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.

Monday, August 27, 2012

The Armstrong fallout: Thoughts and theories

The Lance Armstrong fallout - questions, denials and doping reactions

Friday last week saw Lance Armstrong release a statement that effectively ended his fight against the USADA doping charges, and accept the stripping of his seven Tour de France titles.  It was a significant day for the sport, if only because it forces a look back to the era of cycling that was so tainted by drugs that between 1996 and 2006, the sport has not had a single champion untainted by doping.  The timeline reads: Riis, Ullrich, Pantani, Armstrong, Landis, and here we sit, seven years later with a big asterisk next to the Tour!

The reaction to the USADA case, and Armstrong's statement, has however produced huge debate.  I've refrained from comment here, but have been discussing it at length over on Facebook and Twitter for those interested in the day-to-day thoughts that come up.

But it's time to address a few of the common questions and positions, hence this post.

Read more. . .

Sunday, August 12, 2012

London 2012: Live Men's marathon analysis

London 2012: Men's marathon live analysis and splits

Gold for Uganda, as Kiprotich steals the show

Stephen Kiprotich of Uganda is your 2012 Olympic marathon champion.  In a stunning result, the unheralded 23-year old from Uganda surprised the fancied Kenyans to win the Olympic track and field programme's final medal.

His reward is a gold medal, to be handed out at the closing ceremony.  You'd be forgiven for not knowing much about the new Olympic champion, because his credentials coming in said nothing of what he was about to do.

Kiprotich, at only 23, is relatively new to the marathon, but his last three performances told little of his potential.  He has a 2:07:20 PB in the marathon, run in Enschede in 2011.  He ran a 2:07:50 in Tokyo this year, and was 13th in the IAAF World Championships marathon last year (2:12:57).

His track credentials are nothing special (27:58), but that is likely because he's never run seriously for track times.  His half marathon PB is "only" 62:52, also nothing to scare the sub-60 min Kenyans into thinking he would be the man to beat them today.

That he was, however, and he pulled off a huge surprise to win Uganda's first gold.

An intriguing race

The race was intriguing throughout, set up by early Kenyan aggression, but it became a race of attrition.

Consider for example that the halfway split was 1:03:15 for early leader Kipsang, with the chase group (including Kiprotich) 16 seconds back.  Their split was therefore 1:03:31, and the winning time was 2:08:01.  That means that Kiprotich ran a 1:04:30 second half, a minute slower than the first and he was closest to running an even pace of the early leaders (I'll check later to see if a strong finish was closer, but I doubt it.  Keflezighi, for example, was a very strong finisher, but he went through halfway in 1:04:30 and finished in 2:11:06 (1:06:36))

The rest collapsed under the pressure, and in the conditions. Abel Kirui, a proven championship runner, was second, 26 seconds behind, and Kipsang finished third, a full 1:36 down.  Other big favourites were totally blown away.  The Ethiopian challenge is perhaps best exemplified by Abshero, who was in the chase group until after halfway, but then got dropped and went backwards at an incredible rate.  Having gone through 25km in 1:15:05 with the chase group, he went through 35km in 1:49:22, which included a 5km segment of 19:03, before he stepped off the road.  The other Ethiopians also didn't not finish, with Sifar actually losing contact as early as 10km before he stopped.

A terrible day for Ethiopia, then, and a disappointing day for Kenya, despite getting two men onto the podium.  After what has been a disappointing Games for them, they'd have been hoping that their most dominant event, the Marathon, would at least provide some desired gold, to defend the title the late Sammy Wanjiru won in Beijing.

It was not to be, though they raced much more aggressively than we've seen so far in the distance races in Beijing.

The race as it unfolded

There were echoes of Wanjiru in the race today, because Kipsang took the race lead as early as 10km, with an aggressive 14;12 split from 10km to 15km, and opened a lead that got up to 16 seconds.

The chasers eventually reeled him in, but not before huge damage had been done to most of the field. By the time the race re-formed at the front, around the 27km, there were only three men left - Kipsang, Kirui and Kiprotich. The Ethiopian challenge had been seen off, as first Sifer, then Fekele and then Abshero were blown away.

The three were left to sort out the medals, and there was a moment, around 35k, where it seemed that the Kenyans had broken the resistance of their East African neighbour.  Kiprotich dropped off, not substantially, just by about 10m, but it seemed that it was the first sign of his impending slow down, because it co-incided with the slowest interval of the race (15:48).  However, he was well in control, caught up at 37km, and then went straight past.

At that moment, the Kenyan response simply did not come.  The pre-race favorites, having dealt with the Ethiopian threat, now found themselves trailing a man who they must surely have discounted as a serious challenger.

But Kiprotich was the man on the day, and his final 5km were unchallenged.  He is not the first man from Uganda to win Olympic Gold, but this is surely their greatest triumph.

The entire race, as it unfolded, is in the post below, where I made real-time comments.  Feel free to relive the final event of the London Games' athletic programme

Live splits

Welcome to the London Olympic marathon.  The race is underway, and over the next two hours, I'll post the 5km interval times and some thoughts on the race as it unfolds.

5km - 15:23, projecting 2:09:49
10km - 30:46, so another 15:23 split for the last 5km
15km - 44:58, the last 5km in 14:12. The chase pack is 13 secs back (45:11)
20km - 59:57. Last 5km in 14:59.  Chase pack is 14 seconds behind
Halfway - 1:03:15, with 16 seconds to the chase (1:03:31)
25km - 1:14:58, a 15:01 for the last 5km, and the gap is now 7 seconds
30km - 1:30:15, all together. Kipsang did 15:17, the rest 15:10 for last 5km
35km - 1:46:03. Last 5km in 15:48, the slowest of the race.
40km - 2:01:12. Last 5km in 15:09, and Kiprotich makes the gold-winning move
Finish - 2:08:01.  Last 2.2km in 6:49 (3:06/km)


Uganda have gold in the men's marathon, courtesy 23-year old Stephen Kiprotich.  The Ugandan made the race's decisive move at 37km, and ran the final 5km on his own, with the favored Kenyans in his wake.  His winning time was 2:08:01, quite a lot slower than the halfway split suggested, but it was a race of attrition and Kiprotich was its best survivor!

His final 2.2km were not spectacular (6:49, 3:06/km), but they didn't need to be - he had a substantial lead and managed to increase it to 26 seconds on the line.

Second went to Abel Kirui, proven championship runner, in 2:08:27, with bronze to early aggressor Kipsang in 2:09:37.

Incredible result, huge surprise, and Kenya finish the Games with only two golds, having expected so many more.


Stephen Kiprotich is now 19 seconds ahead of Kirui, with Kipsang a further 32 seconds back.  With only 2km to go, that's the gold medal for Uganda, barring an absolute disaster.

His last 5km were run in 15:09, a significant jump in pace from the preceding interval, run in 15:48.  Kiprotich looks good enough to extend that lead, let alone defend it, and Uganda are on their way to their most celebrated Olympic medal ever.


Stephen Kiprotich now was a lead that looks to be decisive, it's around 100m, and he shows no signs of coming back.  Amazing story.  Here's some information on the Ugandan:

He has a 2:07:20 PB in the marathon, run in Enschede in 2011.  He ran a 2:07:50 in Tokyo this year, and was 13th in the IAAF World Championships marathon last year (2:12:57).

His track credentials are nothing special (27:58), but that is likely because he's never run seriously for track times.  His half marathon PB is "only" 62:52, also nothing to scare the sub-60 min Kenyans into thinking he would be the man to beat them today.

He is only 23 years old, so new to the marathon, but this is just a remarkable run, one which has certainly taken me by surprise.  This will be one of the big surprises of the Games, if not Olympic marathon in recent memory, though this is a race that does tend to produce surprises.


Kiprotich has come back.  Incredible!  Having dropped off by 10 m at 35km, he has come back and made all those early projections look foolish.  He moved past the Kenyans at the 37km mark and soon opened a big, big lead.

Kirui is trying to give chase, but the response is slow.  If Kiprotich can hold this, he'll win gold for Uganda, and what an amazing story that will be.  Did anyone pick this?


As I typed that, it was Kipsang who moved clear again.  It's not a surge as much as a subtle increase in pace, but it was enough to gap Kirui, who is now running 5m back.  The elastic has not broken, but was certainly stretched.  They've now come back together.


The last five kilometers have been covered in 15:48, the slowest of the race by a considerable amount.  Yet despite that, the pace seems to have dropped Kiprotich of Uganda, and so Kirui and Kipsang, who have been talking to one another for most of the last five kilometers, seem to have sorted out gold and silver for Kenya.

Unless Kiprotich can claw his way back, that's how the final 7km will play out.  The pace is not likely to stay at that 15:48 level, it should get faster, so it will be a difficult ask.  Kipsang looks very fluid and may be the favourite now.


The "catch" came at about 27km, and Kipsang's "lonely vigil" suddenly became a threesome, as he was joined by Stephen Kiprotich and Abel Kirui.  The pace then held firm, which is not a surprise.  It's not as though Kipsang was cracking to allow the chase to catch him - his pace has held constant since the 15km mark, and they continue to run around 15km per 5km.

The medals will certainly come from these three, Abshero has continued to drop back, and is 46 seconds behind.  dos Santos of Brazil is another five seconds back and will probably take fourth place soon.  If anything happens to the front three, he's the likeliest other medal winner, but the three in front seem to have this race controlled and will probably fight out gold, silver and bronze.

If people think Kipsang has made his move in this race, they're wrong - he'll be a factor in the final 5km, I'm sure of it, because what he did is not much different to anyone else, and so he's in a good position right now.

Kirui looked spectacular in London when he surged, and then he exploded and finished terribly.  His turn of speed is perhaps the big danger for gold now.  I'd make him the favorite of the three.

The last 5km, from 25km to 30km, were run in 15:17 (for Kipsang, the chase is a little faster - 15:10).

The projected time, meanwhile, is 2:06:56, which means the Olympic record may still be on, but is slipping away right now.  If there are surges, they'll go quicker.


Abshero is gone!  So Ethiopia has lost its final medal contender before the 30km mark, and that is a surprise.  The chase is now Kirui and Kiprotich, with Kipsang ahead.  As expected, it's East Africa to the fore, but possibly, without Ethiopia.

Abshero's chance here relies on the pace dropping once Kipsang is caught, which looks like within the next few minutes.  A lull in the pace may allow him back, as we saw in the women's race where Arkhipova looked off the back a few times, but came back and fought for bronze.


Kipsang is still holding the lead, but the urgency in the chase group has begun to erode it.  It's now 7 seconds, and the last 5km were run in 15:01 by Kipsang, which means the chase group have run a 14:54.

What the chase has done is fragment, and it's now down to only three.  They are led by Stephen Kiprotich, who has been largely responsible for the increase in pace from behind.  The others are Abshero, the lone Ethiopian, and Abel Kirui.


The chase group is now starting to split too.  Stephen Kiprotich took the group through Leadenhall market very aggressively, and the compact group was suddenly stretched.  This is the pressure of Kipsang now filtering its way back to the chase, who have presumably recognized that they need to respond to bring that gap of 16 seconds down.

If anything, the gap is growing.  Perhaps Kipsang has done what he did in London earlier this year, surging at the half marathon mark.  The next split will be interesting.


Halfway has been reached in 1:03:15, with the chase group at 16 seconds down.  It's now a chase group of six, as Eritrea's Asmerom has also fallen off the pace.

The chase is now made of Feleke and Abshero of Ethiopia, Kirui of Kenya, Kiprotich of Uganda, dos Santos of Brazil and Mokoka of South Africa.  Although as I write that, dos Santos has begun to drop off the group and so it's down to five.

At the front, Kipsang stops to go back for his energy drink, which means he lost a second or two at the 22km water point.


Kipsang has relented only slightly, running the last five kilometers in 14:59.

That has helped him defend his lead, which now stands at 14 seconds.  That chase group is now down to seven, with Emmanuel Mutai dropping off the back.  He was the reserve added to the team after the withdrawal of Moses Mosop.  Many felt that Geoffrey Mutai would have been a better pick. Hindsight is easy of course, but I suspect those calls now seem a little insightful.

Meanwhile, the stretch from 17 to 18km claimed two of the three Americans in the race.  First Ryan Hall stopped, holding his right hamstring, it seemed.  Then Abdirahman stopped, so only Keflezighi is left now.


Wilson Kipsang, winner of the 2012 London marathon, has opened a lead of 13 seconds over a chase group of 8.  All the main protagonists are there with the exception of Dino Sefir, who fell out of the group at about 10km.  The chase thus comprises two Ethiopians (Abshero & Feleke), two Kenyans (Mutai and Kirui), an Eritrean (Asmerom), a South African (Mokoka), a Ugandan (Kiprotich) and a Brazilian (dos Santos).

The last 5km were run in 14:12, which explains that huge explosion in the race from 10km onwards.  The time now projects a 2:06:29, but don't expect the next 10km to be run at the same pace, of course.  The Olympic record of Wanjiru is probably still on, however.

Fascinating battle here, because the pressure has been applied by the race favourite.

The pressure has put paid to everyone else - Ryan Hall and Keflezighi are off the back, and they'll hope to run their own race for something in the range of 2:08, and hope that the early pace claims some victims.


Kipsang has actually created a small gap, and we are only at 12km.  Amazing early aggression.  It was expected that Kenya would try to assert control on the race, but to do it this early is very surprising.  They took the lead shortly after 10km, once De Almeida had been caught, and the field split almost instantly.  Sefir and Ryan Hall were the first casualties, but now many have been dropped.

The front of the race has been trimmed to just over a handful, but Kipsang has pulled Abshero clear and there are gaps.  Kipsang is 10m clear of Abshero who is 5m clear of the chase group of maybe six.

Abshero has now fallen back into the pack, and there are now 7 men chasing the leader Kipsang, whose lead has grown to perhaps 7 or 8 seconds.


In the battle of Ethiopia and Kenya, it's first blood Kenya, and that's because Ethiopia's Dino Sefir is off the back of the group.  And so is Ryan Hall.  That's very surprising, because we're only at 11km, and the pace is "only" 2:09:49, yet two of the pre-race favourites are already off the back.

At the front, Kenya have again assumed the lead, with Kipsang pressing the pace.  It certainly seems to have gotten faster, because the front group has been cut to about 12 men already.


There is a breakaway leader, about 50m off the front of the main field.  It's Brazil's De Almeida, but the time for the main group is 30: . That's a 15:23 for the last 5km, identical to the first 5km.  So even paced as can be.

Remember back to the women's race - the pace was very steady for the first half, and then when it picked up, it didn't reach the same kinds of speeds we are used to seeing in the city marathons.  That's a function of the tight turns on the course, and the change in surfaces, which let some of the athletes to call it the hardest race they'd ever run.  That will be a significant factor in the second half of this race.

Right now, it's all rather sedate.


The 5km mark is reached in 15:23, and the Africans are already showing at the front.  Maybe the Kenyans are borrowing from Beijing, where Wanjiru's aggression perhaps changed the way marathons are run.  They're at the front, but the group is big, because the pace is respectable, but not super fast.  Right now, it's projecting a 2:09:49.

Start and preview

The race has been billed as a clash between Kenya and Ethiopia.  2011 was of course Kenya's year - all  20 of the top times were run by Kenyans in an unparalleled show of dominance by one nation.  That included the winners of every single major city marathon, and the world record.

2012 has been more evenly matched - Ethiopia stole the Dubai marathon with a host of fast times, and then also won Rotterdam, and set this race up beautifully.  More on the protagonists as the race unfolds, but it seems that most are forecasting this as a Kenya vs Ethiopia showdown, with Abshero and Kipsang their likeliest champions.  The Americans, in the form of Hall and Keflezighi in particular, may disrupt the battle in the same what that Arkhiopova did in the women's race.

My initial thought is that Ethiopia should be favoured, because their athletes last raced in late January in Dubai, compared to the Kenyans who raced in April in London's city marathon.

For example, Ayele Absehero and Dino Sefer of Ethiopia have had 198 days since their last marathon, whereas Wilson Kipsang has had 112 days.  Those 86 days matter in a race this competitive.  (Thanks, by the way, to Wayne Do Rego for the numbers - I'll do a proper post with his analysis after the race!)

This, plus the fact that it has been a very poor Games for Kenya lead me to think that Ethiopia hold the cards here.  Kenya's athletes have, to me, seemed over-done and tactically poor.  They've been run out of medals in the women's 800m and men's 1500m and 10,000m races, and have seen their big favourites settle for minor medals.  Turning that around will be difficult.

We are coming up to five kilometers, however, in a big group, so let's get the projections going!


Saturday, August 11, 2012

Women's 800m: Analysing Semenya & other insights

London 2012: Women's 800m perplexity, analyzing Semenya's race 

On Thursday night, David Rudisha led home the greatest 800m race we've ever seen - he pulled the field to a world record for every single finishing position, 7 personal bests and three national records.

Tonight, Mariya Savinova led home the women's 800m final, but it leaned more towards the curious and peculiar than the spectacular.  That is primarily because of the manner with which Caster Semenya, South Africa's flag bearer, ran to win the silver medal.  Savinova was, as usually, tactically superb, fast and timed her effort perfectly.  She won in 1:56.19 to add to last year's World title.  The real story, at least for me and all those discussing it on Twitter, was Semenya, and so let's talk about that a little.

If you saw the race, you'll know what I'm referring to - she dropped into 8th place by 300m, and stayed there for the next 300m.  At the bell, she was 1.38s behind the leader.  Down the back straight with 280m to go, when Pamela Jelimo made the race's first move, Semenya was perhaps 12m back, in last place, and not even close to responding, as you can see in the screenshot below.

With 200m to run, Semenya had moved into 7th, picking up a tiring Niyonsaba, but was still well off Jelimo, a pre-race favourite.  Meanwhile, Savinova had by now begun to make her move too.  This was the move that Semenya must have known would determine gold and silver, and in her semi-final, she'd shown the ability to respond to those tactics.  Tonight, in the final, she was distant from the action.

With 140m to go, Savinova was making the race's decisive move, but still Semenya had not responded - she was by now up to 6th place, however, picking off the fading Jepkosgei.  I kept waiting for a move, because she'd shown in her semi that she was not tactically unaware, but it just never came.

Savinova would go on to open a commanding lead, and with 50m to go, the race was over.  Only Poistogova and Jelimo went with her coming off the final bend, while Semenya was still in 6th.

Jelimo's legs imploded around 60m from the line and she went backwards.  By now, finally going forward was Semenya who would move incredibly rapidly through the field and close down everyone in front of her with the exception of Savinova.  Semenya ended with a season's best of 1:57.23, marginally faster than the 1:57.67 she ran to win her semi-final, but it was a race run in a totally different manner.

This led Sports Illustrated's Tim Layden to tweet the following immediately after the race - Semenya was "disengaged".  He's not accusing her of anything, but it's not difficult to see where the next step lies, and that's exactly what has happened since the race.

History repeating itself with Semenya - a common allegation

Unfortunately, this kind of speculation is becoming all too familiar for Caster Semenya.  Last year in Daegu, the race strategy was different, but the result was identical (Savinova-Semenya), and the speculation after the race was the same.  There, Semenya was attentive and ran near the front, before moving into the lead with 180m to go.  Savinova followed, but Semenya looked strong enough to win until the final 30m, where she suddenly slowed and Savinova swept by to win.  The forums were soon buzzing with allegations that Semenya had lost on purpose.

This year, the same has happened basically every single time Semenya has run in the European meetings.  At Diamond League meets, she was often seen languishing at the back, looking "disinterested" but running solid 1:59 to 2:00 times while her major rivals - Jelimo and Fantu Magiso in particular - were running 1:56 to 1:57.  People were accusing Semenya of running slowly on purpose, so that she avoids too much scrutiny, that she is 'scared' to win because of the intense allegation it may bring.

Semenya - evaluated differently because of her past

You see, Semenya is not "judged normally" in athletic circles, and that has everything to do with the sex verification controversy involving her after she exploded onto the world scene in Berlin.  Since being questioned, she spent nine months away from the track, before returning amid much secrecy and with slower times than before.  The speculation bandwagon kicked off, and when she was winning, she was accused of cheating, when she was losing, she was accused of not trying.

She was, and remains, in an impossible situation, because every result and every move is looked at through a filter.  It is a filter that colors her performances according to male vs female, cheating vs throwing it on purpose, and when she produces racing performances like tonight, that filter is rather vivid.

The prevailing "allegation", ever since her return in 2010, is that she is running slowly to stay under the radar, avoiding winning and the questions this would undoubtedly bring.  If that's the plan, then it sure isn't working, because what we saw today (and in Daegu) draws more allegation than a "typical race", in my opinion.  But more on that shortly.

Possibility # 1: Semenya may simply not have the speed

The current speculation (and before accusing people of ignorance and stoking the fires of controversy, just have a look at the forums and Twitter to see the reaction to Semenya's race) is thus fueled by Semenya's history.  Within the ten minutes of the race finishing, I got 34 tweets asking whether she'd "thrown it", or "tried to avoid winning gold".  One person demanded a full investigation into why she was jogging. Another said that he'd never seen someone look so "aerobic" at the finish of an 800m race.

Of course, this may all be a totally misplaced accusation.  Maybe Semenya just didn't have the physiological capacity to run the race tactics people are accustomed to seeing.  Maybe she was just not good enough to go with that early pace, and to respond to those surgest.  Perhaps there is nothing to her performance other than that she runs a more even pace than her rivals.

A comparison between her semi-final and this race is interesting in this regard.  In that semi, she went through 400m in just over 58 seconds, 600m in about 1:28 and then closed the final 200m in 29.5s, looking like she had something in reserve.

Tonight, she went through 400m in 57.69s, then through 600m in about 1:27.1, and then closed in a touch over 30 seconds.  My point is, her performance in the final was slightly faster at every stage than the semi, until she closed slower over the final 200m.  To finish SLOWER than she did in the semi implies that she has little reserve and that she is closer to the limit than she looks.  She wasn't actually that fast over the final 200m, it's just that everyone else was very slow!

It's possible that she doesn't have the speed (or psychological capacity and confidence) to be able to run a 56-second first lap, or a 28 second 200m split, regardless of when in the race it happens.  If you look at Semenya, her running style is very laboured - the commentator described her as "lumbering" and that's about right.  She lacks a knee lift, and her heel-flick is also very limited, so it is possible that she lacks the ability to change pace much, and so I have to put forward the possibility that she may not actually have the capacity to respond to surges, and maybe a 28-29-29-30 race breakdown is as fast as Semenya can go.

The rest of the race, incidentally, went 27s to 200m, then 29s to 400 (56.3 at the bell), and then 29.2s for the next 200m, and closed, for the most part, in 32s.  So, you have Semenya with a 28-29-29-30 (57.69s & 59.54s), running against everyone else with a 27-29-29-32 (Jelimo, for example, was 56.66s & 60.93s).  In this regard, Semenya actually didn't finish the race fast, as much as everyone else finished it really slowly.  The one exception of course was Savinova, who closed the final 200m in just under 30 seconds (57.29s and 58.90s halves).

The rest, Jelimo in particular, were terribly slow over the final 200m.

Not that I'm trying to say that Semenya ran a good race - you simply cannot allow the moves of your two main rivals to go completely unnoticed, but I am saying that it's possible that Semenya does not have the ability to run the race any other way - she may well be at her limit and unable to run those 28s 200m splits mid-race.  The fact that she looks so easy doing it is neither here nor there.  Go on YouTube and look up her race at the World Junior Championships in Poland in 2008.  She finished second last in her semi-final, and looked the same as she did today.  That was long before any controversy, or any need to avoid scrutiny.  Semenya is just a very 'casual', disengaged runner.

The other speculation - let the guesswork begin

That said, there is still much to be debated about the case.  Once you have dealt with that possibility that her apparent "throwing it" and "sandbagging" tactic may just be that she can't match the speed of the first 600m in the race, then you get on to dealing with the other speculation.

I'm going to simplify my answer as much as I can, and then try to go into detail to explain some thoughts and insights.  The simple answer is "I don't know what happened.  Your guess is as good as mine.  And I understand the questions, but there are no answers, we just do not know".

Right, now, having dealt with that, let's discuss the current discussion!  For this, a brief history lesson on her case, which most of you will know, so jump ahead a section.  If you're new, read on.

The history and secrecy fuels speculation

Since day one, Caster Semenya has presented an insoluble problem for the sport.  The biggest problem, aside from the resolving the obvious debate about her performance, is the secrecy which has surrounded her story since the case first broke.  I suspect there is no satisfactory answer to this story, at any level.  Even going back to 2009, when debate first began, it was impossible to say what should be done.  Did she have an intersex condition?  That part would be easy to find out - the science and biology is not that complicated.  In 2010, I wrote a scientific review paper on the subject with geneticist Prof Malcolm Collins, summarizing the history, the physiology and performance implications of sex verification in sport, for those interested.

But, does the condition provide an advantage?  And if it does, should that be the basis for excluding her from competition - it's a natural advantage, after all?  That's a whole lot trickier, and it's an ethical, moral and social debate for which I think there is no consensus.  Each will have their own opinion.

However, the debate still exists, and rather than allowing radical speculation, I hope it is helpful to consider the story in a thoughtful manner, hence my thoughts below.  

A point on context - being in South Africa, I'm exposed to more news and speculation about Semenya than perhaps most, and so my views are kind of informed by years of conversation with people, reports, information from people connected to the case etc.  But I want to stress upfront that just like the rest of the world, with maybe a few exceptions, we are all guessing here.

If you were to right the summary version of this history, it would go as follows:  Q:  "What happened to Semenya in 2009/2010 to allow her to compete?"   A: "We don't know".  Next question: "How do we explain the huge variability in Semenya's performances in 2011 and 2012, where she goes from the back of the field in a Diamond league event and struggling to break 1:59 to being utterly dominant in the major championships?"  Answer: "We don't know.

The short version is that we just don't know anything about anything, and so we speculate as much as possible, maybe in an informed way, weighing the possibilities, but very few people know the truth, and they are not talking.  Should we speculate at all?  Probably not.  We should, in theory, "trust" the IAAF, who were involved in the process from Day 1, and say that if they have cleared her to run, then we should just accept that.  And officially, that would be the correct position to take.

Significant improvements in a short time ask the questions

However, the reality is that just as we SHOULD question performances that we regard as suspect, I think it's naive and 'deliberately ignorant' to ignore the questions that arise from Semenya's case.  Here, it is her performance that asks the questions, not the history of her case.  That history tells us that the IAAF worked with Semenya, cleared her, and she should be treated as any other athlete.  The case is closed, it was resolved and is in the past.  The problem is that the performances re-open that door, and because nothing is known, it leads to speculation and accusation.  The root cause is the secrecy around the case.

The first problem arises out of the sudden improvements Semenya makes at championships.  Or put differently, it's how well off the pace she is in European races, before she arrives to championships looking close to unbeatable (by all but Savinova, it turns out).  This year, Semenya had been "stuck" in the 1:59 to 2:01 range since April, and had run half a dozen races where she was unable to get faster.  Then suddenly, she runs 1:57 looking rather easy, and it is going to cause questions.  

Remember, this is exactly the same thing that was done for Ye Shiwen of China and for Makhloufi of Algeria - they improved significantly in a short time, it was deemed "peculiar" and the speculation of doping began.  Semenya's improvement is similar, if not larger in magnitude over a shorter period, and so the same logic leads to questions.  The difference is that once asked for Semenya, the question will not have us zoning in on doping as has happened for Ye Shiwen or Makhloufi, it will return to the gender controversy, and we will unfairly make accusations about gender, all over again.  Is it right?  No.  It is understandable?  Yes.

The secrecy - the root cause of speculation

And the reason it's going to happen is because of the failure in management of the message, not only by the IAAF, but by Semenya's camp.  To explain, the two key points, which I think are more important than the performance:

1)  The case should never have been leaked in the first place.  Obviously.  That was a mistake for which Semenya will "pay" for the rest of her career, and it has exposed her to the most invasive scrutiny I think anyone can imagine.  I think it is remarkable that she has continued to compete, and how she has stood up under that kind of pressure.  Most would not cope at all, let alone resume their athletic careers.  She's done that, and she was rightly given the honor of being our flag bearer, and the courage and character she shows to run at all is amazing.

2)  Having said this, once the story broke, and the athletics world knew there was a question, then in my opinion, it had to be followed through to its conclusion and made known what the outcome was.  And simply clearing her to compete many months later is not the same as saying that the matter was concluded.  People are notoriously mistrusting of sports governing bodies, and they're even more mistrusting of athletes.  There are too many dishonest athletes to believe what we see with no small dose of skepticism.  So, when Semenya resumed her career in 2010, I felt that it would be important for her to make some kind of announcement to say that the matter had been resolved, and how.  Perhaps this should have been done by the authorities.  But it should have been done by someone, to at least control the message.

But what happened instead was that a veil of secrecy fell over the story, and all of a sudden, nobody was saying anything.  The secrecy grew and grew, until she began running again.  But she was not dominating - having destroyed the best in the world in 2009, she was now 4 to 5 seconds slower, looking sluggish and losing races.  Her subsequent performances was gone up and down wildly and it has been absolutely impossible to predict what is coming next.

Everyone can see this unusual situation, they know that they are seeing 'abnormal' variations in performance, but nobody can say why.  And so they speculate.  The problem is that when you fail to tell people the truth, they tend to make up the truth.  And the made up truth is almost always worse than the reality.  And so now, we sit in a situation where people will either allege that:
  1. Nothing happened in the first case, and she is still a man (this is ignorant, because that clearly was never true to begin with - the biology of sex is far more complex than this), or 
  2. She got treated but it's not working, or
  3. She got treatment but is able to manipulate it to optimize her performance whenever she wants to - it slows her down in a predictable way, so she can use treatment as she pleases to find those improvements, or
  4. She is deliberately losing races to avoid suspicion, as is happening after the Olympic Games

To repeat, we simply do not know what transpired, and therefore we cannot know whether any of the above options is true.  If I were forced to give my thoughts, I'd say that option 1) is impossible - we know something happened.  Reading between the lines, based on the time it took, I'd fairly confidently speculate that she received medical treatment, and probably still is.  Thus, the next three options are possible.  I don't know what treatment might involve, or whether she can manipulate it.  I suspect that it would be possible, just as any doping is possible.  But I'd be surprised if it was this simple.  

I simply cannot see option 4) being true - why would you try to avoid detection by going from last to first?  The contrast in performance is just so enormous that people will notice it EVEN MORE!  If you are going to fly under the radar, then your approach would be to look as normal as possible.  Going from nowhere to dominating is not "normal", and so if they are deliberately slowing down to lose races, then it's a strategy that is not only bizarre, but also foolish.  I just can't see it as being possible.

Unfair, but understandable suspicion and speculation

There is also a fifth option, namely that nothing is wrong, and that she's just getting her training right when it matters, and that her "bizarre" race strategy is nothing more than typical even-paced running, as I explained above.  But people won't make that allegation.  Why?  Because they don't know anything, and they are driven by mistrust.  Therefore, they will settle on one of the four options that 'feeds' their mistrust.  

So they'll go with option 3 or 4 as most likely, and Semenya will face accusations that she is either cheating by manipulating her "advantage" through medical means, or she has been deceiving everyone for months leading up to the Games, and continues to NOT win on purpose. 

Both are unfair, and, I suspect, incorrect.  As I explained earlier, I think it's plausible that Semenya is running as fast as she can, and that 1:57.2 is the "limit" for her, in a more or less even race.  Maybe with a little more confidence, she'll be able to get her fast lap down into the low-57s range and break 1:57 for the race this year, but it's not impossible that a 57.69s and 59.94s is Semenya at her limit.  The sudden improvement in performance is more difficult to explain, but like any other debate based on performance, we must recognize that performance alone is not sufficient to reach a verdict!

It's also not difficult to see why people think differently - they don't know any better.  And that's because of the secrecy around the whole thing, and it forces people to speculate.  We shouldn't.  We should accept the control of the IAAF and trust that they have identified and managed a potential problem. In an ideal world, that would happen.  But I think it's naive to expect that of people.  Until people know, they'll make it up and everyone loses in that equation.  

If Semenya is to win people over, as she should - look at her interview after winning silver in Daegu, and tell me that this is not an athlete who is warm and genuine and worthy of positive sentiment - then the secrecy must be lifted.  Easier said than done, of course.  But what the future holds with these wild variations in performance, given the history of Semenya in the sport, is just not something to look forward to.

The marathon to close it down

The marathon tomorrow - join me at 11am London time for live splits and comments as it unfolds.  It's Kenya's last chance to rescue what has been a miserable Games, highlighted (in a big way, of course) by Rudisha's golden world record and Kemboi's gold.  They'll want gold in the marathon, but Ethiopia will be a stern test.

My money is on the Ethiopians - I think the Kenyans, who raced more recently, will struggle on the twisty course, and Kenya will regret not picking Geoffrey Mutai.

But join me tomorrow to see what transpires!


Thursday, August 09, 2012

London 2012: Spectacular 800m

Rudisha lights up London with a world record. And more 800m thoughts

The London Olympics saw its first World Record on the track tonight, and not surprisingly, it came from David Rudisha, who ran from the front, gun to tape, to become the first man in history to break 1:41.  His time was 1:40.91, and he pulled the other seven men to the fastest, most spectacular 800m race that we have ever seen.  More on that below, along with some other thoughts on his race.

1)  Rudisha's race strategy

The question ahead of the race, for many, was not whether Rudisha would win, but what the margin of victory would be.  There has been no greater favorite in an athletic event at these Games than Rudisha. His form this year has been spectacular, he has won paced and unpaced races, he has run from the front and looked peerless.

The biggest question was perhaps around the tactics he would employ in the Olympic final.  Front-running is the logical choice to most, because when you're about two seconds faster than the next fastest guy, you would want the pace to be beyond them.  Why allow a final 200m sprint, where a different type of physiological attribute can determine success, when you have such dominance over the whole race?  The problem in a final 200m sprint is that when the spread of runners is relatively narrow, the first 600m does enough "damage" physiologically that the person who is running with the greatest "reserve" is not guaranteed to win.  The ability to close in say 25 seconds is not a function of that reserve, which means that a 1:44 man can beat a 1:41 if they both get to that position together.  In Rudisha's case, I suspect he is so superior that he'd win anyway, but it becomes a far more open race than it might otherwise be.

So front-running was the option, and Rudisha was wise enough that he actually started to do this in his European races leading up to London.  We have seen many times how athletes become so accustomed to paced races on the circuit that they seem all at sea during a tactical race - the Kenyans in the 1500m looked this way earlier this week.  But Rudisha seemed ready, he had familiarized himself with the front-running pattern in a few races, including the Kenyan trials, and so everyone expected this approach.  Once he led after the break at 100m, it was clear that he was going for it.  It is easier said than done, however, largely for psychological reasons - putting yourself out as a pace-maker is never easy in an Olympic final.

It takes confidence and conviction, and Rudisha was good enough to do it.  He led the field through the bell in 49.28s, and then began to open the gap with 300m to go.  That's not surprising, because everyone in the race was running above themselves just to reach the 500m mark at that pace.

At 600m, which was passed in 1:14:30 (25.02s for the 200m split). Rudisha was clear, and on course for the record.

He slowed in the final 200m, covering it in 26.61s, but it was enough to break 1:41, and claim Kenya's second gold.  The manner of the win, plus the bronze for Timothy Kitum, will be some consolation for the nation that expected more than they have won to date.

2) The race was spectacularly deep and fast

Rudisha was chased home by a host of sensational performances.  In fact, every single position in the race set a world-place position.  Second went to young Nijel Amos of Botswana (a surprise) in 1:41.73, fittingly equalling Seb Coe's old world record), and then a further three men went under 1:43.  They included the two Americans, Duane Solomon and Nick Symmonds, who would surely not have believed that they'd break 1:43 and not even win a medal.  Even in last place, Andrew Osagie ran 1:43.77.  Only Abubaker Kaki of Sudan, who eventually finished 7th, did not run a personal best.  Three national records were also set.

It was just a spectacularly fast and deep race, and while everyone who was in it might feel stunned at their times without medals, they were part of something truly remarkable.  I suspect many would be wondering if a step up to 1500m might not make more sense, however - Rudisha is only 24!

3) The pacing - a pattern in the 800m

One final point about the race, and it relates to a peculiar pacing pattern that you see when you look at the best ever performances in the 800m event.   Part of my PhD looked at the pacing strategies used in all the world records from 800m to 10,000m, and there's a pretty constant pattern in long-distance races.  The 800m race is different, however, and is paced differently from other middle- or long-distance races.

However, I'm going to hold back on this discussion, for now, because David Epstein of Sports Illustrated will probably introduce it in his piece on the race.  I'll provide you that link, as soon as it comes out, and then I'll add the detail once his article is up.

So that's for tomorrow, a discussion on pacing in the 800m event.

There is much more to be said about 800m running, but on the women's side.  The semi-finals took place tonight, and they introduced us to a controversy that is just waiting to erupt when the finals happen on Saturday.  Semenya is back, having battled all year for half a second here and there in the range of 1:59-2:00, she tonight won her semi, looking incredibly easy and in a time of 1:57.  That's a 2 to 3% improvement, after a long season of many reasons 'stuck' at 1:59.

Cue yet another debate on "unrealistic" performance improvements, like those we saw with Ye Shiwen and Makhloufi.  Except this time, it's not doping that will be discussed.

But that may be for another day, keep your eyes open for the debate.

Until tomorrow, which brings some relay finals, and the women's 5000m, the second part of a Tirunesh Dibaba double, perhaps?


Tuesday, August 07, 2012

London Day 10: 400m success, gene pools and training, and a DQ debate

London Day 10: 400m success for the islands, and disqualification (doping?) debate

Quick thoughts on last night's action:

1) The 400m events were a Caribbean parade. As were the 100m sprints.  Is it the genes? Or the training?

Last night saw both 400m events - the flat and the hurdles, and both were dominated by Caribben islands.  Add this to the 100m sprints for both men and women (where Jamaica won four of the six medals - gold and silver for men, gold and bronze for women), and it's clear where the epicenter of world sprinting now lies.

Here's the finish results for the men's 400m, for example:

The same was true of the 400m hurdles final later, won, amazingly, by Felix Sanchez eight years after his Athens triumph.  That race had a strong American presence (three men), whereas the 400m race was the first time in the history of the Games that the USA did not have at least one runner in the final.

Then here's a graph (source: Sporting Intelligence) showing gold medals won per million people (on the y-axis) and per billion GDP dollars (x-axis) back in 2008, in Beijing.  Zimbabwe did well GDP-wise (thanks to Coventry, who won all four of their medals, but Jamaica and the Bahamas feature well in both categories.  Grenada now have their first gold, the Dominican Republic won two last night, with Sanchez's gold complemented by Santos' silver.  These Caribbean islands made up nine out of the 16 finalists in the two events, winning five of the six medals.  The same debate can of course be had for distance events, with a focus on Kenya, Ethiopia and Uganda as your main protagonists.

Of course, this invites the common debate about training vs talent.  Are these tiny nations, like the Bahamas and Jamaica, so successful because of a deep gene pool, which is somehow related to an accelerated "survival of the fittest" concept, as was discussed in a recent documentary featuring Michael Johnson?  Or is their success a function of their excellent school sports systems, their "culture" for sprinting, their investment into the sport and excellent training programmes?  David Epstein of Sports Illustrated described some of these factors in his report on Shelly-Ann Fraser-Pryce after her 100m gold.

I won't go into the whole discussion of training vs talent, and the 10,000 hour concept of deliberate practice.  Rather, I'll refer you to these two articles I wrote a while back, for those interested in more and who feel like a longer read:

  1. A review of the 10,000 hour concept: Is it valid, even for non-sport activities?
  2. Some of the evidence for genes, and why they matter

Unnecessary polarization

But for now, I will say that the polarization of this debate is unnecessary and wrong.  Why does it need to be one or the other, rather than both?  Why do we insist on discounting the role of genes?  I am sure that most of you reading this already agree, and probably wonder what the fuss is about.  Well, in popular media, in particular, the likes of Malcolm Gladwell and Matthew Syed have propagated the idea that genes don't matter, that it's all in the training, as I described in those previous posts linked to above.

So no, it's not just the genes.  And no, it's also not just the training.  We know that genetic factors influence performance, both in terms of the starting physiology, the adaptations that occur in response to training, and in all likelihood the ultimate ceiling that can be reached.  There's evidence from Bouchard et al, for example, that 21 distinct single nucleotide polymorphisms (or SNPS) affect how our VO2max changes in response to endurance training - if you have 19 or or more of the "right" SNPs, you respond well, whereas if you have 9 or fewer, you are a non-responder.  If you want to be an elite athlete, it's a pretty sure bet that you need to be on the high responder side of that spectrum.  That's just in terms of VO2max, and we know that endurance performance is made up of far more than just VO2max.

The point is that genetic factors clearly affect performance.  More in some sports than others - the "physiologically limited" sports like rowing, running, cycling etc may be more affected than say table tennis, archery, or even sports where a greater range of physiology can succeed (think football, and Messi vs Drogba).  But genes matter, and for speed and endurance, dismissing them is to dismiss scientific evidence

What has NOT been found is a single gene that explains it.  There was a prospect, when a specific variant of the ACTN3 gene was found to be associated with sprint/power ability in a group of European athletes.  That same gene is not associated with sprinting ability in other populations, but that's because of interaction effects and genetic differences between populations that I won't claim to be able to explain to you in sufficient detail.  But this failure to find the gene is often cited as evidence that there is no gene.  I'd argue that the scientific approach to the question is wrong, for two reasons.  First, it's far more complex than just being one gene - if 21 SNPs explain training response of VO2max, then you won't explain something so complex using a single gene approach.

And secondly, the question should not be whether there is a specific gene that some groups have that others do not - this is why geneticists and anthropologists get worked up and annoyed, because this kind of question leads to generalizations that are almost certainly wrong.  If you try to argue that Grenada, or Jamaica, or people of West African descent have a gene that makes them faster, then have a problem when someone NOT of that descent wins.

(As an aside, this discussion also opens a can of political correctness that I've never truly understood - I'd have thought it would be complement to be identified as "superior" in some task.  Obviously, if you're accused of being inferior, it's different, but that's not what's happening when we celebrate the world's fastest sprinters and distance runners.  I guess for every winner there is a loser, but it's funny that the "winners" are the ones who usually pull out the PC-stick! And it certainly doesn't diminish the achievement - it's not to say it's 'easier', because the training is still absolutely vital.  Anyway...)

For example, when Galen Rupp wins a medal in the men's 10000m race, you have to explain why he did, when clearly he isn't of east African descent.  If your position of "it's all in the genes" was based on heritage, it now looks weak.  But it was, in my opinion, the wrong position to adopt to begin with.  It shouldn't be about this population or that population, this descent or that descent.  I would say there is still a chance that some genes that affect performance ARE linked to descent, but for now, that's not even needed to explain why genes matter.

It should be whether the prevalence or frequency of the "favorable genes" is higher in some groups than others.  When you try to find a gene or SNP that you think Jamaicans may have that no other people do, you doom yourself to a negative finding, because that gene might be present everywhere.  You're asking the wrong question.  You should rather be looking to find whether that gene might exist in more people in certain groups, and thus whether the probability of producing a champion athlete there is greater.  This scientific question has yet to be answered, but may hold the key.

If this were the case, then the end result would an enormous difference in final performance because of the additive effect of having more to choose from, plus the system applied to choose it.  Does South Africa possess athletes who could challenge the Kenyans or Jamaicans?  Yes, of course.  But I'd hypothesize that we have a lower probability to begin with, and we don't maximize what we do have.

Based on this, I'd conclude that it is the application of the training system and culture to the right population, where the prevalence of whatever genetic factors determine success, that enables such dominance by a small population group.

That's my conclusion, for today.  The rest is the explanation, but here is a paper I co-authored with a geneticist, Prof Malcolm Collins, recently, where we explain how BOTH genetic and training factors are crucial for success.  Bottom line is that while Gladwell and Syed's fairy-tale that you can achieve anything if you practice sounds good, the reality is far more complex.

2) Track gets its own version of Ye Shiwen in Taoufik Makhloufi

Last week, the action in the pool produced a side debate on the Chinese swimmer Ye Shiwen.  The 16-year old won the 400 and 200m Individual Medley, and because of her age, her world records and the fact that she is Chinese, was deemed suspicious as a possible doper.

That story has no resolution, but now athletics has its own case, in the form of Taoufik Makhloufi of Algeria.  He won his 1500m semi-final, beating defending champion Asbel Kiprop and a host of other athletes, in pretty amazing fashion.  His last lap was 52.5s, with a final 800m of around 1:49.  He also improved substantially in the last year, about five seconds for 3:30.8 this year.  And he comes from a nation that is regarded by most within the sport as being 'suspect'.  In the same way that China is deemed suspect in swimming (and track, for that matter), North African nations have the same stigma.

Apparently, commentator Steve Cram said "That's unusual to see the Algerian run this well. ... I'm not sure what I'm watching with Makhloufi there ...", and the forums on athletics sites kicked off with debate and accusation over the possibility that he was doping.

So once again, you have a debate where some will say it's an unfair generalization (which it probably is), but others will point to history and how we've been fooled before.  Learning lessons from history is often the basis for generalizations, but applying them correctly is a difficult concept!  And once again, as was the case for Ye, looking SOLELY at performance leads to all kinds of conclusions that ARE certainly not fair.  For example, it soon emerged that there have been a few performances where the final 800m have been quite a bit faster than Makhloufi's, and they've often been in faster races.  Big improvements are also not unusual

So judging someone as a doper based on performance is just not feasible.  It was the same for Ye.  Some would say "she's young", but others could easily point to other young swimmers who were not suspicious.  They'd say "she improved by 7 seconds in a year", and others can point to even bigger improvements in non-accused swimmers.  Ultimately, performance doesn't cut it.

The performance does however ask the question, and given the history, it's right.  It's unfortunate for the individual, but history means it's his turn in the spotlight.  The only way to answer those questions is through testing, comprehensive and long-term.  Then, if the athlete doesn't get caught and doesn't slow down, then we must accept it.  If they don't get caught and slow down, we have a hint of an answer.  And sometimes they get caught.  If the testing is done properly, then time will provide the answer.  Of course, the problem is that the testing is not trusted either, because we've learned that it's too easy to get away with doping and not get caught.  But the more the better, it's the best one can hope for.

3) Makhloufi finds himself in a second, unrelated controversy

Then amazingly, the same athlete whose performance was hotly debated, found himself in a second, totally unrelated controversy when he was first kicked out of the Games and then later re-instated, after he was found to have deliberately under-performed in the men's 800m heats.

You'll recall the badminton players who got expelled for deliberately trying to lose to set up more favorable draws in the quarter-finals.  You'll also recall that Japan's women were instructed not to beat South Africa to get a better draw, and you may remember that a British cyclist confessed to crashing on purpose to force a race restart in the men's pursuit (he later retracted the 'confession').

So this has been the Olympics of "slower, lower, weaker", in some respects.  In the case of Makhloufi, he lined up in the heats, the morning after his 1500m semifinal, and presumably wanted to save himself for the final.  So he jogged slowly for 200m, stepped off and was done.  In response, the IAAF expelled him from the Games.

It was a bizarre sequence of events. Three quick thoughts:

  1. There's no consistency in the sanctioning of athletes for "not trying hard enough".  Some are expelled, others are not.  At least get the same method for all.  I realize they're subtly different, in the same way that an athlete who jogs in to qualify in fifth place in a 5,000m heat is different from one deliberately losing a heat like Makhloufi allegedly did.  Playing a weaker team to rest key players, or playing at 90% because you don't care to win a match, is different from deliberately manipulating the result to lose or draw, I'd argue. When you pre-determine the outcome, you're fixing the result, but the format of competitions and the rules sometimes facilitates this.

  2. I can't believe his federation would make him run the 800m heats knowing that he is a realistic medal chance in the 1500m.  The final is a day later, he's already limped off after his 1500m heat (unless he was laying the groundwork for his excuse a day early, that is), and so why push him to run a race that compromises his chances?  I presume the athlete didn't want to run the race, and that the Algerian Federation refused his request to withdraw.  We had a similar case in the swimming, where Chad le Clos had actually qualified for the final of the 200m IM, but withdrew because he wanted to focus on the 100m butterfly event as a better medal chance.  That was not sanctioned (rightly), and seems the common sense approach.

  3. Why does the Olympic programme not enable the 800m-1500m double by at least separating them?  Remember Coe and Ovett?  Their double attempts were a highlight of the Games.  Kelly Holmes won the same double in Athens.  Did those heats overlap with the finals of the other event?  They have done, in which case, I'll take this one back (no time to check, sorry!).  But it seems that it would be reasonable to enable the double with a schedule change.
In any event, he has now been re-instated for the 1500m final.  If he can produce the same 250m as he did in the semi-final, maybe that will provide more 'fodder' for debate on the forums, and another "Is he doped?" debate.

Should be interesting. That final is later this evening!


Sunday, August 05, 2012

London 2012: Men's 100m...It's Bolt!

Usain Bolt defends - 9.63s beats Blake into second

It's been a long day, with a fast finish in the form of Usain Bolt defending his 100m title in emphatic fashion, in a time of 9.63s, ahead of Yohan Blake (9.75s) and Justin Gatlin (9.79s)  A full recap tomorrow, but here are three very quick (Bolt-like) thoughts on the men's 100m:

1) Bolt second half was dominant

I hope someone will produce a statistical analysis of the race, with 10m interval speeds, as was done for the 2009 WR, because the structure of Bolt's race would make a fascinating comparison.  He was trailing over the first 30m in London, and so I'd love to compare this Olympic race to that WR from 40m onwards.  Was his first phase tonight slower than 2009, or were Gatlin and Blake better than the 2009 field?  Remember, when he set the WR in Berlin, Bolt actually reached 20m first! His start and drive phase were pretty good then.

Tonight, he was behind, but good enough to put himself in a position to capitalize, which he did emphatically.  His reaction time, incidentally (since I know this will come up) was 0.165s, which was faster than both Blake (0.179s) and Gatlin (0.178s).  But this is only a small part of the start - how you react is followed by how you drive from the blocks, and that's where Bolt has been found out before - he needs to get those long legs out of the way first.

Once that happens, he's unmatchable.  There was a moment tonight where Gatlin was holding Bolt off, but from that crucial 50 m mark, where most men begin to hit top speed before slowing down, Bolt just has so much more than everyone else.  He moved past Gatlin and Blake so powerfully, and those numbers would be very interesting to see.  Top speed and best 10m interval comparisons would be revealing.

Missing the world record by 0.05s suggests that it might not have been quite up to the top-end speeds we saw in 2009, but if the start was worse, it'll be close.  I'll look out for the numbers and report on them if I find them.  And obviously, if you have them, do let me know!

2) I'm surprised at the margin of victory

0.12s is a big win in a final that was so split before it happened - look at the spread on predictions, it was so evenly divided between Bolt and Blake that a win that size surprised me.  In Beijing, the margin was larger, of course, but Blake threatened to make this closer than it ultimately was.  Blake's 9.75s matched his performance from the Jamaican trials, and with the peak of the Olympic Games, plus all the talk of a fast track, plus the fact that the wind was a pretty decent 1.5  m/s tailwind, he might have expected to go a little faster than he did then.  Admittedly, it was cooler, which is not ideal for fast times, but I'm surprised Blake didn't go a little closer.

Nevertheless, he confirmed his standing as the challenger to Bolt, and the 200m event later this week should be another fabulous race.

3) A world record would have been a huge surprise

If you think about it, we hadn't seen a sub-9.70s clocking since 2009 and the Bolt world record in Berlin.  Prior to 2008, the best performances were in the mid-9.70s, and occasionally, a performance in the low 9.70s would light up the world of track and field.  When it did, it was usually a world record, like Powell's 9.74s, or Bolt's 9.72s in 2008.

Then came Beijing - 9.69s with a celebration.  Then followed Berlin - 9.58s.  We were taken into an era where track fans were eagerly looking forward to the fall of the 9.50s barrier.  Then, almost as quickly as it arrived, that era seemed to depart.  Since then, 9.75s has been the standard once again, and the world leader is a low 9.7-something.

And so going into these Olympic Games, the prospect of jumping from 9.75s all the way to a sub-9.60s was just beyond belief.  No track surface or reasonable following wind was going to allow it.  The times this year, from the big four (Gay included in the list) have frequently been around 9.80s, but never faster than the 9.72s that Bolt carried into Beijing four years ago.

Therefore, it seemed reasonable to assume that even if Bolt (or someone else) peaked for the Games, and produced the same kind of spectacular performance we saw in Beijing, he'd run somewhere in the mid-9.6s.  That Beijing performance, incidentally, was slowed by his celebrations, but we worked out at the time that he probably would have run around 9.64s - 9.65s had he not begun his dance while at the 80m mark of the race.

That was the basis for my prediction of a 9.68s winning time ahead of tonight's final (over on Twitter).  I didn't think Bolt would quite be in his Beijing form, which would have put him at 9.61 to 9.65s.  As it turned out, he was, and ran 9.63s.  But a prediction of 9.5-something, or even faster than that, was the result of being "spoiled" by 2008 and 2009 - we are back on the constant improvement curve now, and this performance is probably exactly where the "normal" would be.  Not that anything about Bolt is normal, of course!

Full recap of day's events tomorrow

I'll do a proper recap tomorrow, including some talk of the women's 400m, the men's steeplechase, and whatever other random thoughts enter my mind overnight!  Right now, I'm Olympics-ed out!


P.S.   Another stat that always seems to come up is number of steps taken.  I'm not sure you can read too much into this, because you need to know contact times, contact lengths, and force applied to the ground, but here are the numbers for those who are interested:

Bolt - took 41 steps 
Blake - took 46 steps
Gatlin - took 42.5 steps

All that really means is that Bolt has longer strides.  Looking at the height of the three men, you'd have predicted this, but people enjoy that stat!

London Women's Marathon

Women's Olympic Marathon: Live splits and comments

Welcome to live splits and coverage of the Women's Olympic Marathon from London.  I'll be updating the splits every 5km, and adding a few comments as the race unfolds

Splits and projected times

5km - 17:20. Projects a 2:26:17
10km - 34:46 (17:26).  Projects 2:26:42
15 km - 52:10 (17:24)
20km - 69:26 (17:16)
Halfway - 1:13:13
25km - 1:26:23 (16:57), projecting a 2:25:48
30km - 1:42:44 (16:21), now projecting a 2:24:30, but with this pace, the Olympic record is under threat
35km - 1:59:29 (16:45)
40km - 2:16:10 (16:41)
Finish - 2:23:07 (final 2.2km in 6:57, and a second half of 69:54)

Race report


Tiki Gelana of Ethiopia has won the Women's Marathon - it's a significant victory, and it comes as something of a surprise given the credentials of Keitany and Kiplagat.  Not that Gelana was an outsider - her PB is 2:18:58, and she won the Rotterdam Marathon earlier this year, so a highly pedigreed athlete.

Her winning time was 2:23:07, and a new Olympic record.  The race was a big negative split, after a cautious first half.  The pace, as expected, was ramped up by the Kenyans soon after halfway, and they put in a 16:21 5km segment from 25km to 30km.  That pulled a group of six away - three Kenyan and three Ethiopians, much as most would have predicted.  However, Russia's Arkhipova defied that script, and reeled the leaders back, and one by one, the east Africans began to fall off.

The 5km from 35 to 40km was run in 16:41 (3:20/km), and by then it was down to four, as Kiplagat had surprisingly been dropped early, and the Ethiopian challenge had been reduced to one.

The final 2.2km were run in 6:57 (3:10/km), and that gave Gelana the record and the win, as she was able to surge to gap a group of four, including pre-race favorite Keitany, Jeptoo and surprise Russian Arkhipova.  The other huge surprise was that neither Keitany, the London champion earlier this year, or world champion Edna Kiplagat, were able to medal.

So that's two women's distance races, and two Ethiopian gold medals.  First Dibaba beat Kenya's best in the women's 10,000m, and now Gelana has "stolen" the Olympic gold from the favoured Kenyans.  Add to this the Kenyan men failing to medal in the men's 10,000m, and the Kenyan hopes for London are being rapidly revised.  Their next chance to get on the board comes in the men's steeplechase later, where surely they'll win their first gold.

But for now, it's an Ethiopian celebration for Gelana.  She was the last surviving Ethiopian, with three Kenyans, but came through in the end.

The race was, also unsurprisingly, a massive negative split - halfway was hit in 1:13:13, which means the second half was run in 69:54. I must confess that I'm a little surprised that this was good enough to win - don't get me wrong, it's incredible running, but when you think of Keitany's London win earlier this year, she ran a 67:43 off a faster first half than today (70:53).

That is perhaps testament to the course, with its twists and turns, which were arguably made even more challenging by the conditions - it was wet at the start, and poured heavily in patches.  It may force a rethink about the men's race, because this kind of course may be ripe for surprise, and while Gelana was not exactly a long shot to win, as I mentioned earlier, few would have expected neither Keitany nor Kiplagat to win a medal.

As it unfolded


Keitany is dropped!  The pre-race favourite is in trouble, and is gapped, and there will be no coming back from this - too quick now and that is too big a drop-off for her.  It's Gelana doing the work now, capitalizing on that surprise.  Jeptoo is also being gapped and it's going to be an Ethiopian gold if this keeps up!

It's four women for three medals, and only 1.5km to sort them out in.  No one making the decisive move, and so this may boil down to a final 1km, like a track race.  Will Arkhipova (or Petrova, for those with confusion) manage to disrupt an African podium?  It'll be an amazing effort if she can.


Last five kilometers were done in 16:41, which is fast.  But here again, that doesn't tell the story of the fast-slow tactics being employed.  Arkhipova has been a good indication that those 16:41 have not been even-paced, which makes them even more impressive.  There were some undulations in that section, however, and they may also explain some variability in pace.


Arkhipova is being stretched here, every time there is a surge, she's off the back.  She's fighting for all she's worth, but the signs are there that when the big, final push comes, she'll find herself in the fourth position.

It's the east Africans who are being aggressive again, with only 3 km to run.


With four in the lead group (Keitany and Jeptoo for Kenya, Gelana for Ethiopia, and Arkhipova for Russia), the final 5km holds some serious questions.  The pace, for these women, has not been quick, so there should be surge and countersurge for the next 15 minutes.

Keitany to the front again, but it's not for a decisive increase in the pace.  So far, everything has seemed rather subtle.  It's been enough to crack those who are weak, but there's no indication, yet, of which of these four is on the limit.

I'm surprised the Kenyans have allowed the race to fall back into this kind of pace (17:15 per 5km).  It only allows runners in, and they have such superiority in terms of performance ceiling.  Then again, this is the Olympic marathon, and with much at stake, perhaps the caution is understandable.  But there's no question that when you are a 2:19 runner, as they are, then running 2:24 creates an "unpredictable" race, and that leaves a door open.


Last five kilometers in 16:45, so it has slowed ever so slightly, and that explains why Arkhipova got back to the group.  Interestingly, at 35km, the Russian went to the front, and Kenya was slow to respond!  Gelana made the first move along with the Russian, and so that is interesting.

The slow pace from 30 to 35km may have moved the Olympic record beyond reach, because they're now projecting 2:24:03, which means the 2:23:14 from Sydney may just survive.

Kiplagat seems to have been dropped now, and the only way she gets back is a drop in pace.  With 7km to go, that seems unlikely, and the Kenyan sweep is no longer on.


The group is now five, the small gaps to Kiplagat and Arkhipova have disappeared and so we once again have five.  The mile-by-mile splits would be fascinating, because I suspect the leaders are going fast-slow-fast-slow, rather than it being a case of digging in and returning to the pace of the front-runners for the Russian and Kiplagat.


The pace seems to have been lifted again.  The stimulus for that may have been the arrival of Arkhipova, because no sooner had she joined and Kiplagat was dropped, and now the gap is just starting to appear to Arkhipova.  The medalists may be taking shape, because if these two can't pull through this phase, then it's down to three - Keitany, Jeptoo and Gelana.

And a Kenyan is off, and it's Kiplagat!  That's a surprise as well, because she pressed Keitany in London earlier this year.  Now it's Keitany and Jeptoo, and the possibility of a Kenyan sweep seems gone, unless she can somehow recover.  The elastic is stretching, but not broken just yet.


Now Dibaba of Ethiopia has also dropped off, and so it's three Kenyans against a sole-surviving Ethiopian challenge.  That comes from Tiki Gelana, who now faces Jeptoo, Keitany and Kiplagat.

And Arkhipova has closed the gap to the leaders, and so now the lead group is back up to five.  It dropped to four, but it was effectively a substitution - Dibaba is out, Arkhiopova is in.


Keitany is pressing the pace, she is at the front and has been for the last 5km, and so is feeling confident enough to lead.  Nothing dramatic, if you don't count a 16:21 5km split as not being too dramatic!

Of the chasers, Flanagan is 17 seconds back and hoping for a blow from three or four women in front.  Arkhipova is also in front of Flanagan, and so she represents the first non-African threat to the medals.


30km is reached in 1:42:44, so the last 5km were done in 16:21, and that's been the source of the damage.  That's a pretty significant increase in pace compared to the first half (3:16/km compared to 3:27/km), but it only tells part of the story - the surges are doing more damage, and it's the reason we're seeing three Kenyans against two Ethiopians for the final 10km.  This is the section of the race that Keitany covered in 31:53 when winning London earlier this year.  Again, that was a different course, with fewer twists and no water, but the question is whether she has that in her again.  Or possibly even more, given the slower pace to date.


Now Zhou of China is also just dropping off slightly, and so it's five again.  That's a function of a slight increase in the pace at 27km, and we'll see how the 30km split reflects that.

No single person is assuming the leadership of the race, but it's Kenya at the front.  Side by side, the three Kenyans are pressing now, and the race is becoming just what everyone thought it would be - a direct clash between the Kenyans and Ethiopians.  So far, this marathon has had so much more tactical intrigue than the 10,000m race last night.


One of the Ethiopians has dropped off - Mergia, who won the Dubai Marathon has lost about five seconds to the lead group.  That group now has three Kenyan, two Ethiopian, one American (Flanagan) and one Chinese women in it.  Flanagan is being stretched by the pace though, a gap just starting to appear.


The last five kilometers were done in 16:57.  That's not breathtaking, but most of it was in the last two kilometers.  That was when first Keitany eased the pace up, and then Kiplagat surged, and the result is that split.  Goucher, Flanagan, the Chinese, Mikitenko are all off the back, but they may be able to claw back if the pace slows. But I'm expecting more of the same in the not too distance future.  The non-Africans will actually provide a nice barometer of the pace - if they drop, it's been ramped up again.  Look for the splits to get faster and faster now.

Now the big move, by Edna Kiplagat and within one kilometer, it has become an African race and a "rest of the world race".  The lead group has gone from 12 to 6, and not surprisingly, it is made up of three Kenyans and three Ethiopians.  Here comes the race!

The front group is now being headed by Mary Keitany, the champion in London earlier this year, and so perhaps the race is now about to become an African affair.  The lead group is thinning, but still large, with about 12 runners in it at 25km.

Some real race developments since halfway.  Biggest of all is that Shobhukova has stopped, clutching her right hamstring, and so an injury presumably explains her poor showing.

Keitany must be an enormous favourite here - apart from that second half of 67:43 in London this year, she covered the final 10km in an astonishing 31:53.  You can see those splits here


The most interesting development so far is that Lilya Shobhukhova is off the front group - she is 16 seconds down at half way.  There are 22 in the lead group, but she is not one of them, and that is a big development!

The split at halfway is 73:13, so it's on for a mid-2:26, but the pace may ramp big time from here on in.  Interesting stat - when winning the London Marathon earlier this year, Mary Keitany ran the second half in 67:43, and that was off a 70:53 first half.  Today, the first half is over 3 minutes slower, and so that gives an indication of what Keitany can do in the event.  If she produces anything like that speed, nobody in this race will match her.  The question is whether she can do it on the different course, in the wet?


Still no action, other than the teams working together to get water, and the dropouts.  The latest to step off the road is van Blerk of South Africa at about 18km.

The group is thinning, but only very, very slightly, as the 2:30 runners begin to drop off the 2:27 pace.  It has been an incredibly consistent pace - 3:27 to 3:29 for every 5km interval so far.


Not much change, literally.  In fact, the pace has been almost identical the whole way.  The first 5 were 17:20, then 17:26 and now 17:24.  So nothing to speak of yet.  The rain has relented, leaving only wet roads to contend with.

First 10km

At 10km, the lead group is still large, about 20 women, but there's still no real split in the race.  All the big names are there - three Kenyans, three Ethiopians, Shobukhova and the two American's featuring prominently at the front.

The slow early pace is no surprise - unpaced marathons rarely go out hard, and with the stakes as they are, and the field as deep as it is, expect a cautious game until just after halfway.  It is not inconceivable that the second half might be in the range of 68 minutes, depending when the attacks come, because the first looks likely to be in 73 min.

The biggest factor so far has been the weather - it has been wet since the start, with period around 5km where it poured down.  Women were jumping puddles.

The Kenyans are known to dislike running in the rain more than most, and so this may well prove to be an equalizer of sorts.  We'll know after halfway, as that's where the increase in tempo is likely to happen, at least based on Keitany's marathon past.

The other news is that of a few significant drop-outs.  Mara Yamauchi of GB pulled out with what appeared to be a leg injury, and the USA's Desi Davila pulled out even earlier, having carried a hip injury into the race.

London 2012: Day 8 recap

London 2012 Day 8 recap: Great Great Britain, Men's 10,000m and the fastest woman in the world

Five very quick thoughts after an amazing Olympic Day...

1) Great Great Britain

What a day for the hosts. In track & field, they got three golds in an amazing hour, first with Ennis, then Rutherford in the long jump, and then Farah in the 10,000m.  That followed two gold medals in the morning's rowing and another track cycling gold medal, and Great Britain was in a frenzy by the time Farah kicked to gold.  They lost a penalty shootout to Korea in the football, but little will dampen the enthusiasm.

2) Ennis delivers under pressure, and with style

Jessican Ennis carried with her the expectations of everyone, as she was made the face of the Games in the year leading up.   She produced a spectacular performance, filled with PBs and charisma.  Her medal ceremony would have raised the roof on a closed stadium, and must be among the all-time highlights for a host nation.

3) Farah wins the 10,000m gold, with Rupp in second, for the worst African showing in decades

First of all, the gold went to a deserving champion in Mo Farah.  He may have been denied in Daegu last year, but he has been the world's premier distance runner for two years, and he duly bagged the gold that his status warranted.  He did so in a peculiar race, at least by my watching.

Going in, everyone had seen the movie before - go slow, wait to the last lap, and watch Mo Farah sprint away from you.  Kenenisa Bekele had seen it more than once, and though it was not always Farah out-kicking him, it was someone, and so he needed to avoid a re-run.  Yet that's exactly what happened.  The first 2km were painfully slow - 6 minutes.  It got better, first with Zersenay Tadese and then the Ethiopians pressing the pace into the 2:40s per km.  But in truth, all they were really doing was dangling off the front, and 5km was reached in 14:04.

After that, it was clear that Bekele and the Kenyans (well, everyone) needed a even faster pace to disrupt the 'script' that Farah would have been hoping for before the race began.  It never came.  Ethiopia tried, sort of, but again, they had men on the front, but the pace stayed firm but unspectacular.  The laps counted down and the opportunities to change the script were ticked off one by one.  By the time the bell was reached, there were still 11 men in the lead group.  I don't recall ever seeing a group that size at the bell in a championship 10km.  It was testament to the lack of punch at the front of the race.  And if it was not doing damage to ten men, then Farah, comfortable with a considerably faster pace, was certainly not going to be put into difficulty.

Farah was, of course, unmatched in the final sprint, winning it easily with a lap of 53.48s, and a final kilometer of 2:28.  Those are fast, but not eye-poppingly quick numbers.  I suspect that in the end, nothing anyone could have done would have beaten Farah - a 26:50 and he would have won, a 27:30, and he did win.  But the second half of this race was done in 13:26, so for all the "front-running", it was just a really tepid tactical effort.  Perhaps the east Africans just didn't have the weapons to change the structure of the race, because Bekele was clearly not at the kind of level to be aggressive off the 64 to 65 seconds per lap pace that was set for most of the race.

In the end, then, it was the other Bekele, Tariku, who got closest to Farah, but even that challenge faded, and in the home straight, it was Galen Rupp who came through for the USA's first medal in the event since Billy Mills in 1964.  That will no doubt inspire almost as much celebration in the USA as Farah's win for GB, and it will be interesting to see if either can grab a second medal in the 5,000m.  Farah looks a good bet in that event, and with Dibaba looking like a women's double winner, we may well have a repeat of Beijing, where two double golds are seen.  The difference is, it won't be Bekele.

4) Fraser Pryce defends

In the final track event, Shelly Ann Fraser-Pryce defended her 100m gold with a dominant performance of 10.75s.  She had a tailwind (1.5m/s) in cool conditions, and used her explosive start to set the race up.  Carmelita Jeter challenged, and came close, with a 10.78s good enough for silver.  Veronica Campbell-Brown took the final medal in 10.81s.  It was a fast, deep race, with six sub-11 second clockings.

The track is clearly fast, then, which augurs well for the men's race tomorrow.  There, all the big names got through, and possibly saw the addition of a 'new' name to the mix, in Ryan Bailey.  He is of course the third American, and you don't get into that team without having real aspirations on a medal, but it was a surprise to see his 9.88s PB in the heat.  Admittedly, he ran all the way through the line, whereas all the others - Bolt, Blake, Gay, Gatlin, Powell - were able to coast for at least part of the race.  Bolt seemed to barely break out of a fast stride, looking in his heat like Fraser-Pryce did in hers.  Blake also looked mighty impressive, and tomorrow's sprints should bring the times down into the 9.80s for everyone.  Whether Bailey has a "jump" will be tested.

While on the matter of times, I thought I'd throw out a predicted time for the men's final - four years ago, Bolt won in 9.69s, but he cost himself between 0.03 and 0.05s with his celebrations.  So call it a 9.65s time.  One year later, he ran 9.58s, but has since returned to the 9.7-range.  So too have other men - Blake is in the low 9.70s, Gay seems to be capable of getting there.  Back in 2008, Bolt came to Beijing off the back of a 9.72s world record.  That is slightly faster than the times we've seen this year, though the big contenders have not shown their hands just yet.

However, given the trend in sprinting in the last few years, it's difficult for me to see a sub 9.60s clocking in London - a performance jump of 0.15s (which is more or less what it will take) seems too large, given that three years haven't produced anything under 9.70s.  The fast track is making me reconsider this, but then the cool conditions in London by 21h30 mean slower times too.  So if I had to guess right now, I'd say that a time of 9.68s will win gold tomorrow night.

Then again, I may be totally wrong! I hope I am, and I hope we see a record.  What I do know is that we'll see an incredible race - like the men's 10,000m, it has so many story lines in the race.

5) Tomorrow's action - a women's marathon to kick off

It's been a good games for defending champions.  Given how rare it is to defend an Olympic title, to have seen three in the first eight events is quite something.  Bolt will be aiming to be the fourth, although Ezekiel Kemboi will be hoping to do that when he races the steeplechase just before tomorrow's programme closes with the 100m final.

Also tomorrow, the women's marathon, and what should be an incredible race.  Here's a good pre-race preview, courtesy Letsrun.  My pick, for the sake of it, would be Keitany, followed by Kiplagat, in a repeat of London this year.  Shobukhova's performance will be fascinating, but I think the racing aggression of the Kenyans wins the top medals tomorrow.