Wednesday, April 3, 2013

Optimal Pacing?: Lessons From The Best

Lately I've been thinking about heading out to a race and leaving my watch behind. I know some people that have the confidence to do so. My wife has been running without a watch all year, preferring to trust her judgement and also not being pressured into chasing splits. She is an ultramarathon runner and in some ways, certainly outside the level of elite runners, as one heads into longer distances the margins are less fine. I'm focusing my peak performances on 5km and 10km races, using 15km and a few half marathon races for stamina and endurance training. For me the margins are fine. If you lose 30 seconds on a kilometre in a 50km race, one can strategize on the fly to get that time back. For me 30s in a 5km and even a 10km race will take a serious effort to claw back, with these races lasting ~18:00 and ~38:00 for me.

Pacing is such an integral part of racing. There are many ways to run a given time for a distance, but one way will often be a bit more comfortable than another. I've been spectating at a lot of races, supporting my wife and my club mates from Varsity Kudus, and I'll often seen two athletes finishing side by side, yet in contrasting condition. When one peruses running resources, the advise is very clear that for middle, long and ultra distances, an even pace, where conditions will allow, is most ideal. It suggests increasing effort as a race progresses but if the effort has been distributed properly one should be able to maintaining pace at the end of a race and even have a bit in the tank for a little kick!

I came across an article posted in Sweat Science on the Runner's World site called The Disadvantages of Perfect Pacing which really got me thinking. Last year I inadvertently perfected pacing. I started off running very negative split races, like running the last 11.1km of a half marathon quicker than the first 10km to eventually running even splits on a course with some climbs. Part of it was that because I was in a rapid phase of improvement I could rarely take a similar strategy into races even a few weeks apart. So I tended to find a comfortable effort and hold on to it. But now while I'm still improving the gains are becoming comparatively marginal and I'm having to think about what I'm doing constantly. The Sweat Science article really stood out for me because it presents a scenario where very very good athletes are shown to eschew the sensibilities of even pacing and rather push the envelope. That is encapsulated by this rather excellent quote from the article:

According to conventional physiology, steady pacing is the most efficient way to ration your energy stores. Its logic has been taken for granted since Aesop and his tortoise. But it is inherently limiting: to run at an even pace, you have to decide on your final finishing time, and thus set a ceiling on your potential achievement, before the starting gun fires. As a result, even pacing may produce better results on average, but it is less likely to produce dramatic outliers: jaw-droppingly fast (or slow) times.
That quote really speaks to me. My goals for the year and with my margins for error becoming finer as I get faster suggests that at some point I will HAVE to ditch sensibility and go for it. In fact when I ran 37:46 at the Great Run last year that's precisely what I did. At the start I found myself in a group of runners that at races have typically run around or just faster than my pace. On this day, I felt really good, so I surged ahead and chased another group with more runners I recognized who I only really see at the start and the finish of races! My goal was not to catch them, but to keep them in my sights. As the kilometres ticked by, this group was edging further away but next thing I was 8km into the race. My legs started to burn but I hung on and shattered my PB, producing by my standards an outlier performance. Nothing in my training had suggested I could run that fast. Had I raced as I had planned, I would have run 38:30. Sometimes it pays to be adventurous.

I suppose the question depends really on the ambition of the runner. I was watching the Lisbon Half Marathon just the other day, and the Kenyan Bernard Koech set out with a mission to to break Zersenay Tadese's astonishing 58:23 world record. He set out a blistering pace, going through 10km in 27:17, 57:34 pace! But as the race wore on he was still running quickly but by 16km was already outside world record pace, in 45:20 and by the finish over a minute outside the record in 59:54. His pace for the first 10km was a mind boggling 2:44 min/km but the for the remaining 11.1km 2:56 min/km. Contrast this with Tadese who went through 10km in 27:53, 2:47 min/km but covered the next 11.1km in 30:30 at a pace of 2:45 min/km. He just ramped his pace up and even his final 1.1km was under his average pace for the WR. And having seen both races he looked in a lot better shape at the end than Koech. And remarkably enough, Koech was 'only' 2s/km faster than the WR for that first 10km but thereafter 10s/km slower.

One might think that Elite running has nothing to do with us mere mortals but I'm a firm believer that if I'm going to be a better runner these are the people I want to be learning from. It's not about emulating them but taking the principles of their approaches and applying them appropriately to my situation. And on that pacing front I've certainlu been more Bernard Koech that Zersenay Tadese in all but two of my six 10km races this year. Possibly like Koech's approach, I always look to bank time by building myself a buffer but neglect to appreciate how quickly one can fall off the cliff even in a race as short as 10km. Have a look at these splits from my last 10km race:
  1. 3:24
  2. 3:26
  3. 3:30
  4. 3:35
  5. 3:39
  6. 3:44
  7. 3:45
  8. 3:44
  9. 3:42
  10. 3:45
These splits are from the Bonitas 10km Challenge from 23rd March in Cape Town. This was 1) the flattest race I have ever run and 2) at the coast. The final time was 36:13, a 23 second PB which is about the only real positive from the race. All things considered this was a real opportunity for me to real test myself and I ruined it by going through 3km in 10:20. That was excellent pacing for a 5km. Once I had hit 4km in 13:55 I would have had the confidence to put out a 3:20 final kilometre but alas I still had 6km to go. A 5km split of 17:34 would be as good as it got as I limped home in 18:39 for the remaining 5km. Consider that at Deloitte I finished with a 3:28 final kilometer on course with ~100m elevation gain.

My goal for the Bonitas race was simply to break 36:00. I fell 14 seconds short and in hindsight my fate was sealed at that 3km mark. Had I gone out at 3:36 min/km pace, I would have gone through in 10:48, and I am 100% certain  I wouldn't be running at 3:44 pace in the second half, outside my average pace for most races I have run this year. A 10:48 first 3km, which would been closer to 18:00 for 5km would have still been aggressive, but with flatness of the course would have been conservative enough in the circumstances given that I still finished with an average pace of 3:37 min/km. A 3:27 min/km pace for the first third, when aiming for 3:36 in average, was in running terms suicidal and in some ways I'm surprised I got that close to 36:00 in the end.

My take home message from the Sweat Science article, studying some elite performances and then looking at my own running is that a little bravery goes a long way but must be married to some sensibility. I'm certainly on the quest for that outlier performance. I believe that at this point in time, I am a regulation 37 minute 10km runner  and if I work a little bit harder I will drop into the 36 minute range. But with a little bit of sensible aggression that 35 minute 10km will surely come, even in the thin air of the highveld! Then I can shift my regulation performance to 36 minute and target that sub 35 minute dream.

This weekend I have the Mazda Athletic Club 10km. I did the 5km last year and broke 20 minutes for the first time. The 10km is described by Runner's World as 'flat, yet challenging', so standard fare for the highveld then! I never like to make excuses for my runs but I was ill before the Bonitas 10km Challenge then had an 18 hour drive to Cape Town two days before. And though I don't believe it affected my performance, it has meant that training has not gone according to plan. The 29:01 I did at the Two Oceans 8km Fun Run however suggests that my form is good, if a little rusty, and as I said a few lapses in concentration aside, I could have run around 28:50 on the day, not bad effort for a 15:00 race. However, the quality has evaporated from my training in the last three weeks. On Saturday though I still plan to be aggressive as I search for my outlier performance of the season but plan to be more Tadese than Koech, sensibly aggressive, looking for closer to 11:00 for the first 3km and staying away from the 10:30 level. I certainly believe in running close the edge and hopefully having enough in the tank for one finally kick. BUT, no more 5km PBs in 10km races!!!

Happy racing!


  1. I'm a believer in going for that outlier performance! To run a sub-36 you need to be on track at 5km (maybe 17:34 is too quick for now but it was at sea-level and you wouldn't have had much experience of the altitude effect and how it might play out). For an interesting read, check out
    for some interesting pacing analysis at the top level.

    In my opinion, the 18-hour drive to CT would have had an effect.

    1. Maybe I am underplaying the effect of the drive, I'm still knackered now after the drive back up! I definitely think I got excited at the first km, it really was so easy! I don't think I've opened with a 3:24 before, usually 3:30-3:35, so perhaps a 17:45 but a more even one with splits in 3:32-3:36 range would been better, then I could have closed with a second half closer to 18:00. I'm still learning :)

      Shot for the Science of Sport link. I found a couple other posts but this one is more detailed.