Don’t believe every table you see: Noakes on running potential

Posted by Dave on July 11, 2011 | 8 Comments

Lore of Running

930 pages of science!

This summer I’m reading Tim Noakes’ massive volume Lore of Running, bit by bit, then reporting the critical bits back to you. Last week I covered VO2 max, the idea that every runner can consume a relatively-fixed maximum amount of oxygen. But while VO2 max can’t be improved much with training, I held out hope that running efficiency (referred to in the scientific literature as “economy”) can be.

Unfortunately, Noakes spends the remainder of the chapter making a fairly strong case that even running efficiency can’t be improved much. Noakes argues that a distance runner’s performance in events longer than 5K is almost perfectly predicted by her or his performance in the 5K. Thus, the preponderance of tables and calculators all over the internet predicting your marathon time based on your best 5K time, or your time for any other distance. For example, I can visit the McMillan Running Calculator and input my 5K PR of 21:32, and it will spit out my projected times for every distance from 100 meters (18.3 seconds) to a marathon (3 hours and 29 minutes). Of course, I prefer to enter my 8K PR, which gives me a projected marathon time of 3 hours and 23 minutes!

McMillan takes pains to point out that you can only achieve your projected times if you train for the particular race; if you have only trained for a 5K and your PR is 21:30, you’re not going to go out tomorrow and run a three and a half hour marathon — you have to do your best job training for the marathon first.

A lot of people take a look at something like the McMillan calculator and start thinking about hypotheticals: “If I can get my 5K time under 20 minutes, that means I could run a 3:15 marathon!” Noakes suggests that’s the wrong way of thinking about it. Your 5K pace isn’t going to improve a whole lot, he argues, because at that pace you’re running at about 95 percent of VO2 max, and, if you recall, it’s very difficult to improve your VO2 max. All training will do, Noakes seems to be saying, is allow you to attain your maximum potential given your (relatively) fixed VO2 max. Just as some runners have genetically higher VO2 max, some runners are genetically more efficient than others, and just as you can’t do much to improve VO2 max, Noakes claims, you can’t do much to improve efficiency, either. Efficiency, Noakes says, has a bit to do with your stride length and how much you bounce up and down while you run, but other than that, it’s primarily due to genetic factors like your muscles’ ability to store energy and resist fatigue. Is that really true?

I find all of this rather confusing, and I expect you might too, if you’ve ever progressed from a relatively unfit phase of your life to a more-fit phase. A year and a half ago, I ran an 8K in about 43 and a half minutes. This year, I ran one in 34 minutes, improving by nearly 22 percent. If VO2 max can only improve by 5 to 15 percent, how is that possible?

The reason is simple: These tables are based on individuals who are trained to their maximum potential. If you’re completely out of shape and you try to run a 5K, you probably won’t be able to finish without walking most of the distance. You might not be able to finish at all. Suppose you manage to complete the race in 50 minutes. McMillan’s calculator dutifully predicts you’ll finish a marathon in 8:07:26, an average pace of 18:37 per mile, which is no more than a brisk walk. But I guarantee you, if you spend the next year training and avoid injury, you’ll be able to complete a marathon (and a 5K, for that matter) much faster than that. In other words, if you’re not already in the best shape you can possibly be in, these fitness calculators tell you very little about your potential.

Elite marathoners typically are extremely lean, about 4 to 7 percent body fat for men, and 6 to 15 percent for women—about half the level of the average person. Many people with anorexia have more body fat than that! (However, as Dan points out in the comments, although some runners do suffer from anorexia, a true elite marathoner manages his or her diet properly and can safely maintain a lean body mass) and bordering on some definitions of anorexia. If you have, say, 35 percent body fat, you can get a whole lot fitter just by losing some of your excess body fat; there’s no way a pace-predictor can account for that. This isn’t to say losing that weight will be easy—but it’s not impossible! What is more difficult is determining when you’re as fit as you can be. You can always lose more weight, but at some point losing weight becomes unhealthy.

So despite the somewhat fatalistic statements in this section of Lore of Running (I’ve only read through page 91 out 930), most runners will be able to improve considerably through training. While it is true that few of us have the potential to be elite runners, nearly everyone can be much better than they are now.

What this means for you: Running pace predictors can be good fun, and they can be useful for making short-term goals. But unless you are already in top shape, they don’t define absolute limits for performance. Nearly everyone can improve through training.

Noakes, T. (2003). Lore of Running (4th ed.). Champaign, IL: Human Kinetics.


8 Responses to “Don’t believe every table you see: Noakes on running potential”

  1. Matt Williams
    July 11th, 2011 @ 8:40 pm

    And what does a person have to do to become “are trained to their maximum potential?” I suppose unless you are able to live, eat and train like an elite, you may never hit your true “maximum.”

  2. Dave
    July 11th, 2011 @ 8:51 pm

    Matt: I’m hoping we will find out somewhere in the remaining 840 pages!

  3. Dan Way
    July 12th, 2011 @ 5:39 pm

    I really enjoyed reading this article. Well-written, informative and with practical application. But a few things to note:
    1) Suggesting that elite (or even non-elite) runners are anorexic based simply on their percentage body fat is extremely misleading and problematic. Anorexia is a clinical disorder with 3 (or 4) criteria and varying degrees of severity. Only one criteria specifically pertains to body weight (that being maintained at less than 85% of ‘normal’), and says nothing about body fat levels. Thus it is extremely reductionary to assume the condition based on weight alone without considering the complex psychological features (fear of fatness and undue influence of body weight on self-evaluation). Anorexia is likely more prevalent in runner’s than other athletes, but judging this based solely on weight is incorrect.
    2) The whole point of the prediction calculators are to take one’s current fitness level (including those who are not ‘maximally’ trained) and provide a moderately reliable estimation of what one is capable of. I am only as fit as my latest race suggests and this is a good indicator of how I should train to either maintain current fitness, or preferantially to improve it. If you give an honest effort at a race (which would be expected), your time is the ultimate indicator of your current fitness and future potential. Whether these calculators are relaible in taking that value and predicting outcomes at other distances is certainly suspect and will not be accurate for all distances nor for all individuals but I would argue that they are usually pretty effective (if for nothing else than to motivate me to achieve those ambitious targets)
    Great post and look forward to the next one!

  4. Dave
    July 12th, 2011 @ 6:18 pm

    Dan, Good points. I certainly didn’t mean to suggest that elite runners are anorexic. Sorry for the confusion there. I will modify the post to make it clearer. There is a good set of criteria for anorexia at the end of the Running Times article I linked.

    I have to dispute your second point. You can’t use the calculators to predict your pace in longer races based on a time in a shorter race if you’re not at a high level of fitness. The assumption of the calculators is that if you train for the specified race, then you can perform at the level specified. But if you’re not currently very fit, then once you train for that race, you improve your level of fitness, and the pace predictor will be wrong.

    There are some uses for the calculators in the short term, though. You could use your current 5K pace to predict your equivalent marathon pace, then use that pace as a guide in your training. But as you become more fit, you should probably recalibrate your training goals based on running another 5K.

    What Noakes is arguing in the book is really something different. He’s making the case that these tables are fairly inflexible once you approach a maximum fitness level. Suppose an athlete trains for several years and manages to run an 18-minute 5K. She has reduced her weight to an optimal level, done everything right, and this is the best she can do. Noakes says that she has no hope of ever being a world-class marathoner, because a pace-prediction table shows that the best she could do in a marathon, even after marathon-specific training, is 2:55. Maybe she could shave a few minutes off of that with some enhancements to running economy, etc., and she could even win some second-tier marathons, but she’s not ever going to be in the 2:20 range that women like Davila, Kilel, and Flanagan are running.

  5. Dan
    July 13th, 2011 @ 1:32 pm

    I guess the point I was trying to make was that based on a recent race (which indicates current fitness), the pace calculators are fairly reliable for giving predictions for other distances IF one is willing to train for them and put in the work. They are particularly useful for determining HOW one should train. If you don’t like what you see, increase your fitness and the numbers/times will go down! Certainly when one achieves his/her highest level of fitness (impossible to know?!), they will only ever be able to accomplish certain times at any distance. I would like to add that for any runner hoping to improve, these prediction tables are excellent, reasonably reliable (and free!) resources for training in general and for specific races.

  6. Dave
    July 13th, 2011 @ 2:23 pm

    Dan, it sounds like we’re agreeing; I’m just emphasizing different aspects of the calculators and trying to put them in the context of Noakes’ claims about the maximum potential of a runner.

  7. Fiona
    August 27th, 2011 @ 6:15 am

    Reading this with interest. One question/thought: Elite 5 km runners are not always very good marathon runners. There must be a point at which the reliability of these tables starts to be inaccurate. An amazing track sprinter who can do super-fast 5 km race times could train for a marathon, but, if they have a muscular sprinter’s build (lots of high twitch fibres, bigger muscles) and not do all that well at the marathon. Does this author consider this? Something to think about…

  8. Stephen
    August 27th, 2011 @ 7:43 pm

    McMillan specifically says on his site that anything under 800 meters is more like a guess. I entered my times as an example. I recently ran a 19:46 5K. The table says I should run a 5:42 mile and 1:10 400 meter. The mile time is basically dead on what I can run, but I can run a 58 second 400 meter which is almost 20 percent better than what he expects.
    So I definitely agree with McMillan that the shorter distances are different enough mentally and physically that it would be hard to compare them to longer ones. I’m really curious if I would also underachieve what would be expected on longer races based on my 5K time, or if that’s sufficiently long enough to be useful. Perhaps the tables could be improved by making you enter times from multiple distances.