How Should We Think About Training?

The science, the tradition, and the individual

19 min read

Most training advice skips a step.

It jumps straight to what you should do — which intervals, how many easy kilometres, what intensity split — without addressing a more fundamental question: how do we know any of this is right?

This matters more than it might seem. If you have ever read one article arguing for polarised training and another arguing for threshold-heavy blocks, and found both convincing, you have already encountered the problem. The disagreement is not just about training. It is about what counts as evidence, how much weight that evidence can bear, and what happens when the evidence runs out — which it does, sooner than most people realise.

This article is not a training guide. It is an examination of the foundations that any training guide rests on. How should a serious runner think about the knowledge that informs their training? Where does that knowledge come from? Where is it strong, where does it thin out, and what fills the gap?

The answer involves three things: sports science, coaching tradition, and the individual. Understanding how they relate — and where each is trustworthy — is the starting point for training that is genuinely well-reasoned rather than just confidently asserted.

What sports science gives us

Start with what works. Sports science and exercise physiology have given us a remarkably detailed understanding of how the body responds to training.

We know the physiological determinants of distance running performance: VO₂max (the ceiling of aerobic energy production), the lactate and ventilatory thresholds (the intensity you can sustain before fatigue accumulates rapidly), running economy (the oxygen cost of running at a given pace), and increasingly, durability (the ability to maintain those qualities deep into a race as fatigue mounts). We understand how these relate to each other and why improving one often benefits the others.

We understand adaptation mechanisms — the specific biological processes by which training stimuli produce performance changes. Mitochondrial biogenesis (more and better cellular power plants), capillarisation (denser blood supply to working muscles), cardiac adaptations (a heart that pumps more blood per beat), connective tissue remodelling (tendons and bones that tolerate greater loads), neural efficiency (muscles that activate more effectively). Each of these responds to specific types of training, and we can describe those relationships with reasonable precision.

We understand fuelling: how the body selects between carbohydrate and fat at different intensities, why glycogen depletion limits performance in longer events, and how nutrition supports recovery and adaptation. We understand injury at a mechanistic level: the relationship between training load and tissue capacity, the biomechanical and systemic factors that predispose runners to breakdown.

This knowledge is real. It is hard-won through decades of careful research. It provides the physiological logic that underpins everything a coach does — or should do. When a coach prescribes a session, they should be able to explain which physiological adaptation it targets and why that adaptation matters for the athlete's goals. Sports science makes that explanation possible.

None of what follows is a diminishment of this achievement. But honest engagement with the evidence base requires asking a harder question: what can it not do?

Where the evidence thins out

The limitations of sports science for individual training prescription are not failures of the researchers. They are structural features of how research works — built into the method itself. Understanding them is not scepticism for its own sake. It is the most scientifically rigorous position available.

Isolation versus integration

A controlled study typically isolates a single variable. One group does threshold intervals twice a week. A control group does steady-state running. After eight weeks, the researchers measure who improved more on a specific metric.

This is good experimental design. It tells you what threshold intervals do when that is the primary intervention. What it does not — and structurally cannot — tell you is how to combine threshold work with easy volume, a long run, a tempo session, strength training, and recovery within a single week for a specific person who also has a stressful job and two young children.

That integration problem is where coaching actually lives. Not in whether threshold training works — it does — but in how much, how often, alongside what else, and for whom. The study answers the first question. It is largely silent on the rest.

Population means versus individual response

Research reports group averages. A study might find that a particular training protocol improved VO₂max by 3% on average over eight weeks. That average is real, but it is an average. Within the group, some individuals likely improved by 6–8%. Others may have improved by 1% or not at all.

This is not a flaw in the study. Individual variation in training response is a well-documented phenomenon. But it means that a finding like "threshold training improved performance by X%" tells you something about the population and less about any specific person within it. You might be a strong responder to threshold work. You might not. The study cannot tell you which.

For a coach working with an individual, population-level findings are starting points — informed, evidence-based starting points, but starting points nonetheless. The individual's actual response is the data that matters most.

The population problem

A large proportion of exercise physiology research is conducted on recreationally active or untrained participants. This is partly practical — recruiting experienced athletes for controlled studies is harder — and partly statistical, since untrained individuals show larger and more consistent responses to training interventions, making effects easier to detect.

The consequence is that much of the evidence base describes what happens when relatively unfit people start training. It describes this well. What it describes less reliably is what happens when a runner with five years of consistent training and a well-developed aerobic system applies the same intervention. The magnitude of response, the time course, and sometimes the direction of effect can differ meaningfully between trained and untrained populations.

This does not make the research useless. It means the findings require careful translation. A protocol that dramatically improved VO₂max in previously sedentary participants may produce a much smaller — or even negligible — improvement in an already well-trained runner.

Short studies, long development

Most training studies run for six to twelve weeks. This is a reasonable timeframe for detecting acute physiological changes, and it is often what funding and logistics permit.

Distance running development happens over months and years. A fully developed aerobic base — the product of years, potentially a decade or more, of consistent training — is what separates good runners from the best. Connective tissues — tendons, ligaments, bones — adapt on timescales of months to years, not weeks. The compounding effect of consistent training over long periods is arguably the single most important factor in distance running improvement, and it is almost invisible in short-duration research.

This creates a subtle but important distortion. The training approach that produces the best measurable outcome over eight weeks is not necessarily the best approach over eight months or eight years. A protocol that maximises short-term improvement through aggressive intensity might come at the cost of injury risk, burnout, or diminished long-term development — costs that a twelve-week study simply does not run long enough to observe.

The prescription gap

Even where a finding is robust, replicated, and drawn from relevant populations, there remains a gap between the finding and a specific prescription for a specific person.

"Threshold training improves lactate threshold" is a well-supported statement. But it does not tell a coach how many threshold sessions per week this particular athlete should do, at what duration, alongside what other training, during which phase of their preparation, given their injury history, their current fatigue state, and the race they are preparing for. The move from a general physiological principle to a specific training session on a specific day for a specific person requires a series of judgments that the research does not — and is not designed to — make.

This is the prescription gap, and it is where science hands the problem to coaching.

What it means to take science seriously

There is an irony in how sports science is used in popular running discourse. The most common way to invoke science — "the research shows you should do X" — is often the least scientific way to apply it.

A genuinely scientific mindset includes understanding the limits of your methods. It means distinguishing between what a study actually demonstrated (a specific effect, in a specific population, over a specific timeframe, under specific conditions) and the much broader claim that often gets attached to it ("science says polarised training is best").

No individual study shows that polarised training is best for all runners in all contexts. What several studies show is that polarised intensity distributions produced favourable outcomes in certain populations over certain timeframes on certain metrics. That is a meaningfully different statement. It is useful. It is informative. But it is not a prescription for your training.

The runner who reads a study abstract and adjusts their training accordingly is not being more scientific than the coach who considers the study alongside twenty other inputs. They are being less scientific — because they are treating a narrow finding as a broad prescription, which is precisely the overgeneralisation that scientific training teaches you to avoid.

Taking science seriously means using it as it is designed to be used: as an evidence base that informs reasoning, not as an authority that replaces it.

Coaching tradition as knowledge

If sports science defines what we can demonstrate under controlled conditions, coaching tradition represents what has been learned through decades of systematic practice with real athletes over real timeframes. It is a different kind of knowledge. It is not less real.

Consider Arthur Lydiard, the New Zealand coach whose athletes dominated middle and distance events in the 1960s. Lydiard's central insight — that a deep foundation of high-volume aerobic running produced dramatic improvements across all racing distances, including the 800 metres — was radical at the time. The prevailing wisdom favoured high-intensity, interval-heavy training. Lydiard's athletes did enormous volumes of steady aerobic running for months before introducing any speed work.

He was right. Decades later, sports science provided the mechanistic explanation: mitochondrial biogenesis, capillarisation, improved fat oxidation, enhanced cardiac output. The physiology confirmed what Lydiard had observed empirically. But the knowledge came first. He watched athletes, experimented with their training, observed the results, refined the approach, and produced world-beating performances — all before any laboratory could explain why it worked.

This is not folklore or guesswork. It is empirical knowledge generated through systematic observation and iterative refinement, applied to real athletes in real competitive contexts with real consequences. The method is different from a controlled trial, but the process — observe, hypothesise, test, refine — is recognisably scientific in its logic, even if not in its formal methodology.

The Norwegian double-threshold model offers a more modern example. Norwegian endurance coaches, working primarily with cross-country skiers and distance runners, developed an approach built around multiple days per week featuring two threshold sessions in the same day — a morning session targeting around 2.5 mmol/L of blood lactate, and an afternoon session at roughly 3.5 mmol/L. Both intensities sit below the second lactate threshold, deliberately so. The logic is a trade-off: by keeping each session below the intensity that would impose heavy fatigue, the athlete can accumulate a large volume of quality running — far more than if they pushed to or above threshold and needed extensive recovery. More quality volume, at the cost of lower peak intensity per session. This emerged from elite training environments where coaches combined regular lactate monitoring with practical observation of athlete response.

The approach is now the subject of growing research interest. But its practical application — how to structure the sessions, how to integrate them with other training, how to progress them across a season, how to adapt them for different athlete profiles — still extends well beyond what any controlled study has validated. The coaches are ahead of the science. They are reasoning from physiological principles and accumulated practical experience, using research to inform but not to constrain their programming.

This is what coaching tradition does that controlled research structurally cannot: it addresses the integration problem. How do you combine multiple training stimuli within a week? How do you periodise across months? How do you progress an athlete through years of development? How do you balance the competing demands of volume, intensity, recovery, and life? These are questions about the integration of many variables simultaneously, applied to specific individuals over long timeframes. They are precisely the questions that single-variable, short-duration studies are least equipped to answer.

Coaching tradition answers them — imperfectly, sometimes incorrectly, but with a depth of practical knowledge that no other source provides.

The individual as the final calibration

Sports science defines the space of plausible approaches. Coaching tradition narrows that space through accumulated practical wisdom. But the final calibration is always the individual.

Two runners with the same 10K goal time may need fundamentally different training. One might lack aerobic endurance — their fitness fades in the final third of the race. Another might have a strong aerobic base but a weak top-end speed — they cannot change gears when the pace lifts. A third might have excellent physiological capacity but poor running economy — they are spending too much energy per stride. The limiter is different. The training should be different. The science and the coaching tradition provide the menu of options. The individual determines the order.

This is where training prescription becomes an act of informed judgment. Not guesswork — informed by physiology, informed by what has worked for similar athletes, informed by the individual's own history and responses. But judgment nonetheless. A coach choosing between two threshold sessions or one threshold and one VO₂max session for a specific athlete in a specific phase of preparation is making a judgment call. The evidence informs it. Experience guides it. But no study definitively answers it for that person.

This is not a failure. It is the reality of applied work with complex biological systems that vary between individuals, respond differently to the same stimulus, and exist within life contexts that no study can replicate.

The appropriate response is not to pretend the uncertainty does not exist. It is to be transparent about it. A well-reasoned training recommendation comes with an honest account of the confidence behind it. Some things are well-established: consistency over months and years is the primary driver of improvement. Aerobic development is the foundation. Progressive overload works. These are high-confidence claims supported by both research and extensive practical experience.

Other things are judgment calls: whether this athlete would benefit more from a third easy run or a second quality session this week, whether to introduce VO₂max work now or build the threshold a little further first, whether the reported fatigue is a sign to back off or a normal part of adaptation. A good coach makes these calls clearly and explains the reasoning — but also monitors the outcome and adjusts when the evidence warrants it.

The transparency is the point. An athlete who understands that "this is my best judgment based on your training history and how you responded to the last block — we will adjust if it is not working" is better served than one who is told "the science says to do this." The first is honest. The second overstates what the science can say.

Acting well under uncertainty

The inability to prescribe objectively optimal training is not a limitation to be embarrassed about. It is the condition of every applied domain that works with complex, variable, real-world systems.

Medicine works this way. A doctor choosing between treatments for a patient is drawing on research, clinical experience, and the individual patient's history and response. The best treatment for the average patient in a clinical trial may not be the best treatment for the person in front of them. The doctor exercises informed judgment, monitors the outcome, and adjusts. This is not a failure of medicine. It is medicine working as it should.

Engineering works this way. A structural engineer designing a bridge uses well-established physics and material science, but also applies safety factors, accounts for local conditions, and makes judgment calls about uncertainties. The calculations provide the foundation. The engineering judgment determines the specific application.

Coaching is no different. The foundation is strong: we understand the physiology of training adaptation, we have decades of practical wisdom about what works, and we can observe the individual's response in real time. What we cannot do is eliminate uncertainty from the process of applying that knowledge to a specific person. And that is fine.

What distinguishes good coaching from bad coaching is not the elimination of uncertainty. It is the quality of reasoning within it. Good coaching uses the best available science without overstating what it proves. It draws on the deepest available practical wisdom without treating tradition as infallible. It calibrates to the specific individual without ignoring the evidence base that applies broadly. It is transparent about confidence levels. And — perhaps most importantly — it adapts based on what actually happens.

A good training plan is not one that was right from the start. It is one that was well-reasoned, honestly communicated, and intelligently revised as new evidence arrived — from the athlete's own training.

This is how we think about training at Runaid. Not as a problem that has been solved, but as a problem that rewards careful, honest, adaptive thinking. The science provides the foundation. The coaching tradition extends it. The individual completes it. And the conversation between all three is ongoing.

This is the first in a series of articles exploring the principles behind Runaid's coaching approach. Next: the physiological model of distance running performance — what determines how fast you can run, and how the pieces fit together.

References and Further Reading

The claims and frameworks in this article draw on a broad body of research and coaching literature. The following sources are recommended for readers who want to go deeper into the topics covered.

Physiological Determinants of Endurance Performance

Joyner MJ, Coyle EF. Endurance exercise performance: the physiology of champions. Journal of Physiology. 2008;586(1):35–44. — The foundational review on how VO₂max, lactate threshold, and running economy interact to determine endurance performance.
Bassett DR, Howley ET. Limiting factors for maximum oxygen uptake and determinants of endurance performance. Medicine & Science in Sports & Exercise. 2000;32(1):70–84. — A comprehensive treatment of the physiological ceilings on aerobic performance.
Jones AM. The physiology of the world record holder for the women's marathon. International Journal of Sports Science & Coaching. 2006;1(2):101–116. — A case study illustrating how the determinants of performance operate in an elite individual.

Training Intensity Distribution

Seiler S. What is best practice for training intensity and duration distribution in endurance athletes? International Journal of Sports Physiology and Performance. 2010;5(3):276–291. — The most cited review on how elite endurance athletes actually distribute their training across intensity zones.
Stöggl T, Sperlich B. Polarized training has greater impact on key endurance variables than threshold, high-intensity, or high-volume training. Frontiers in Physiology. 2014;5:33. — A controlled comparison of different training intensity distributions in trained endurance athletes.
Muñoz I, Seiler S, Bautista J, España J, Larumbe E, Esteve-Lanao J. Does polarized training improve performance in recreational runners? International Journal of Sports Physiology and Performance. 2014;9(2):265–272. — One of the few studies examining intensity distribution in recreational rather than elite athletes.

The Norwegian Model

Casado A, Hanley B, Jiménez-Reyes P, Renfree A. Does lactate-guided threshold interval training within a high-volume low-intensity approach represent the "next step" in the evolution of distance running training? International Journal of Environmental Research and Public Health. 2023;20(5):3782. — A detailed description of the lactate-guided threshold model and its physiological rationale, including the training patterns of the Ingebrigtsen brothers.
Haugen T, Sandbakk Ø, Seiler S, Tønnessen E. The training characteristics of world-class distance runners: an integration of scientific literature and results-proven practice. Sports Medicine – Open. 2022;8:46. — A comprehensive review of how elite distance runners train, including the Norwegian approach in context.
Kelemen B. Norwegian double-threshold method in distance running: systematic literature review. Scientific Journal of Sport and Performance. 2024;3(1):1–18. — A systematic review of the double-threshold approach, including training structures and lactate targets.

Strength Training and Running Economy

Balsalobre-Fernández C, Santos-Concejero J, Grivas GV. Effects of strength training on running economy in highly trained runners: a systematic review with meta-analysis of controlled trials. Journal of Strength and Conditioning Research. 2016;30(8):2361–2368. — Meta-analysis demonstrating a large beneficial effect of strength training on running economy.
Denadai BS, de Aguiar RA, de Lima LCR, Greco CC, Caputo F. Explosive training and heavy weight training are effective for improving running economy in endurance athletes: a systematic review and meta-analysis. Sports Medicine. 2017;47(3):545–554. — Further meta-analytic evidence on the mechanisms by which strength training improves running efficiency.

Individual Variation in Training Response

Bouchard C, Rankinen T. Individual differences in response to regular physical activity. Medicine & Science in Sports & Exercise. 2001;33(6 Suppl):S446–S451. — The landmark paper on inter-individual variation in training response, demonstrating the wide distribution around population means.
Mann TN, Lamberts RP, Lambert MI. High responders and low responders: factors associated with individual variation in response to standardized training. Sports Medicine. 2014;44(8):1113–1124. — A review of the physiological and genetic factors that contribute to why individuals respond differently to the same training.

Coaching Tradition and History

Lydiard A, Gilmour G. Running with Lydiard. Meyer & Meyer Sport, 2017. — The definitive account of Lydiard's training philosophy, updated from his original texts.
Daniels J. Daniels' Running Formula. 3rd edition. Human Kinetics, 2014. — The standard reference for VDOT-based training pace prescription and the systematic relationship between fitness and training intensities.
Canova R, Arcelli E. Marathon Training: A Scientific Approach. March 1999. — Canova's treatment of progressive specificity and the integration of race-specific work within high-volume programmes.

← All articles Next: The Physiology of Running Performance→