THANK YOU FOR VISITING SWEATSCIENCE.COM!
My new Sweat Science columns are being published at www.outsideonline.com/sweatscience. Also check out my new book, THE EXPLORER'S GENE: Why We Seek Big Challenges, New Flavors, and the Blank Spots on the Map, published in March 2025.
- Alex Hutchinson (@sweatscience)
***
‘Tis the season for championship track racing — and with it, the annual moaning about slow, tactical middle-distance races. At both the U.S. and Canadian national championships last weekend, the men’s 1500-metre races went very slowly (until, late in the race, they suddenly started going very fast). On the message boards, people started the usual criticisms of everyone who didn’t win, saying that they should have taken the lead and made the race faster from the start — much like Christin Wurth-Thomas did in the U.S. women’s 1500m. (The fact that Wurth-Thomas, who had the fastest seed time, was passed by three women in the final straightaway and thus failed to qualify for the World Championships, seems lost on these critics.)
Anyway, as I always do, I got sucked into the debate too, in a thread on tnfnorth. Given that USATF results now offer complete splits for every lap of every race, it’s possible to do a much more detailed analysis of tactics than it used to be. Out of interest, I looked at the three semifinal heats of the USATF men’s 1500m. There were three intermediate splits (at 300m, 700m and 1100m) taken in each race, which means a total of nine intermediate leads recorded. Seven men filled these nine leading spots; none of them qualified for the final.
Of course, this data didn’t convince anyone. Just a fluke, they said. So I’ve decided to take it a bit further. I looked back at World Championship results between 1997 (the earliest year for which intermediate leaders are listed in the results) and 2009 (the most recent championship). Here is how the first-lap leaders (after 400m) fared in the 23 quarter-final heats of the men’s 1500m in that timespan:
As expected, there’s a full range of results — leading the first lap doesn’t guarantee either success or failure. But there’s a pretty pronounced tilt toward the right-hand side of the graph. Indeed, 35% of first-lap leaders managed to either hang in the top six or take their race out fast enough to get a time qualifier. In contrast, 65% failed to move on to the semis.
Now, the tactics in qualifying races (where the goal is simply to place within the top N) are obviously different than those in finals (where the goal is to place as highly as possible, with every place counting). There’s not as much data for the finals, but here they are nonetheless:
Let me emphasize that race tactics are enormously complicated, dictated by the individual’s physiology, psychology, abilities relative to the rest of the field, weather conditions, and so on. I have beliefs of varying strengths regarding many of these factors — but I don’t believe this data answers any specific questions. It does, however, give a snapshot of how all these factors play out in the real world when they’re mixed together. And it suggests (to me, at least) that taking the lead during the first lap of a championship men’s 1500m race rarely ends well.
Suggestion:
For each runner in each event, compare their position at the end of the first lap, to their position at the end of the event.
Scatterplot these four items: entire data set, median finish position for each first lap position, average finish position for each first lap position, and percentage of each first lap position that qualified.
I’m thinking those graphs will show that being very near the front, but not in front, is optimal.
I’m also thinking that the leader in lap 1 has a better outcome than the last position at the end of lap one. 🙂
Isn’t this a bit of a self-fulfilling prophecy? If conventional wisdom says it’s bad strategy to lead the first lap, then the best runners aren’t going to lead.
How do you know there’s not teamwork and pacing being done in the early rounds even of a championship race? It’s not impossible that some of those leaders on the first lap were sacrificing themselves for a team mate by forcing a faster pace.
@Jim Ley
Very good point, which is exactly why I suspect that – broadly speaking – position at the end of lap one correlates positively and strongly with the final position, and why I suspect that – broadly speaking – the top spot at the end of lap one is not the “best” position for predicting the top spot at the end of the race.
An analysis like this can’t be done so simplistically as “what happens to #1?” It has to include a bit of what happens to all positions, and ideally should be looked at in quantiles, in terms of averages/medians/odds, etc.
@lurker
As it happens, I did exactly the scatterplot you suggest for the USATF 1500m results. The problem is that data in that form requires a pretty big data set to draw meaningful conclusions. My SUBJECTIVE impression of what was essentially a Rorschacht blot was that (a) being right at the front was bad; (b) being very close to the front (~2-5) was good; (c) being in the back half of the field was bad. But there wasn’t enough data to really draw that conclusion.
Unfortunately, the IAAF data only lists intermediate leaders (unlike the USATF data that gives complete splits for every runner). What we’re really interested in, of course, is not “What happens to the first lap leader?” but rather “What strategies were most commonly followed by the runners who ended up qualifying/medalling/winning?” But the IAAF data doesn’t allow us to analyze that; and the USATF data isn’t a big enough data set.
@Dave Munger
There’s certainly a very long list of possible confounding factors in this small data set. But I don’t think there’s any reason to think that the best runners in this field are any more aware of or susceptible to conventional wisdom than the worst runners. This is the world championships: even the “losers” are typically national record holders who’ve been racing internationally for years. Everyone in the race understands what usually happens to leaders.
Having said that, it’s reasonable to assume that the lowest-ranked runners in the field, who are the least likely to qualify or medal in a “business as usual” scenario, have the greatest incentive to shake things up and try to catch people by surprise by leading. Similarly, if you’re the 7th or 8th fastest runner in a qualifying heat, and thus unlikely to qualify automatically if the race goes on form, then you have have an incentive to make your heat faster than the other heats in the hope that 7th or 8th spot will get you one of the qualify-on-time spots.
Everyone in the races understands all these dynamics. The question is: do they work? That’s what I hope the data gives us a peak at, though it’s obviously too small a sample to draw firm conclusions.
@Jim Ley
Pacing and teamwork were certainly factors in the 1500m finals in 1999 and 2001, when El Guerrouj had a teammate pace him. So this skews the finals data. In general, I’m pretty sure there was little or no teamwork in the quarterfinal data — none the eventual finallists would have had much trouble placing in the top six of their heats, and it’s generally rare for teammates to be in the same heat (max three runners per country). Moreover, if a teammate is going to help you win a medal, he needs to make sure he himself gets through the qualifying rounds: a “sacrifice” in the quarterfinals would be a fairly meaningless gesture.
That being said, factors like this are certainly possibilities that can’t be eliminated in this type of analysis. All you can do is collect as big a data set as possible and look for patterns.
My question is more for those who have no chance of winning because their seed time is a good 10sec slower than the top guys. Why go out slow and finish 10th when you can go out fast and finish 10th but with a better time?
@John Lofranco
To make the question more realistic, it should be “Why go out slow and finish 10th when you can go out fast and finish 11th with a better time?”
When you put it that way, we’re no longer arguing about tactics — we’re arguing about goals and values. (And I think it’s a confusion between those different ways of judging race results that leads to many of the message-board debates on 1500 tactics.)
Personally, I ran in four 1500 national championship finals, with finishes of (I think!) 4, 8, 9, 12. Every one of those races was tactical, and in every case I ran with no goal other than to place as highly as possible. The entire rest of my season (at Jerome, McGill, etc.) was devoted to running as fast as possible, but this one race was devoted to placing as high as possible. And believe me, though I was never a threat to win, there was a HUGE difference between how I viewed those finishes. The 4th place finish was the highlight of my career, and possibly the greatest race I ran (though it was by no means the fastest). The others were relative disappointments depending on my fitness at the time (8th was a huge disappointment, 9th was actually a moderate victory give that I’d started running after a stress fracture only a few weeks earlier).
I guess what I’m saying is: IF you value time more than place in a national championship, then of course you’re right that runners shouldn’t sit back and let the pace go slow. But a key assumption there is that the runners in question would rather place 11th in a fast time than 10th in a slow time. And as a competitor, that wasn’t my reality.