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.

Thursday, September 25, 2008

2008 Berlin Marathon

The Berlin Marathon: A world record for Gebrselassie?

Breaking news: The Record is gone: 2:03:59. Click here for our race report, including the splits and pacing strategy by Gebrselassie as he became the first man under 2:04!

This Sunday sees the first of the Autumn marathons, and it's a big one, with Haile Gebrselassie making a big effort on his own, 1-year old record in the Berlin time-trial...er, Marathon.

Whenever an athlete of Geb's caliber lines up, the talk of world records is inevitable, but this particular race is even more in focus than usual, because this is the event that cost us the opportunity of seeing Gebrselassie in the Olympic Marathon in August.

You'll recall that it was earlier this year that Geb announced that he would not run the Olympic Marathon because he did not want to "damage" his long term-health in the polluted Chinese capital - many were sceptical then, suggesting that the big pay-day of a Berlin World record was the real reason he didn't run. And now, 6 months later, that potential pay-day has arrived.

Geb will line up in a race that really does lack competition, at least in terms of what he's probably going to be asking of his pace-makers. The World Record, you'll recall, stands at 2:04:26, and averages 2:57/km. Given that the next fastest man in the race has a 2:07 PB, and that the half-marathon mark is likely to be reached in 62 minutes, it's hard to see how anyone will be within a kilometer of Gebrselassie by the time he approaches the Brandenburg Gate and the finish line.

Looking back to look forward: Last year and this year's predictions

The talk is all about the record - the race is secondary (or non-existent, unless we have a miracle performance by someone, or a disaster run by Geb). Can he go sub-2:04? One year ago, as tends to happen when world records are broken, people waxed lyrical about the first sub-2 hour clocking. Never mind that it was "only" 4 and a half minutes away. The more "conservative" types were talking of a 2:02 within the next few years, and Gebrselassie himself said 2:03 was on.

And that's become the magic number for this year's race - 2:03-something. In order to become the first man to break 2:04, Gebrselassie will have to run just over 0.5 seconds per kilometer faster than he did last year. Sounds pretty straightforward - after all, 0.5 seconds represents about 3 m for every 1000m. When you look at it that way (a 0.3% increase), the sub 2:04 seems a reasonable guess. But as you'll know, it's far easier said than done.

Have a look at the splits from last year's race, which we analysed and reported in this post.

The race was a textbook example of how to pace a marathon race - the early pace was a little bit quick, but settled up to halfway, which was reached in 62:29. The pace then picked up in the second half, which was run in an incredible 61:57.

Between that race and Sunday's, Gebrselassie ran one other marathon, in Dubai. There, his pacing was not quite so assured - the first half was run in a suicidally quick 61:27. He died somewhat in the second half, coming home in 63:26 for a 2:04:53.

That Dubai race is an illustration of just how fragile the pacing strategy for the marathon can be...then again, the Berlin race last year demonstrates the same thing. Note how Gebrselassie never ran any 5km segment slower than 2:59, but that his fastest split, in the final 2.2km, was run at 2:53/km. The consistency in pace is remarkable, and it illustrates just how close to the limit Gebrselassie was.

For that reason, talk of a sub-2:04 may be a little optimistic, unless his training has been much better for 2008 than it was last time around. Given the "distraction" of the Olympic 10,000m, I find that unlikely.

To run a sub 2:04 would require a half-way split of around 62:00. At the slowest, I'd say 62:15, because to run a second half in sub-62 minutes off that pace would be a very difficult task, if not impossible. But an over-ambitious start, say 61:50 for the first half, may also see record chances disappear. The margin for error is thus so small that the pacemakers have an enormous responsibility - 15 seconds either way, over 21km, is their margin for error.

The other pivotal factor is how long they'll stay with Gebrselassie. Needless to say, running a 62 minute half marathon is no mean feat, and ideally, the pacemakers would want to continue onto the 30km mark (at least). Last year, Geb was alone from 30km onwards. He'll need at least the same performance this time around, at a faster pace. He'll also need the weather to be perfect - so far, it looks pretty good, with very little wind predicted, and temperatures between 13 and 16 degrees Celsuis (55 and 65F). The weather may yet become a factor...or an excuse...

Gebrselassie's last shot? And our prediction...2:04:40

The other story is that Gebrselassie may not have too many more shots at the record - he is running great times over 10,000m this year, an encouraging sign, but at the age of 35, it's hard to see that he has more than 4 more great races in him. Then there's the emergence of some real challengers - Sammy Wanjiru ran 2:06:32 in Beijing's heat and humidity, in a tactical race with a brutal early pace. Incredible performance, and maybe Geb is lucky he didn't try to run that race - he would likely not have won anyway. And there's Martin Lel, beaten in Beijing, but invincible before. He'll be back and I still favour him to run the world record close.

Regardless of who it is, I feel that this time of 2:04:26 will be beaten within the next 18 months, and so Gebrselassie will want to "pad" it with some reserve.

My prediction:

Geb will get the pacing slightly wrong and go out a little bit too fast (around 61:45). He'll pay for that in the second half, and fall just short of the world record - 2:04:40.

Then again, I might be completely wrong! Feel free to fire away with your own predictions - we'd offer a prize, but for now, you'll have the satisfaction of winning the "Nostradamus" award for closest to the actual time...

Either way, we'll do a "real-time" analysis of the race on Sunday morning, so the splits and race report will be up almost as soon as it's over!

Join us then!


Andrew said...

How much did Geb actually train for the Olympic 10K? Could that have been just a "training race" for him?

2:04:20 is my wild guess...

I figure that Dubai was a test to see how fast he could push it. I bet he'll attempt your "over-ambitious start, say 61:50" on the theory that it's slower than his Dubai start and that his training will enable a sub-62:00 finish; however, he'll fade more than he wanted at 62:30. Still a WR, but not under the magic 2:04.

And yes, Wanjiru or Lel will be the first under 2:04, perhaps in 2009. Those guys amazed me in the Olympics to run off the front like that!

Mark Boen said...

I disagree that the Olympic 10,000m race was a distraction. Instead, I think it further increased his turnover rate for this race, which is probably exactly why he skipped the marathon. That and a big fat check on Sunday.

I predict 1:03.58 if the whether and conditions are near optimal.

PecosPeet said...

Berlin weather looks great - low Sat night of 8 climbing to a high of 17 with sunny skies.

I'll assume the temps at the end of the race will be about 13/14 and predict a 2:04:02. Sorry Geb, but not quite -- this year.

Ross Tucker and Jonathan Dugas said...

Hi everyone!

Seems I'm the "pessimist" - three predictions, three world records, one sub-2:04...time will tell!

The impact of his Olympic race will be interesting. It's possible that he didn't adjust his training at all - he was in seriously good shape for the 10k earlier this year when he broke 27min in Holland (in May, I think), which is a sign that he'd have done some specific work. I imagine he'd have done a similar thing leading up to Beijing, although there were reports that he was suffering from an Achilles injury that hampered his speed work. So maybe his training was unaffected...

And maybe the 10km focus will have helped his marathon - we do know that the best predictor of marathon performance is 10km performance, so in theory, improving his 10km (relative to last year's) would have a positive effect on his marathon. However, I'm still taking the pessimistic view that any lack of volume will have an effect on his ability to sustain what is a very fast pace in the last 7km of the race.

I recall that Geb has done marathon attempts on a number of occasions - there was the successful attempt last year, then Dubai, plus Berlin in 2006, and Amsterdam, and another in Japan, if memory serves me, and the attempt was unsuccesful because on each occasion, he just couldn't sustain the pace from 30km. Until Berlin last year, that is, when he pulled it off.

So on the balance of history, I guess you could say he's learned. Or (being the pessimistic), I could say he's got 1 out of 5 so far, Sunday will continue the trend!

We'll find out on Sunday!


Anonymous said...

Oh, oh, i want to predict! I predict that he'll get the WR but "only" in 2:04:10. He's running so close to the edge, and he wants this record soooo bad, enough to skip the olympics, that he'll go out to hard. I think he'll come through half in 61:40, and run a second half of 62:30

Anonymous said...

A Marathoner has a limited number of great personal performances. I cannot recall where I read it from, but Alberto Salazar and Dick Beardsley were mentioned as examples. Nowadays athletes taper and recuperate more than the past but still the Marathon takes its toll, especially when running at such incredible speeds.

I would certainly love to see Geb break the record and flash a smile at the finish since he is a favorite athlete of mine. But this is very difficult. Even cutting 42 seconds off his previous WR, requires running at 2:56 per km, which is extremely difficult as you wrote.

After all the Marathon is very hard to predict: You train, you run kilometer after kilometer and you reach your peak. But in the end you just hope that all goes according to plan and that your hard work pays off, there is no warranty. The Marathon is a very risky investment for all athletes, even if they are at least an hour slower than Geb like myself.


Ron George said...

I'm watching it live right now as we speak. Heck, you were dead on right. They're going way too quick for the first 3-4 miles. Incredible pacing. Lets see how it unfolds.

Unknown said...

Hi Mark you had it almost spot-on. Final time 2h57m I think.

Mark Boen said...

I was off by one freakin' second. Where's the love everyone? I knew he could do it. That man is a pacing machine. Is Samuel Wanjiru going to be the first man under 2:03? Looks like he may have 15 years yet to show us his potential.

Anyway, you have to say that the Olympic 10K helped versus hindered. It was always part of the master plan.

Ross Tucker and Jonathan Dugas said...

Hi Mark

Yes, indeed, much respect, one second is not bad...better luck next year with that accuracy!!!

No, good estimate! It was impressive. I don't think we must get too carried away with Wanjiru's chances - I think a mid 2:03 is on the cards, but Wanjiru is a racer, so the big problem for him (and Martin Lel, who everyone seems to have forgotten about) is when they run this time - Berlin is clearly the best race to do it in, but Geb won't have those guys in the race with him next year. They run London anyway, and that's defintely not as quick as Berlin, I don't think. Especially because it become tactical...

So when does Wanjiru take his shot? I hope soon, because I think he could run 2:03:40, thereabouts. Going much faster than this, well, that will be something. I suspect I'll be an old man by the time we see a sub 2:03! But then I was wrong two days ago!


Anonymous said...

Sorry, but Mark was off by 1 hour and 1 second... Mark, you were way too optimistic.

Take the love back.

Mark Boen said...

I see that Ray. I thought I wrote 2:03.58 and that's what I meant. But I do the same thing with gas prices. I tell my wife I bought gas at 1:45 and she corrects me and tells no, you mean $3.45 (in Oregon USA).