Chicago 2009: Thoughts on pacing and the limit to performance
Well, I have arrived in Chicago after a great couple of days in Washington, and now the build-up to the Chicago 2009 Marathon can begin! A little later today, Jonathan and I will be meeting with some of the Marathon organizers to discuss race-day logistics and to figure out how (if at all) we can go about doing live 1km race split reports here on The Science of Sport.
We're also meeting with a local health insurance company for a Q & A and a short presentation, and then tomorrow, I am at the UIC to give a talk to the department. So a busy few days have begun!
However, the marathon is the focus, and particularly, the presence of Sammy Wanjiru, Olympic and London Champion. There was talk of sending pacemakers out at 61:40 to halfway, in a quest to break the world record. The most recent reports suggest that he won't go for the record, but will rather focus on solidifying his lead in the World Marathon Majors. Is that an attempt to deflect attention and downplay the occasion? I have a feeling he'll go for it though, only because he's spoken of his desire before, and if he's in good condition, Chicago represents a real opportunity. Wanjiru is also so aggressive - remember his Beijing performance in the heat - that I can't see him holding back if he has good form.
So far, the weather forecasts look favourable - it won't be too hot, that's for sure. The only concern may be around the wind, which has been pretty gusty since I arrived. Regardless, 1km splits will be interesting, because if it's a slower race, then the surges, which Wanjiru used in both Beijing and London will make for an interesting analysis.
Pacing strategy analysis
And it's pacing strategy that is the focus of today's post. At least, some thoughts on pacing and the limits to performance, which were inspired by the talk I was preparing for UIC tomorrow.
Below is a graph of the pacing strategy from every men's world record in the 800m, mile, 5,000m and 10,000m events. "Interval number" on the x-axis refers to the lap number for the 800m and mile races, and the kilometer number for the 5,000 and 10,000m
So, looking globally, in the 800m event, the first lap is on average 2.4 seconds faster than the second (52s vs 54.4s). That's why I've often written that if you run an even-paced 800m, with a second lap that is as fast as the first, you're underperforming (more on this another time).
However, from the mile up to the 10,000m, the strategy (at least globally) is similar. Fast first interval, then a gradual reduction in pace in the middle, before what is called an "endspurt" in the final interval, be it the last lap, or the last kilometer (note that in the long distance races, that fast end is probably also due to the last lap, not a whole last kilometer).
Physiology of pacing - not as simple as it may seem
Now, the physiological basis for this pacing strategy is very interesting, but not really the topic of this post. Briefly, if you've always been told that you slow down because you become anaerobic, because lactate affects your muscles, or because of other chemicals that run out or accumulate, then you have a very difficult time explaining how the fastest part of the race comes at the end. Does the muscle suddenly overcome the inhibitive effects of the chemicals? Does the muscle suddenly 'tolerate' being anaerobic to allow a shift in pace that is often as large at 6 seconds over the final lap?
The peripheral fatigue explanation doesn't allow that. The only explanation is that the athlete had some reserve during those middle kilometers when they were slowing down. This is obvious to most people - except if they're taught otherwise! Most athletes and coaches will readily agree that there is reserve, but this reserve was not acknowledged in exercise physiology, amazingly, until quite recently.
What does this reserve look like? Well, there is evidence that the activation of muscle is increased at the end of the event, when the finish line is in sight. Also, it's pretty obvious that when you are running a 10,000m race, you are not going "all-out" and so there is without doubt a reserve capacity. During the middle part of the race, for reasons I won't go into here, you don't use that reserve, but at the end, your brain allows you to use it and you speed up. So now, you should be asking how large that reserve is, how it is maintained, and whether you can 'eat into it' to run faster? That is another series of posts altogether, and I must confess, I don't know the answers to those questions, I can only theorize!
Simple observations, complex physiology! I won't go into more detail on this, but if you'd like to read further, then I'd recommend two review papers which I had published earlier this year, in which I did theorize and propose a model for how the brain regulates performance (Tucker, BJSM, 43(6), 2009)
Pacing and performance limits - a change in how records are paced
In any event, this post is about the limits to performance, and for that, we must look more closely at the 5,000m world record. The graph below somewhat arbitrarily breaks 32 world records into three eras:
- 1920 to 1953 (9 records)
- 1954 to 1973 (11 records)
- 1974 to present (12 records)
It will be pretty obvious that something has changed in the way the world records have been paced. The first era, from 1920 to 1953, reveals a typical pacing strategy where the start is fast, the middle gets slower and slower, and the final kilometer is very fast.
The second era starts to look a little different - the drop in pace from 1 to 2km is not as large, and overall, the mid-race slowdown is nowhere near as pronounced. However, the same pattern exists, with kilometers 1 and 5 being significantly faster than the middle kilometers.
And then comes the third era. There is no longer a slowing down in the middle part of the race, and the only 'pacing anomaly' is a faster final kilometer. This period, from 1974 onwards, has seen the 5,000m WR fall from 13:13 to 12:37 (36 seconds), but the difference in the first kilometer is only 4 seconds.
So what is the significance of this change? Well, there are a lot of differences between the present era and the 1950s, for example. Training methods, technology, diet, equipment including shoes and track surface, genes (since the population is different in the third era - it's dominated by African runners), and dare one say, doping...do these differences explain why not only the pace, but the pacing strategy has improved? Perhaps - that is an interesting debate!
(Incidentally, if you look at the 10,000m record, you find exactly the same thing - a distinct trend towards even pacing has driven most of the performance improvements in the world records)
Eroding the reserve to reach the limit
So what does this mean for limits to performance? Well, it suggests that the absolute limit may be reached when the optimal pacing strategy is achieved. That would be the point where the reserve is entirely 'eroded' and the athlete is performing to the maximum of their physiological capacity on that day.
Of course, you may be thinking that the reserve and 'maximal physiological capacity' is dependent on training and fitness. And I agree, which is why it's possible that as you read this, you may believe you're already pacing yourself optimally, but you fancy that you can still improve with training. Your reserve (both on the day, and over months of training) is larger.
But we're talking about the elite here, and the very best elite athletes in the world at very close to the peak of their fitness. And once you're looking at this group, the notion of knocking 2 seconds per kilometer off becomes a lot more difficult. Their reserve, collectively, is much smaller, and even months of training doesn't change that. Haile Gebrselassie illustrated this in Berlin, when he ran close to perfect even pace the whole way and then imploded in the final 7km. That suggests that he had run just a little too fast for the conditions and for his physiological capacity, which means he's very, very close to the limits of what he is able to achieve.
Someone else may come along (maybe Wanjiru this weekend), but they're not going to blow records away to the tune of even 2 seconds per kilometer anymore! The reserve available to an elite athlete, even with many months of training, is much smaller than to your 20-min 5k runner. That's why the only improvements in the last ten world records have come from running the middle kilometers faster, and not just running the whole race faster, which is what you'd do if you went out and ran 19-minutes after 3 months of good training.
The key is in the reserve, and for the very best runners in the world, at the very peak of their fitness, there is very little of it. But that which exists represents the margin by which performances can be improved, until performances are almost perfectly even-paced and only minute improvements will be possible. Perhaps the limit to performance is near.
Chicago Marathon - even paced for the WR?
So it's in that light that I'm looking forward to Sunday's race, in the hope that Wanjiru does go out and try to get this record. If he hits halfway in 61:40, I dare say it would be too fast. That would put him on course for a 2:03:20, which is a 40-second improvement, and I believe that's too great. I'd rather suggest a 61:55 for the first half, with something close to that on the second. Time will tell!
More to come from Chicago in the coming days!