Welcome to the Science of Sport, where we bring you the second, third, and fourth level of analysis you will not find anywhere else.

Be it doping in sport, hot topics like Caster Semenya or Oscar Pistorius, or the dehydration myth, we try to translate the science behind sports and sports performance.

Consider a donation if you like what you see here!

Did you know?
We published The Runner's Body in May 2009. With an average 4.4/5 stars on Amazon.com, it has been receiving positive reviews from runners and non-runners alike.

Available for the Kindle and also in the traditional paper back. It will make a great gift for the runners you know, and helps support our work here on The Science of Sport.

Friday, August 22, 2008

Beijing 2008: Men's 800m

Men's 800m: Anyone's race and a discussion of 800m pacing physiology

The final day of athletics competition in Beijing brings with it one of our most anticipated races of the Games - the Men's 800m event. As we wrote yesterday, the event has already thrown up some huge surprises, because three of the big favourites failed to even qualify for the final! Yuriy Borzakovskiy Mbulaeni Mulaudzi, and Abubaker Kaki Kamis all failed to advance from their semifinals in what must be one of the most open, competitive and unpredictable events in athletics.

The race without the favourites should be fascinating, and any one of the eight athletes could win it. Most impressive in qualifying has been Wilfred Bungei (who has the experience to boot), and so I'd probably bet on him if forced to. But really, any one of the other seven could beat him - look for Kirwa Yego (also of Kenya) to be competitive, and for a "dark-horse", keep an eye on Gary Reed of Canada.

Today's post is not so much an event preview, however, as it is an examination of the fascinating physiology of the event. So here is the physiology of 800m running, which will hopefully provide some more insight on the events that will unfold in Beijing tomorrow.

A tactical game: Pacing and strategy as vital as physiology

The 800m distance is fascinating because it straddles the divide between what people usually refer to as “sprinting” and “middle distance” running. To some, it is the first of the middle distance events, whereas to others, it’s the last of the sprints. Of course, using such jargon can pose challenges, but generally, when people refer to a sprint, they refer to an event where the athlete goes ‘flat out’. This is of course never true, because even in a 200m race, there is some pacing, as evidenced by people who go out a little too fast and end up faltering in the final 40 to 50m! By failing to pace herself properly, Sanya Richards demonstrated the value of pacing in the 400m event, when she got it wrong and was reeled in by two athletes in the final earlier this week in Beijing. So pacing is certainly as vital in sprints as in middle distance races, which brings us to an analysis of the 800m event.

Coaches (and physiologists) have often spoken of an aerobic-anaerobic divide for different events, and they often refer to the 800m distance as being a 50% split for each. That is, they say that approximately 50% of the energy comes from aerobic sources, 50% from anaerobic. This is a contentious issue in itself, one that I would argue with, as recent evidence suggests there is no black and white split between the energy sources. It's more likely, based on recent work, that the aerobic component is far larger - even in 400m running, it's almost 50-50.

What is optimal pacing?

It is often said that the ideal way to run an endurance race is to aim for what are called ‘even splits’. In other words, the first half and second half should be run in the same time. If you are a 10km runner, for example, the ideal strategy seems to be to run even pace the whole way. An underperformance happens when you either start too fast and slow down (called a positive split – the first half will be faster) or you finish very fast, running the second half faster than the first (called a negative split). For most recreational runners, aiming for a negative split is probably the prudent approach, recommended as a safety first option.

To fully appreciate pacing strategies, you have to look at a range of different events, and so we'll look briefly at the track events from 800m upwards. We'll do it in reverse order, and work our way backwards to the 800m event.

The graph below shows the pacing strategy from world records for 800m, the mile, 5000m and 10000m. Of course, we make the assumption that the pacing strategies in the world records are optimal, which is of course not always true. Sometimes, world records are set in tactical races, and that obviously affects the strategy. Sometimes, they are set IN SPITE of, rather than because of pacing strategy. But generally, if you take 30 world records, you can be pretty sure the guy running the time is running on the limit of performance, and so the pacing strategy is at least close to ideal.

10 000m distance – even pace is the way to go, with a fast finish

In the men’s 10000m distance, 34 world records have been set in the modern era. It’s quite clear from the graph that on average, the race is even paced, with a fast start, more consistent period in the middle, and the final kilometer is fastest. In fact, in 33 out of the 34 world records, the final kilometer was the fastest of the race. What this means, practically, is that even the elite have left themselves something in reserve for the final kilometer. You may be thinking that this indicates that the athletes are not performing maximally, because surely, if you have enough for a sprint at the end, you might have been able to go quicker in the middle part? And you’d probably be right, and that's one of the big unknowns in exercise science - what does this reserve mean, and how can it be accessed sooner (or more fully?)

We did discuss this issue in our series on Fatigue earlier this year, and you might spend some time looking at that for futher insights. The very summarized version is that the brain prevents the runner from ever fully accessing muslce motor units until the end of the bout, when the "danger" has passed. The "sprint" at the end is the manifestation of that reserve, but it's not a simple matter of accessing sooner, because it serves an important regulatory function as it protects against possible damage during exercise. The great atheltes go closer to the "limit" than others, it's part (though not all) of their advantage.

Men’s 5000m – similar to the 10000m, the pace is even, with a final kilometer kick

There have been 32 world records in the 5000m event, and the pacing is very similar to that of the 10000m. The first and final kilometers are faster than the middle three kilometers, which again suggests that the middle kilometers are somewhat ‘conservative’. The final kilometer has been fastest in 21 of the 32 world records. In the other 11, it has been the first kilometer that is fastest. So for those wondering about tactics, they are certainly in play, but they have never once produced a fastest kilometer in the middle of a 5km race.

Men’s 1 mile – a much more even pace

The men’s mile event begins to get down into the range where speeding up at the end is a lot more difficult to do. On average, the final lap is run in the same speed as the first lap. In fact, in more than half the world records set, the first lap is actually faster.

There is still a drop in pace in the middle part of the race, but the overall strategy is even, in contrast to the longer races, where the fast final kilometer ensures that the second half is usually faster. So here, in the even that lasts about 4 minutes, we see a subtle change, which has physiological relevance. Because the even pace in the mile suggests we are getting down to the point where speeding up at the end is becoming increasingly difficult for optimal performance. Which brings us to the 800m race…

800 m – it’s not possible to run optimal times with a faster second lap

In the 800m event, 26 world records have been set. The graph below shows the average lap times in these 26 races. It’s immediately clear that the second half is quite a lot slower than the first. Some of you may be thinking, hang on a moment, what about the 200m splits? Unfortunately, they are not available for the 26 world records, but in the ones they are available, they follow the same pattern – the first 200m is fastest, followed by the second, and the pace gets slower and slower.

So this is a departure from what we’ve seen before – suddenly, speeding up at the end of an 800m doesn’t happen. In fact, in the 26 world records, the second lap has only been faster than the first on ONLY two occasions. Therefore, a world record seems to require that you run a fast first lap, and then hang on in the second, but speeding up does not appear to be an option.

Some of you may now be questioning this statement. Among the biggest challenges would be the assumption that you’re seeing ‘optimal performances’. And of course, this is true. If a guy goes out and run 1:46, who is to say that is not optimal? Perhaps it is. However, I still maintain that with this pacing strategy observed in 24 out of 2
6 world records, the best way to run the race is to run the first lap faster than the second. On average, the difference is 2 seconds. This means a first lap of 50 seconds would be followed by a second lap of 52 seconds.

What is even more interesting is that the two fastest second lap times ever achieved in 800 m world record performances were run in 1972 and 1966 respectively. The graph below shows the lap times from all the world records, and if you look at the panel on the right, you will see that the second lap time of a world record performance has not improved in 35 years, since Dave Wottle broke the world record with a time of 1:44.3 (min:s) and a second lap of 51.40 seconds in 1972.

The current world record holder, Wilson Kipketer, has broken the world record on three occasions, with second lap times of 52.12, 52.90 and 51.80 seconds. Therefore, a 3.2 second reduction in the world record in the 800 m event between 1966 and 1997, from 1:44.3 to 1:41.11, has been achieved by running the first lap significantly faster, rather than an improved ability to increase running speed on the second lap.

The Figure below shows the lap times for the 26 world records in the 800m event. The left panel is the first lap, the right panel is the second lap

Another interesting fact is that the second lap is slower even in the Olympic Games, where the first lap is often tactical and slow. In other words, a slow, tactical first lap is still followed by a slower (on average) second lap, despite your perceptions that the athletes are "sprinting" for the line! The average first lap in the Olympic Games finals is 52.8 seconds and the second lap is 53.4 seconds.

What this suggests is that the ability to run faster during the second lap of an 800 m is limited, and so the optimal pacing strategy may consist of a faster start followed by a relatively slower second lap. I’m going to go out on a limb here and say that if you are an 800m athlete, or you are coaching an 800m athlete, if you want that athlete to run their best, you have to plan for a second lap that is about 2 to 3 seconds slower than the first. So, if the goal is 2 minutes, it’s not good enough to aim for a first lap of 60 seconds. It has to be 58, because if your athlete is going maximally, then he should slow down to a 61 something on the second lap, giving him a final time just under 2 minutes.

Similarly, if you want to break the world record, forget about running the second lap in 51 seconds. It’s not going to happen. Therefore, you must plan for a second lap of 52 seconds, which means the first lap must be 49 seconds, or faster. This is also an indication of the sort of speed needed to challenge Kipketer’s world record – you have to be able to run a 400m in 48.5 seconds as part of an 800m race. Your basic 400m speed therefore needs to be down in the 45’s, maybe 46 seconds (but that starts cutting it fine).

So looking ahead to the 800m final…

The 800m final is more likely to be a tactical affair, unless someone has decided that his best shot is to follow in the Women’s champion’s (Palema Jelimo) footsteps and go as fast as possible from the gun. Jelimo found a willing ally in Jepkosgei, which is unlikely in the men's race. So therefore, a slightly slower race might be expected, and this means the race will be roughly evenly-paced.

I suspect the first lap will be run in about 52 seconds (wild guess), meaning that the second lap will probably be run in about 52.5 seconds. You’ll note that this is still a slower second lap, which is really interesting from a physiological point of view – why can you speed up in a 1500m, 5000m and 10000m race, but not the 800m race? Is it a different type of fatigue? Again, I’ve said this before, but it’s worth repeating – we don’t really know what causes this, and if anyone tells you they do, they’re lying! It’s quite a mystery.

But for the race, it means that the athlete who has the ability to maintain speed is likely to come out on top. In the past, the former Olympic Champion, Yuriy Borsaikovsky, was the master of the "fast finish". I hope that by reading this, you now appreciate that this is not 100% correct – the truth is that he wasn't necessarily the fastest finisher, but rather that he was the least slowing finisher! So his strategy was to hang back over the first 400m, which added about 1 second to his lap time. But he was able to maintain a pace on the second lap that was much closer to the first lap, and appeared to have an incredible finish. I don't have the data, but I would suspect that in his major victories, he ran roughly even splits (between 0 and + 0/5 s), when the rest of the field were running a positive of at least 1 to 2 seconds.

Of course, when Borzakovskiy or anyone else goes out for a fast time, he has to run the first lap in 49 seconds - the speed is "limited" on that second lap and simply can't be much faster than a 52.

Anyway, returning to Beijing, the race will come down to the final 300m, where tactics will be incredibly important. As will speed-endurance – the ability to sustain a fast pace on the second lap, running as close to even splits as possible is a unique physiological ability. That is what makes the race so unpredictable. We saw the fastest man in the world (Kaki Kamis) "disappear" from the race in his semi-final, and so calling it will be difficult. But I'd guess that Bungei comes through, winning perhaps from Reed and Kirwa Yego. Then again, I thought Kaki Kamis and Borzakovskiy would be racing for gold!




Unknown said...

I think that Lydiard told Halberg that the 5,000 meter field would be resting up for the final lap with about 2-3 laps to go and so that was where he was to start his kick. He would catch them napping and the field would be unable to catch him at the finish. Somewhat similar advice was given to Snell who was the slowest man in the Olympic 800 meter field. It looked like he outkicked everyone else but actually he was slowing down the least.

Anonymous said...

Hi, Ross.

I like this entry. Although you may want to hold back on the "lying" conclusion. People can be misinformed, for example.

And to try and zero-in on the mystery of deceleration in the 400 and 800m, I'd like to throw in an observation made in the book by Frans Bosch and Ronald Klomp: race speed decreases suddenly instead of gradually. The authors attribute this to "an inability to maintain a reactive ['elastic' is the word you use] running technique". So it must be a neuro-muscular limitation, or even a property of tendinous structures. (Tendon resting length increases after maximal isometric muscular contractions. Could several fast and strong pulls in succession fundamentally turn it into a worn rubberband, at least temporarily?)

Of course, this assumes that in the shorter distances, there is no restraining order from the brain. Do you suggest this in your piece or were you really leaving it at "mystery"?

On a lighter note, how do you get any "real" work done these days?


Anonymous said...

What do you mean when you say that Snell was the "slowest" man in the Olympic 800-meter field? Was his 800 PR the slowest?

Anonymous said...

Who falters in the final 40-50 meters of the 200m race? Olympians? I ran track in college and don't know of a single person who paced themselves in the 200m. I'd have to imagine olympians have at least reached similar fitness levels and don't need to pace themselves either.

Anonymous said...

To the last poster first.

I think even if Bolt ran a straight line 200m without bend, hit the 100m (without showboating this time) in 9.60 then the second 100m would still be faster (due to running start) but his 10m splits over last 50m would be slower than the middle 100m.
So 'pacing' to an extent is done to achieve an optimal outcome, slower 100m but less slowing down over last 50m.

OK, re the 800. As I always say, observations and statistics don't tell you a cause-effect...that has to be measured in some other way. Just because an event occurs 99 times of 100, doesn't mean you can conclude that its theoptimal way.

Wottle's record all those years ago with a 51.3 last lap is a relatively (well all WR's are relatively same) great run.

Who is to say that a last lap could not have improved to less 50.5 (less of an improvement than total WR)in the meantime if runners conciously were doing first lap slower than that.

Who is to say that Wottle wasn't optimal.

Good food for thought though.

Ross Tucker and Jonathan Dugas said...

Hi Lem

Thanks a lot. On the note of "lying", that is intended for the experts who tend to get into one's face with the "fact" that they know the cause of fatigue. Of course, you're right that people are also misinformed (including us, that's the nature of science, after all - trying to be less misinformed by virtue of research). I'll actually go back and change that one.

To respond to your next observation, I'm not sure that it is sudden. A study by a guy called Ferro in 1999 looked at pacing in the 100, 200 and 400m events, and he found a gradual slowing down in pace. I think that "proper" pacing shows a gradual decline in speed in the sprint events (up to and including 800m, I might add),whereas "bad" pacing is characterised by a sudden drop in some cases, and a gradual drop in others.

I've not seen the data for 800m running to back this up, but certainly in 200m and 400m, the pace gets gradually slower, whereas Sanya Richards might have been more pronounced in her Beijing final, for example!

In the shorter distances, I think there is likely a restraining command from the brain, but it's regulated in the context of the overall exercise bout, and that means that a slowing of pace is "allowed" as a result of the faster earlier pace which ultimately maximizes performance.

In longer events, a fast early pace is not only performance limiting, it's also physiologically harmful (overheating, glycogen depletion etc.). These "limitations" are not in play during short duration exercise, and so the early pace can be faster, because the only thing that is happening is accumulation of hydrogen, depletion of ATP etc, but it's never allowed to become "disastrous", because the event is too short.

This is, however, what is responsible for slowing down, which is why in a 400m race, the first half is always a second or so faster than the second. And that is with a standing start on the first 200m. So the second half is in fact much slower than the first. The body allows that because optimal performance and "physiological protection" are not necessarily the same thing.

But, on the whole, it is a mystery, and we don't know the real explanation. I tried to develop a model on this in my PhD thesis, which I'd gladly send you digitally if you'd like (it's a much more "dry" scientific document, but I think you'd get through it no problem).

As for "real work", stress of my life! I need to figure out how to make a living off this, so it can become my "real work". Then I can do 2 articles a day, and have a lot more fun doing it! If anyone would like to start a fund for "Ross' real work replacement", please let me know!

Thanks for the support!


Ross Tucker and Jonathan Dugas said...

Hi Owen

I read your comment and I was confused, because I didn't recall writing that! Then I realised you were responding to another commenter! I don't know what he means. I suspect he means that Snell was the least accomplished "sprinter" in the race, perhaps with the slowest 400m time. That's often the "jargon" used, isn't it? Like saying that Abeylegesse of Turkey is the "slowest" in the women's 10,000m, because her sprint in the final 200m is not up to that of Dibaba's. Though she proved that wrong...

But I honestly don't know...


Ross Tucker and Jonathan Dugas said...

Hi Mike

Indeed, Olympians, world record holders, everyone - they all "pace themselves". You might not perceive it to be pacing, because it's not a conscious strategy, but this is not the pacing we are referring to. What we are saying, according to the latest literature of fatigue and pacing in the exercise physiology, is that pacing is part conscious, part unconsious (or sub-conscious).

So in a sprint event - 200m or 400m, the athlete cannot afford a "maximal" effort from the start, because the body's energy stores cannot sustain muscle contraction for 200m if the athlete is running at 100m intensity. If the muscle is depleted of ATP, then the muscle would fail to relax, and rigor would develop. That doesn't happen, which is one indication that the recruitment of muscle is reduced for the 200m - that is pacing.

There are studies on this, by the way. Ferro et al (in 1999) looked at pacing in 100m, 200m and 400m events from the World Champs. That race was won by Maurice Greene, and he, like the other 7 men, and 8 women, all slow down in the 200m event, and never hit "maximal" speeds. Both these are characteristics of pacing, so it's very clear that sprint events are "paced", though not in the same way as a 10km race would be.


Ross Tucker and Jonathan Dugas said...

HI Oliver

You're quite right - as I said in the post, the analysis assumes that records are paced optimally, which is of course not necessarily true.

However, what I find most interesting is that even the Olympic finals have this same problem of a slower second lap, when everything would predict that the second lap should be super fast. Wottle's run is amazing, but the fact that his second lap has never been repeated does suggest that something of relevance is responsible.

And most of all, the graph that shows how the record has fallen by 3 seconds entirely as a function of the first lap seems to indicate that the second lap is already at a limit. If it was possible to run the second lap faster, then surely someone would have done so, either in the Olympics or in another race?

I must point out that there are probably many 1:43 times (which are not world records now, but would have been in the 1970's) which are set with final laps of 51 seconds, so the whole analysis is somewhat restrained by the fact that I looked only at world records.

But yes, interesting food for thought, but I really don't think any 800m athlete will ever be able to run two laps in 50.5 seconds - the first must be 49, the second 52.


Anonymous said...

Ain't no pacin' yourself in a 200 metres, I can assure you.

And Dave Wottle's win in '72 is enough to bring tears to any track lover.


Check out the leg speed to get in position, and then the kick home.



orsevrim said...

Your entry was really good, but we saw an exceptional case today with Bungei having the first lap in 53.35 and the last one in 51.30, i.e. a negative split with about 2 seconds. A bit unlucky for you to encounter such a race that counter-examples the general fact that you have stated in your entry, but of course we all know that exceptions do not violate the general truth. Thank you for your remarquable efforts here, and greetings from Turkey.

Ross Tucker and Jonathan Dugas said...

HI Orsevrim

Thanks for the comments!

Yes, quite right - the day after my post, it's proved completely wrong! I was watching it and when the second lap time came up, I realised that my timing (of the post) could not have been worse! I guess that's how it goes...

I will say that this is the SIXTH time in Olympic history that the second lap has been faster than the first. There have been a total of 17 Olympic 800m finals, and so the overwhelming majority are positive splits!

So as you point out, we'll call Beijing 2008 an "exception" and stick to our guns!

Thanks so much for the visit and kind words. Good to have you along - we had another visit from Turkey yesterday. You guys had a good games, especially on the track where Abeylegesse was one of the few to actually race smart against the Ethiopians - gutsy runner, makes for a great race!


Ross Tucker and Jonathan Dugas said...

Hi to anonymous

And I can assure you, there is most definitely pacing in the 200m event. Just because you don't "feel" that there is pacing does not mean that you are still being "regulated", so I think the difference is the defintion of "pacing".

All you need to look at to appreciate this is the top speeds achieved in the races and you'll see right away that your top speed in a 200m is not as high as in a 100m race. Second thing to look at is whether the speed is sustained, and of course it's not (that is a statement backed up by a lot of data - the data doesn't lie). Those two things combined suggest that there is "pacing".

Where you are correct is that there is not CONSCIOUS pacing as you likely are thinking of. But then, as you'll see from response to Mike two above this one, that's not the "pacing" we're talking about here.

Thanks for the link to Wottle's run, great video!


Anonymous said...

Ross wrote:
"There are studies on this, by the way. Ferro et al (in 1999) looked at pacing in 100m, 200m and 400m events from the World Champs. That race was won by Maurice Greene, and he, like the other 7 men, and 8 women, all slow down in the 200m event, and never hit "maximal" speeds."

Never been one to shy from a PhD, so how do you know how much the 'curve' affects the slowing performance -- compared to a 100m? Was this a variable that Ferro accounted for?
Exactly why cannot a 200m be run like a 100m and all the slowing occur as energy systems, neural systems fail? Or are you saying the 100m races are 'paced' as well.
I understand your distinction from 'conscious' pacing.

Anonymous said...

First I have a question on the 800m splits. Is the first lap split for the leader or the winner? Second, has any data been collected or analysis done on the 800m runner types, i.e. 400/800 vs 800/1500, in the way they run their best races? I would expect the 400/800 types to be more likely to run a faster first lap, while the 800/1500 types try to run more even splits.

Anonymous said...

Hi Guys

I just thought I'd toss in my 2cents worth as a former 800m runner. The Olympic final went to script; the slower start was always going to have a faster finish (in lap times). This is quite normal in Championship racing over all distances, but is much more common in non-international championships (National, Provincial, etc). However, if you compare Bungei's winning time against his PB, you'll see that he was alot slower. There is definitely a point at which the positive split hypothesis is reversed, my guess would be when the athlete runs his first lap about 10% slower than his PB average pace. But even then, his 2nd lap will not make up the difference enough for him to set a new PB. It doesn't really matter how much slower your first lap is than PB average, the energy cost by the last bend will be too high to make up the deficit.
The same can be said about going out too fast - too quick and you'll "die" from the early effort.

Henry Su said...

Ross and Jonathan, greetings. I happened upon your site today while browsing for articles on Bolt's 9.58 WR. Excellent analysis and a great follow-on to the IAAF biomechanics study.
Now for my request -- I am wondering if you might post more on your 800 meter pacing analysis in connection with your coverage of the 2009 World Championships. I think your conclusion about needing to run a slower second lap is enlightening but I have to ask whether a different strategy is nevertheless possible, perhaps for someone who is world class at the 400m/800m, a la Alberto Juantorena. One doesn't see such specialists today, and I am wondering whether world-class speed over 400 meters would make a slower, more aerobic first lap (say, 56-57 seconds) followed by a second lap sprint a viable strategy. Thanks.