This past week at UH I had the pleasure of sitting in on a talk by Dr. Ed Coyle on ‘how fast could we run a marathon.’ Coyle is a well-established researcher, but he’s most well known for his infamous research on Lance Armstrong that showed improved efficiency over his career. Coyle still maintains the validity for this research for some reason, which to many dampens his work, but that’s for another topic.

The reason I’m writing this is because the talk and the discussion after demonstrates a fundamental flaw in the world of sports performance and exercise science.

I sit in this weird in-between zone of scientist and coach. I consider myself more on the coach side, since that’s my primary job, first love, and how I got into this sport, but still I live full time in both worlds. There are a lot of really good scientist and coaches who dip their fingers in both sides and they’re to be commended, but I think my situation is slightly unique in that I’m truly fully engaged in both sides at the same time. I have to be. So it’s these two interacting worlds that I battle. I’d like to use this post as a learning tool. An almost “A scientists guide to how to interact with coaches.” Then in the next month or so, put up the counter point and produce a guide for how coaches should interact with scientist.

Before we get there, let’s use Coyle’s talk as a framework to understand the disconnect.

The Talk:

The talk centered on what physiology predicted we could ultimately run in the marathon.

Going into the talk, I mentioned the things I hoped it wouldn’t be about (namely 2 hr marathon is going to happen soon, psychological barriers, and Paula Radcliffe…why these things? Because it’s seemingly the necessary checklists for discussing the marathon now in the scientific world) to my friend and fellow doctoral student, Justin, and low and behold those were the topics it was about. It’s not to disparage those topics and to say that they aren’t important, but it marginalizes a really tricky argument on how fast we can run a marathon.

Briefly, the talk centered on the standard metrics of VO2, %VO2max, Running Economy, and Lactate Threshold and how these components could predict performance to a degree. Coyle broke down a few studies that showed their importance in elite running and then how they could be used to predict performance. While my gripes with some of these components are well known, from a traditional scientist viewpoint I can see why they were used and don’t have a big problem with it.

The problem is the jumps that are made. For instance, when discussing economy, Coyle pulled out his work on Lance Armstrong on how efficiency can improve over time. There are obvious problems with this, which are so blatant I don’t even need to go into them. Then to no surprise, he pulled out the famed Andy Jones work on Paula Radcliffe to demonstrate how VO2max doesn’t change with performance and that LT and economy do. And it’s these changes that led to her world record performance. The problem we have here is that we’re basing our entire view of world class marathoning on a single person, an extreme outlier, which elites certainly are, but I’m not sure if that is the best approach. For instance, Coyle jumped to conclusions that because Jones reported that Radcliffe could run at 88% of her VO2 for a marathon while men could only run at 85% of their VO2, that the fastest women’s marathon time possibly might be 2:04, while the men’s might be 1:57 high.

Which, if you study anything about the typical gap in performance between men and women (or even look at today’s WR), seems a bit off…

The disconnect:

I’m not blaming Coyle on this, his job is a scientist, but I think the questions asked after the talk and the general feeling you got provides a unique view into why scientist and coaches often don’t click together.

Seeing the talk from the coaches side for instance, was slightly maddening. Post talk, part of the discussion was around Sammy Wanjiru and how he was thought to be someone who could challenge the 2-hour barrier because of his aggressiveness and his “fast half marathon times.” Now, if any coach heard this and realized that Wanjiru’s HM times were fast but not unlike any other world class marathoner and fit really well in line with his actual marathon performances, then it could drive you made. It demonstrates the disconnect that occurs. For instance, if Wanjiru’s HM time of 58:35 signified sub 2 hour potential, then what did Tadese’s 58:23 WITH a research based running economy of 150 that was simply insane. It certainly didn’t predict his lack of marathon success…

As coaches, we pride ourselves on understanding rough equivalencies in performance. It’s innate. I know that if someone runs A for 800m and B for 3,000m, he’s going to be roughly in C range for 1,500m. Similarly, we can be in a rough range of knowing what kind of HM time it takes to run close to certain marathon times. There is a gap in speed necessary.

With researchers not familiar with the sport, this innate knowledge isn’t there. Which leads to a disconnect in times. You have people think that just because someone ran a half marathon a decent chunk under 2 hrs, why can’t they just run slightly slower for double the distance?  It’s akin to taking a 1:57 half miler and saying, why can’t you break 4 for the mile with some more training?!

Which is why you get people using linear performance curves to predict that sub 2 hours will be run in 2019. Which leads me to another common audience question, that went something along the lines of “it seems like they’re taking around half a minute off the marathon times lately each year, so it should be within a few years then that sub 2 hours happens.”

It’s the disease of linear thinking. It’s an in built cognitive bias that leads us to think that if things are progressing in a certain way right now, they will always follow that trend.

But history shows that’s not true. Without even using elaborate statistical methods, we can see that the marathon times are essentially just catching up to the track 5k/10k times. Which have stagnated for the past decade.

And this makes complete sense. Why? Well, the money is on the roads, there aren’t many 10k’s anymore, and so forth. SO instead of having a guy like Geb move to the marathon in the twilight of his career to set a WR, you have guys who most likely would have been Oly medalist type 10k runners going to the marathon in the prime of their career to attack the marathon. It’s no longer do everything you can on the track then move up. It’s now move up, see what happens and make some cash.

So when that “sports knowledge” isn’t taken into consideration, you get crazy predictions like sub 2hrs before 2020 based on a simple linear model.

This leads us to the central problem. Claims like these are why most coaches who don’t put much stock into scientist.

It’s because, it is so outlandish that no one takes it seriously. It’s like me walking up to Asbel Kiprop and telling him he has a 3:37 mile in him if just trains correctly over the next few years. It’s so far out there, that it puts you into the “out there/crazy/disconnected from reality/believes in alien abduction’ area.  In other words, as coaches, we’d just shuttle you off into the “smart but doesn’t know anything about the realities of sport, so let’s just ignore them” category. Or simpler terms, we won’t listen to you.

You’ve negated your expertise by making big ridiculous claims. And what happens, inevitably, when the question arises of how we’re going to get the sub 2 hr marathon in 5 years that these graphs predict, the white knight of science comes swooping in.

The post talk questions surrounded how science could lead to breakthroughs “as we certainly aren’t evolving at such a rapid rate to see improvements”.  As a researcher, I have no problems asking this question and talking about how Asker Jeukendrup helped Geb with his fueling leading up to the marathon. But what I have a problem with is things like this:

The Sub 2 hour project

What it does is marginalizes coaches. What it implies is that we can do a better job of training athletes than you can. Just look at the training sections and claims. It’s almost stating “If we use a scientific training method, we’ll get there.” My argument is that if you use training methods based purely on research, you will get worse, because you’re ignoring a century of “evolution” of training methods towards a more optimum state. While you’re relying on training methods validated over 6 week studies on college recreational athletes for the most part that tell you to do HIIT all the time…but I digress.

And that’s the problem, as it was portrayed in the questions after this talk, and in many talks I’ve seen. They’ll state something akin to “If the East Africans only had modern science and coaching….”  “They don’t use any technology (HR, lactate, VO2, etc.)”…

The problem is from a scientific perspective; this is seen as a bad thing. But if you asked me, it’s a wonderful thing. They learn naturally how to run by feel, listen to their bodies, and therefore have an in built perception system that is second to none.  That’s the “secret” to East African running; the psychological approach to running and not stressing about it, not having barriers, and not obsessing, among other things.

But from a researcher standpoint it’s seen as a negative. To get to the point, you’re putting down the coaches that work with these athletes, the athletes themselves, and the teams surrounding them because you’re making the egotistical assumption that ‘If only they had this science, they could run sub 2 hrs’…

With that in mind, let’s look at how scientists should interact with coaches.

My Scientific Guide for Interacting with Coaches:

So I’m going to end this with a point and that is to use the above as a framework of why coaches typically don’t enjoy working with scientist and use that as my “how to interact with coaches” guide for scientist. (for those scientists, don’t worry, in a future blog post, I’ll attack the other side from a scientist point of view)


  1. Don’t make outlandish claims

It makes you look silly and undermines credibility if you say things like “I can help you run sub 2 hours” or “the 2 hour marathon will happen in 5 years”.

Why? Because it shoots up red flags that you have no clue about how the sport actually works.

  1. Don’t tell us everything we do is wrong.

We coach. It’s our job to make people faster. You might think our insistence of running long runs is ludicrous because there’s no “research” on it. Don’t tellus that because of that we need go do HIIT training all day.  It would be like a coach walking into your office and telling you that you need to do your research protocol in a certain way. You’d think he was crazy and be offended by this fool not in your specialty telling you how to do your job. That’s what happens when you tell coaches how everything they do is wrong.

  1. Stop discounting history and respect it.

There’s a reason we’ve ended up at the training design that we have. Training has evolved over decades to get to the current point it has. We’re much better now than we were 40 years ago and partly that evolution is to thank for that. It doesn’t mean we have it all figured out, but don’t  discount it’s importance. Similarly, don’t come screaming to me about how HIIT is the greatest thing since sliced bread when track coaches have been doing a variation of it and refining it for a century. It has it’s place, just like everything. If you know history, you know not to fall for this trap.

  1. Don’t sell Science as the savior of the world

In this talk, a common theme was what role does sports science have in running performance. The Andy Jones work was pointed to and I even mentioned Asker Jeukendrup’s work with Geb as more evidence. But there was this undertone of “if only the East Africans, who relied on primitive techniques and coaches could be saved by the white knight of science, then we could break through.” If that was the case, then all of our American and European athletes who had Sports Science help would be in the mix. They’re not. The East Africans run incredibly fast without much science. That tells you something in terms of training….Now when you get special things is when you help refine items like fueling (as in Geb’s case for the marathon) with science. Please take note.


1.Help me refine my processes

If you engage with a coach, help him refine his current practices. Don’t tell him the workout he needs to do, help him refine how to do that workout better. Or to recover off of it better. Refine, tweak, don’t try and tear us down.

  1. Help on things outside of my realm of expertise that can be practically applied

Great examples of this are on the nutritional fueling side for marathons or on what supplements work. Or perhaps on the timing of strength versus endurance training. You see how these are all things that have been researched and matter? They also don’t invoke wholesale changes in a training philosophy.

  1. Understand the intricacies of our sport OR don’t claim to be an expert

When working with sports, you can do two things. Either understand the intricacies and history of the sport so that you have a slight grasp of the coaches intuition OR claim ignorance and simply offer an outside perceptive, that could be ludicrous at times, but should be presented in that way to the coach. Either way, if you’re a research, don’t claim expertise in a sport if you don’t have it.

  1.   Implement practical changes

Remember, the goal is performance. Don’t simply test everyone’s VO2max to give some esoteric intensity zone because that’s what you learned in exercise testing class a decade ago. Focus on items that will result in practical, tangible benefits, that don’t take obsessive amounts of time or energy. You’re goal is to seamlessly integrate with the program. If you can track something without any hassle, go for it. But don’t make athletes fill out hours of things that might not make any difference at all. Time and energy matter.

  1. Know your role

The point is, know what role you play in the system. What are you supposed to do, help with, and how are you supposed to make athletes better. Do that. It’s simple.

So what?

DO it the right way. One of my PI’s Dr. Simpson likes to point out that research generally checks out and validates training types with innovation coming from the sport. One of my go to guys for any questions is Dr. Trent Stellingwerf, and he put it perfectly in a recent episode of the “we do science” podcast where he implored researchers to go pick up some training books to understand periodization.

Trent’s point was that you have to understand the framework in which you are working. From a nutrition standpoint, if you don’t understand how periodization works and shifts training demands, how can you understand how the nutritional requirements shift.

So if you make claims on the 2 hour marathon, understand what the heck people have done to run 2:03 in the marathon. Once you understand the training demands there and WHY that athlete has done that, then maybe you can understand how to improve upon that. Understand how fast one might have to be at HM to run sub 2hrs or 10k or 5k.

The point of this whole post I think is one of domain dependence. The idea of domain dependence is that we see things through the lens of our own world-view, or framework.

And what I’m saying is step back, look across domains, see the 2 hour marathon through the lens of the coach and see what happens. Putting an ego aside is a wonderful thing.

For other talks on this, I highly recommend Ross Tucker’s on the sub 2 hr marathon

Get My New Guide on: The Science of Creating Workouts


    1. Robert Osfield on April 1, 2015 at 10:12 am

      I think part of the problem is that the Sport Scientists with the most outlandish headlines/predictions get the media coverage, but these same guys often are the poorest of the scientists. Most science isn't glamorous, it's hard, often mundane little steps of improving knowledge. You spend more time observing the world around you than talking about.

      All walks of life has it's Charlatan's, I've read enough dodgy claims from "Sport Scientists" that there are plenty of Charlatan who simply don't deserve the label "Scientists". I don't think it's so much a poor connect between coaching and science, a big chunk is simply dubious "Scientists" trying to big up their work/ideas far beyond they deserve.

      One needs to hold their feet to the fire. Point out the dubious reasoning. This post certainly is good step in the this direction, Ross Tucker's post on the 2hr marathon nonsense is a great one too. The media needs to listening more to the scientist that are good at their job, less time with the charlatan's, this goes for all walks of life.

    2. CoachRivas on April 21, 2015 at 7:31 pm

      Coyle’s work with Armstrong was disputed on how he analyzed his data. You sort of mentioned it so not sure how much weight his talks carry and should not be taken as a whole on how all scientist view coaching.

      I agree there is a disconnect between coaches and scientist, but I don’t agree with how you viewed it. The “science” if that’s what you want to call it is so outdated because there is no money in understanding how to get someone to run faster or jump higher. Therefore physiologist can’t make a career out of it. Additionally, when the current tools of analyzing exercise performance have not changes since the 1920s other than the technology of analyzing vo2/co2 in real time and not in a Douglas bad, tells how far the ”science” of exercise performance is. Also, respectfully, coaching is no different than the “bro science” in the weight room – anecdotal evidence from coaches on how train an athlete is not really science, is it? It holds no true value in the science world. The cutting edge understanding of exercise and adaptation would be to look at changes in molecular signaling or epigenetics – which like you have stated are done in healthy or disease populations, not in elite athletes, though I have read a few in collegiate athletes that have done this.

      You should look up work from Michael Joyner, MD from mayo clinic. Dr. Joyner was a marathoner himself, and is interested in the physiology of elite athletes, but surely has not made a career out of it.

    Leave a Reply