which leads to significant degradation in running economy.
• However, consistent downhill training can build resistance to that damage. Based on that, I’ve been experimenting with runs through hilly terrain and, in some cases, structured downhill intervals.
• For example, instead of a traditional hill session with fast uphill efforts and recovery jogs down, we might reverse the structure—running fast downhill to create muscle stress. I must emphasize that this is still experimental and not without risk.
• On the glycogen depletion side, more aggressive fueling—both the day before and during the run—can certainly help. I've also experimented with depletion runs in a few cases.
• These involve training in a fasted state, without breakfast or mid-session fueling. While not advisable for most people, this approach can place a substantial demand on the body’s glycogen storage systems and potentially enhance its capacity over time.
• As for central fatigue, I believe that long, fast runs at around 90–95% of marathon pace are particularly effective. Unlike low-intensity runs, these efforts place the body in a fundamentally different biochemical state—above LT1—which better simulates race conditions.
• Running 30 to 35 kilometers at an easy pace won’t replicate the internal stress of a marathon pace effort. These near-race-pace long runs seem key to improving resilience against central fatigue.
• I’m also intrigued by the training of top American marathoners like Conner Mantz and Clayton Young. They often structure workouts with 10–15 km at about 90–95% marathon pace—sometimes including hills—followed by faster intervals on the track. This hybrid format could target both durability and speed, though I haven’t used it extensively enough to make a strong recommendation yet.
• Regarding overall volume, I think the endurance community—particularly in cycling—has shown a growing recognition of the benefits of long-term high training volume for durability.
• Although conclusive evidence is still developing, it’s clear that runners doing 100 to 120 miles per week experience fewer issues in the final stages of a marathon compared to those training at lower volumes. 
• Of course, high mileage can be challenging to achieve for many runners, but it does offer clear physiological benefits.
• Increasing volume may be a “blunt tool,” but sometimes that’s exactly what’s needed. Higher mileage leads to greater muscle damage and adaptation, and there’s good evidence that it also boosts glycogen storage capacity.
• In many ways, I see overall training volume as the necessary foundation upon which to build these more refined, specialized interventions.

Ironman marathon training

1:19:41 -

• I don’t consider myself a triathlon expert, but when I approach the topic from first principles, a few key points stand out. The most obvious is the issue of fueling and glycogen depletion.
• Going into the run after working hard for five hours places the body in a significantly fuel-constrained state. This naturally elevates the importance of proper fueling—not just on the bike, but also during the run. It also limits how aggressive an athlete can be with pacing strategies.
• In a standalone marathon, it’s often feasible for well-trained athletes to pace conservatively early on and close strongly, perhaps running the final 10K faster than the first.
• But that strategy doesn’t translate directly to an Ironman, where there’s already a substantial endurance toll before the run begins. In that context, the same pacing flexibility isn’t available.
• From a training perspective, managing rest and recovery becomes especially complex when trying to fit in not just run training, but also cycling and swimming. It’s interesting to note that most elite long-distance runners don’t engage in cross-training.
• Even though it might seem logical to add low-impact aerobic volume through cycling or swimming—especially if injury risk limits run volume—this isn’t common practice. If cross-training truly gave a substantial advantage, we would expect someone to do it and outperform the rest. That hasn't happened.
• My suspicion is that running inflicts not just mechanical but also deep physiological stress, meaning that adding more aerobic work through cross-training may compromise the recovery needed to actually benefit from running sessions.
• When applying this concept to Ironman training, it becomes clear that structuring training in a way that allows for real recovery is crucial. The goal is to create space within the training week to regenerate, rather than simply accumulating fatigue.
• Fueling remains central to performance. We’re seeing athletes consume increasingly high amounts of carbohydrates—150 grams or more per hour on the bike, and somewhat less on the run.
• In my experience, fueling aggressively on the bike is much easier than doing so while running, particularly at high intensity. Runners often struggle to take in more than 70–75 grams of carbs per hour due to gastrointestinal limitations and the mechanical demands of the activity, such as vertical oscillation and reduced blood flow to the gut.
• Therefore, it seems that one of the smartest strategies is to maximize fueling on the bike. Aggressively front-loading carbohydrates while cycling may set the foundation for a stronger run.
• Recovery strategies in training, including adequate fueling and perhaps the use of modern "super shoes" that reduce mechanical strain, may also contribute to faster recovery. That said, it's notable that Ironman run times—even with improved gear and strategies—aren’t significantly faster today on average than they were in the past.
• This may be because athletes are now riding so much harder, shifting the balance of where performance gains are made.

Listener questions

1:25:12 -

When do you introduce intensity after a running injury?

Answer:

That really depends on the severity of the injury. If it’s a significant one, I prefer to start with at least a couple of weeks of purely easy running before reintroducing any form of intensity. That said, I do tend to bring intensity back a little earlier than some—but only in very small, controlled doses. For example, I might include some progression runs or short "cut-down" repeats—something like five or eight sets of two minutes. These usually start at easy pace and gradually build to something like 10K or 5K effort, assuming the runner is tolerating it well.

The key is flexibility. If someone starts to feel pain—whether it’s in the Achilles, the knee, or the IT band—they can ease off or stop altogether. The goal is to reintroduce intensity in a way that’s structured and non-aggressive. We’re not going straight to the track to rip 400s.

Avoiding intensity for too long can actually backfire. When intensity is reintroduced after a long absence, it may feel so foreign to the body that it increases injury risk. So instead, I prefer a gradual, cautious reintroduction that respects the athlete’s current condition.

Listener Question (Toby):

What race or test distance should you use to establish training zones?

Answer:

I prefer the 5K distance for a couple of reasons. Anything from 3K to 10K can work, but 5K strikes the best balance. It’s familiar to nearly every runner—high schoolers, college athletes, adults, even trail runners—and people generally have a good sense of their 5K pace.

More importantly, from a physiological standpoint, it’s long enough to be predominantly aerobic but not so long that fatigue and durability become the primary limiting factors. Unlike shorter distances like the 800m, where a substantial portion of the performance depends on anaerobic energy (which varies widely between individuals), the 5K is mostly aerobic and offers a more reliable indicator of endurance fitness.

In contrast, using marathon pace as a benchmark can be misleading because it's influenced not only by your aerobic capacity when fresh but also by your ability to maintain that fitness over time—a quality we call durability. Some athletes may run excellent marathons with modest 5K times, while others with great short-distance speed may underperform in the marathon due to a lack of durability. That’s why I find the 5K an ideal middle ground for setting training zones.

LINKS AND RESOURCES:

• David Roche – The training and racing strategy behind his epic Leadville 100 course record | EP#444 - inspiring episode about ultra-running, with lots of interesting thought experiments for any endurance athlete
• Michele Zanini (part 2) | EP#394 - a deep-dive into Renato Canova's training methods, from somebody who spent a lot of time working together with Canova
• Run training talk with Jon Green | EP#315 - a great interview on run training with the coach of Molly Seidel, bronze medalist at the Tokyo Olympics