I realize I’m well off the news timelines as far as London goes, but the race, co-incidentally, did a lot to provide context to the question of pacing and the limits to human performance (that is, fatigue), so it’s a lead in to the presentation which is at the bottom.
London Marathon – pacing precision
First, London. Won by Tsegay Kebede, the time of 2:06:04 the slowest since 2007, the race was notable for an attritional second half that saw the lead change five or six times, clear breaks come back and eventually, the athlete who died least, possibly because of a stitch at around 25km which prevented him from responding to the early surges, came through to win.
The story is in the 5km splits shown below. Kebede’s splits are shown in blue, while those of Emmanuel Mutai, who finished second, are in red.
The first 10km in London is always fast, but this year was particularly quick. The result was that even with a slight drop in pace from 10km to 20km, the split at halfway was 61:34. That’s not necessarily a disaster, but it was constructed “badly” in the sense that it was a little too fast early and was produced slowing down. Interestingly enough, I spoke to one of the pacemakers at the Official after party, and he said that they were asking for the pace to be slowed, but Emmanuel Mutai was driving them to go faster. He said that a 62:00 at halfway would have been perfect.
But, 61:34 it was, and then the race’s ultimate slow time was created, because the pace was actually lifted. A 14:30 split (1:59:28 marathon pace, so a significant ramp up in pace) from 20km to 25km broke the race open, and from then on, it was always going to be a matter of survival. 14:49 for the next 5km is what saw the big time gaps appear, and Kebede was actually dropped, later blaming a stitch for his inability to follow that pace. That’s where their lines part company in the graph above – 19 seconds was the gap at 30km, because Kebede dropped off faster and was outside the top four.
At the front, Emmanuel Mutai, then Biwott and Abshero and Lilesa, and then Mutai again, all took turns in the lead, making what appeared to be breaks, but they were reeled in, despite a progressively dropping pace. When you see a lead that keeps changing even though the overall pace is getting slower and slower, then it means that leads are being established not because of breaks, but rather because of failures, and that in turn means it’s a matter of time before the wheels fall off in a major way.That happened to Lilesa, then Biwott (they lost 1:05 and 2:35 in the final 7km), but Kebede was able to hang onto something like a respectable pace over the final 7km. You’ll see in the graph that from 25km to 35km, he was slower than Mutai, losing time. But from 35km to 40km he clawed some of it back, and then the big change happened in the final 2km, where Mutai really did fall apart.
Mutai’s final 2.2km were run in 7:46, and that’s where a lead of 28 seconds was turned into a deficit of 29 seconds by the finish line! For comparison, Priscah Jeptoo, who won the women’s race, covered this segment in 7:23.
So, the men really did pay for the fast early start, but more than that, it was the attack at 20km, off that fast start, that did the damage. It remains a fact that only once in history has a man run both halves in a marathon in under 62 minutes. That was Patrick Makau, who broke 62 min twice on route to his current world record. Others have run negative splits with a 61:xx second half, but the London race highlighted just how precise the pace needs to be before it becomes ‘suicidal’, at least for record purposes. Racing is a different story, of course.
The physiology of pacing strategy, and the limit to performance
That then leads into a discussion of pacing, fatigue and the limits to performance. At the London Marathon conference, a fellow speaker, Doug Casa, and I had some great discussions about athletics, and while we agree on many things, one that we diverged on was the possibility of a two-hour marathon. Doug firmly believes it is imminent, and that he’ll see it soon (Doug also believes women will go under 4-minutes for the mile soon, which is absolutely not possible in this lifetime).
My opinion is different – I told him that unless he can figure out how to cryogenically freeze himself and watch London in maybe 80 years from now, he has no hope of seeing that happen! Even then, I’m not convinced.
The basis for my saying that is at least partly found in the graph above. It shows us that even the very best fail when they don’t get the pacing right, and that means they are right on the limit of performance. If you consider the pace in London, you had 61:34 at halfway. That was set up by 14:23 and 14:33 splits for the first 10km. Too fast, but by how much? Perhaps 10 seconds per five kilometers, so we are talking a margin for error of about 2 seconds per kilometer being the difference between a complete blowout of the world’s best runners, which reduces 2:05 runners to running 17:30 5km pace, and maintaining the pace to run something under 2:05.
Now, if that is what happens when they run at 2:03 pace, and then surge to a 2:00 pace, imagine how much longer we will have to wait to see a runner capable of running every single 5km segment at 2:00 pace? It is, for now, inconceivable that anyone can run 14:30 per 5km eight times consecutively. In London 2013, that happened twice and it broke one of the best fields ever assembled into splinters.
Remember, you didn’t have one runner fail at 61:34 pace with a 14:30 surge. It was everyone. Maybe seven or eight of the best 15 men in the world pushed their physiology over the edge with that racing strategy, and not one was able to come through it without some damage. Kebede was the best survivor, but even he “limped” home with a 64:28 second half (and final 10km of around 32:00). A 2-hour marathon is nowhere near imminent, it’s a long way away, and breaking the race down into its component paces is one way to show that.
The other is to recognize that a runner who is capable of running back-to-back half marathons in 59:59 will be a guy capable of running a single half marathon in about 57 minutes. Currently, a 59 min half marathon runner can hit 62 min in a half and maintain the pace. So, until there is a 57 min half marathon runner, don’t hold your breath. And of course, a 56 min half marathon runner is capable of running back-to-back 10km races in about 26:30. That is a runner who would be able to run a 10km in about 25:40. So when we start seeing 10km and 21km times drop to 25:40 and 57 minutes, then I’ll agree with Doug and the other 2-hour marathon advocates!
What pacing means
Back to pacing, which leads us to the question of how that precision is achieved? What physiological basis is there for such a “fragile” line between optimal and ‘failure’? What is the body responding to in order to slow a runner down when the difference between holding the pace and crashing is as small as 1% too fast early on?
The theory is that we pace ourselves because we are selecting the optimal exercise intensity that allows us to:
- use our available energy at the optimal rate, not too fast or too slow
- gain heat slowly enough that we’ll finish, but not so slowly that we aren’t performing at a high enough intensity
- accumulate metabolites at a low enough rate to not be overwhelmed by them
- meet oxygen requirements of muscle, brain and other tissues
- compete with other runners, the clock or whatever other motivational factors impact on performance
Pacing, then, is the physiological equivalent of sticking to a budget. There is a plan, one which we are not fully aware of, but which covers all aspects of physiology, every system in the body. It then manages our intensity, by adjusting how much muscle we are able to activate (we measure this as EMG, as you will see in the presentation below), so that we don’t deplete reserves or accumulate limiting heat or metabolites. Doing that would result in, in order:
- A failed performance because we’d reach a critical level of hyperthermia, or energy depletion, or metabolite accumulation (or any other factor, depending on the context of exercise, see slides below) before the finish line. That’s called a bad day out, and it happens because performance is ultimately going to be limited by one of more physiological systems. Pacing aims to ensure that this never happens
- Bodily harm. In theory, it is possible to push so far beyond those performance limits that we run ourselves into physiological trouble. The line for this is higher than it is for performance – we would fail at exercise before our bodies fail, but it does happen.In fact, a really good opinion insight on this has just been written by pacing researchers led by Zig St Clair Gibson and Carl Foster, and it’s called “Crawling to the Finish Line: Why do Endurance Runners Collapse? : Implications for Understanding of Mechanisms Underlying Pacing and Fatigue”. I recommend it as a good discussion of this very topic
The presentation – pacing, performance limits and fatigue
As for the rest, they are details. Fascinating details, of course (in my biased opinion), and they’re the subject of the presentation I gave in London, which you can see below. My focus is on heat, because that’s a great model to illustrate the difference between a regulated system and a limited system. In exercise physiology, both exist, but understanding performance regulation is the recognition that pacing has a physiological purpose, and that we don’t ‘dumbly’ run to the point of collapse.
I also devote quite a bit of time in the talk to some really interesting studies by Amman et al, who have shown that the body is trying to defend peripheral muscle function, and that if you block Type IV afferents, you can play havoc with pacing strategies and those muscle properties.
These presentations require talking, and so they may not quite make sense when being viewed like this. The gaps can always be filled, however, and hopefully this introduces the concepts of pacing and fatigue and how various factors, including motivation, competition, emotional state, and of course the various physiological inputs affect our performance limits and pacing.
Here is that talk, enjoy!
More reading and forthcoming attractions!
And, for some more reading on this, two reviews I wrote, published in the British Journal of Sports Medicine:
- An anticipatory model for the regulation of pacing strategy – how perceived exertion controls physiology
- A review of pacing strategies used by athletes in a range of different events, and their physiological significant
And here’s that paper from St Clair Gibson et al, published just last month: Crawling to the finish line
And finally, as mentioned, I’m off to the UK again next week, first to join the SA Sevens team for the season ending tournament, and then at the invitation of the English Rugby Football Union for a symposium on talent ID and development. I’ll share what I can, when I can.
This post is part of the thread: Marathon Analysis – an ongoing story on this site. View the thread timeline for more context on this post.