Kenenisa in 26:17 Form Could Run 2:00:36? So Say MIT Guys

avian expert wrote:

2-03

There is no reason to believe bekele is 3 minutes faster than this top talent pool

I didn't miss the point. You just can't admit that your point is incorrect. You are claiming the Kenyans aren't close to 2:00, but the Kenyans haven't finished improving yet. You keep saying they have maxed out at 2:03 but they haven't. They have run 2:03 and have broken the record at least 5 times in the past 10 years. That's more frequent than any other event.

The other part of your argument about Kenyan top talent moving to the roads makes little sense. Just because you are the top Kenyan doesn't mean you are as fast as the previous top Kenyan. Or in american context, the annual #1 highschool miler isn't always capable of a 3:53.

Do you believe the Kenyans are going to 'keep on improving' as you put it in a linear fashion until the record hits 2-00?

It doesnt work like that

No one in history has had 2-00 ability

avian expert wrote:

Do you believe the Kenyans are going to 'keep on improving' as you put it in a linear fashion until the record hits 2-00?

It doesnt work like that

No one in history has had 2-00 ability

I didnt say it worked like that. I also did not say anything about a "linear fashion" nor did I say I believed they would keep improving to 2:00.

You don't know how it works.

You don't know anyone's ability.

Duncan, I can at least agree with the statement that "If an athlete runs 12:37 and 26:17 while in Marathon training, then it converts to 2:00:32" but of course, that is not K.Bekele.

Almost all athletes need a transition period to adjust their engines from their 5k/10k best to Marathon, which bring back their 5k/10k times. Then, they fall in line with the statistics of being 26:50/13:00 guys. 25'45/12:25 athlete may be able to do it, so I agree with Ventolin.

Do you even know probability, BRO?

Well, considering that their's a 3.5 min deviation per performance, this statistic holds true to a 3 hour guy. It's very likely someone can be a 3:03 or a 2:57 guy based off 17:00/37:00 best. I am sure accuracy can be held true for 90% of the population, but elites gets tricky. The probability doesn't make sense when you apply it to 12:37/26:17 to get 2:00:32 because its either 2:04:00 (expected) or 1:57:00 (unhuman). Unlike the 3:00hr athlete, where probability is pretty spread out around the standard deviations, its only one-sided with K.Bekele.

The variance on normal performances can not be the same as the variance on human limits. So while an athlete (or 4 like I've pointed out) have refuted the statistic, it is still valid for the masses.

avian expert wrote:

2-03

There is no reason to believe bekele is 3 minutes faster than this top talent pool

He was a lot more talented than them. Do you think he would have let Mo p1ss all over him like this lot every champs. This lot might have been good on the track no way they had the pace or sprint to match Bekele. Same as Geb if he had concentrated on marathon at peak his WR would have been at least a min faster

Whether that translated to a top marathon is debatable for Bekele but he was unmatched over xc indicating superb extended endurance. He has run well at marathon as well despite being past his best

The 3.5 min deviation ALMOST certainly is not constant across all times AND does not hold true for a 3 hour guy.

You literally know nothing about statistics, machine learning, prediction algorithms or probability. Go find anther thread to yap on. I can barely even comprehend your jabbering. Do you know what structured thought is?

avian expert wrote:I see that renato has emphatically debunked ventolins claim that top marathoners are in 26-30 10,000 shape

I see that ventolin is also unsurprisingly persisting with this myth

eh ???

ask renato

how fast 10k track shape sammy was in peking ?

or

how fast 10k track shape geoffrey was in awesome new york '11 where i even had to bend knee at the awesomeness offered !!!

wow !!!

he jogged along for 30++ on toughest course out there & busted oblivion-like-hell-hath-no-damnation last 10k to almost dip into

2"04's !!!

this is spanish version of highlights of race

ask canova what he reckons geoffrey's intrinsic 10k speed was that day

if he says slower than 26'30, i promise not to post again until he says so

Duncan Blythe wrote:

ventolin = function(T1,D1,T2,D2,Dnew) {

return (T1/D1 + (T2/D2 - T1/D1)*log(Dnew/D1)/log(D2/D1) )*Dnew

}

I tried this on the data we looked at in the paper and got RMSE around 0.1.

Thus this does not seem to be a viable approach

eh ???

i said that looks wrong for algorithm, but i'm maths-interest not code, but try to force to myself to learn some chi-square

can't you as

"big einstein"

work out algorithm for

1'43.00 / 13'00 ->

2'12.07

3'27.15

4'44.75

7'25.21

27'43.07

???

net wrote:Maxed out at 2:03? What are you talking about? The world record is 2:02 and has been broken at least twice in the last 1.5 years. None of those guys has demonstrated the 10k ability of a peak Bekele.

eh ???

none of them have run a track 10k in peak M shape !!!

tadesse, who turned out crap M runner was angry with his 26'37pb when beat-up by scrub kogo

he complained after, pace was too damn slow & he lost a sprint & he had no sprint, which is no shock

he was right

it was crap pace to 5k

he had no 5th gear

he was an all-out 3rd - 4th gear guy

( stick-shift anyone ?? )

he expected to run that day. but smashed by drudgery to well above former :

~ 26'20 / 26'25

ask canova what he reckoned about zedster's expected time that day...

I'm very close the evaluation of Ventolin (also if a little bit more... humble about performances of Marathon runners !).

In my opinion, 26'30" is not a very strong time. I had athletes, of sure less talented than the best Marathon runners of today, and NOT fast (for example, Albert Chepkurui, in Qatar Ahmed Hassan Abdullah, who ran 12'56" and 26'38" with a real, personal value in 1500m of 3'42", something possible for Wilson Kipsang and Kimetto too, of course with two months of training including some specific workout of speed endurance, or John Korir, 5th in OG 2000 in Sydney when only 19y, able running 26'52" with a PB in 5000m of 13'09"), who ran at a speed today considered almost ecceptional.

I don't want to speak about Nicholas Kemboi, who, in my opinion, was one of the most talented athletes I saw in my life. He was able to move from 28'19" to 26'30" when we had the opportunity to work together in St. Moritz for two following months in 2003, and of sure in two months he could not achieve his best, but after that period had to face several problems, and never was able to run according his talent.

When Ventolin continue to remember the 10000m of Samuel Wanjiru in Bruxelles 2005, immediately behind Kenenisa (passage in 13'10"), we need also remember that, after 5 km, the small leading Group included Nicholas too, who was no more able to run at good speed after 7 km, but in any case finished in 26'51".

If I consider Nicholas' approach to this race, you can understand why, for me, 26'30" is not a great time :

Nicholas was in the best shape of his life before WCh 2005 in Helsinki. He did all the preparation together with Moses Mosop, but was clearly stronger as speed and specific speed endurance, and at the same level as long endurance (we went two times for long run, over 30 km, and they ran 1:33 without specific preparation for Marathon).

Since Moses won bronze medal in Helsinki, running the fastest last lap in the competition (faster than Kenenisa and Sileshi Sihine, but did the mistake to be 5th at the bell, and had to overtake Zersenay Tadese and Boniface Kiprop, wasting time), I can suppose in 2005 also Kenenisa could have had problems to win the race.

But with 4 laps to go Nicholas was spiked, and finished the race limping, almost jogging, in 27'16" (9th position). Immediately after the race, he had 16 Staples, because the injury was very wide, and went to involve, fortunately in very little part, his Achille's tendon too.

The doctor of WCh put the leg of Nicholas inside a plastic for 15 days, for not using his feet, and he walked using crouches.

Immediately at the end of the Championships, we went back St. Moritz with all the Qatar Team. Exactly 12 days after the race, Nicholas, still with the plastic in his leg, told me he wanted to run Brussels because didn't want to waste all the training we did before. I refused, because the risk was too big.

The next day, Nicholas came to me without plastic : he went for himself to a doctor (not the doctor of Qatari Team) for removing plastic ant Staples, and started to walk, telling me : "You see, I can walk, and in two days I can try running". At the end of everything, he went for 40' jogging, and I told him : "I want to be sure you can run, Tomorrow we need to try some 1000m on track with training shoes, for controlling the reaction".

Next days (Brussels was on 26th August, on Friday, and we speak about Tuesday...), he started with 3'07", then 3'03", and at the end ran 12 times 1000m with the last in 2'42". I can't refuse again to put him in the race, but I advised him : "Kenenisa goes for the WR, and want a passage of 13'10". There is the rabbit for the second Group going in 13'28", please, go with the second Group, your approach for this race doesn't allow you to run so fast" and he answered "Be tranquil, I stay with the second Group".

Instead....

The fact this times seems very tough today, is because the event is no more frequented at high level, and there are no more competitions. Another athlete at 26'30" was Abebe Dinkessa, who of sure was not more talented of Geoffrey Mutai and Wilson Kipsang.

About the Geoffrey of 2011 (Boston or NY is the same), I think he could really run very close 26'30", the same with Moses Mosop.

When, at the end of June, we had in Boston the race of 10 km (road), and Geoffrey won with 27'19" after the first 5 km in 14'02", running the second 5 km in 13'17" and the last 3 km in 7'51". Of sure, the Geoffrey 2011 was able running like Abebe Dinkessa !

Looking at these situations, don't think 26'17" was the real value of Kenenisa. He told me NEVER prepared the event, but won a lot of medals because was the strongest in the world. Kenenisa told me that, the two times he attacked and bettered the WR, was not in top shape, and not in perfect health : nothing to do with the shape he had when bettered the WR of 5000m in 12'37". We discussed about what he could have run in several events, looking at his training, and my opinion about his real, hypothetical value (I don't like to talk about what "could be, if...", but this time I can do an exception and I feel myself another Ventolin...), knowing his training (something nobody else knows, since Kenenisa was Always very much jealous about his training), I can say 3'28" - 7'20" (his main regret is never had the opportunity to try fast 30000m when in top shape) - 12'33" - 25'55" - 57'40. I don't include in this analysis the Marathon, since his type of training in the first part of his career never had something to do with this type of event. However, I already pointed that, before Dubai, in spite of a training not completely adequated as volume and specific intensity, because the tendon problems, he showed 3-4 workouts able to allow an athlete to run 2:04. This means he has, of sure, the possibility, moving to Marathon many years ago, and using specific training, to run under 2:02.

But 2:00:36.....

Renato

I agree with much of what you write but wonder how Kenenisa could not have been in top shape for his 26:20 when it was 9 days after his 12:37. It seems he WAS in top shape.

ventolin^3 wrote:

can't you as

"big einstein"

work out algorithm for

1'43.00 / 13'00 ->

2'12.07

3'27.15

4'44.75

7'25.21

27'43.07

???

No that would be mathematically impossible. You need

to provide at least one other prediction from 2 different times.

Then we could check the formula is correct

'I'm very close the evaluation of Ventolin (also if a little bit more... humble about performances of Marathon runners !).

I agree that peak Bekele if training only for marathon could have run sub 2:02

Kenenisa told me he was not in good health during his first WR. After his 5000m had cough and fever, and the situation was by far not the same of the WR in 5000m.

About the second record (Brussels 2005), instead, it was really the shape his problem, and when he explained it, I understood why, in Helsinki, he didn't impress me, and I thought Moses could also win, if his position was number 3 at the bell.

Kenenisa clearly told me his best shape for 10000m was when he won his first WCh title in Paris 2003 (second 5000m in 12'57"...), and never he reached the same level on the distance in the next seasons.

Renato,

Would it be possible to post the training of Kemboi and Mosop leading up to the WC 10,000 in Helsinki or at least post a few workouts?

Thank you for your contributions.

Renato Canova wrote:

knowing his training (something nobody else knows, since Kenenisa was Always very much jealous about his training), I can say 3'28" - 7'20" (his main regret is never had the opportunity to try fast 30000m when in top shape) - 12'33" - 25'55" - 57'40....

But 2:00:36.....

Renato Canova,

2:00:36 sounds very fast, but perhaps we aren't fully appreciating Dennis Kimetto's real ability during the 2:02:57. Using this algorithm,

you can calculate real ability with better pacing. This makes Kimetto's 2:02:57 ~> 2:01:30 (The marathon is not an option on the website now, but I will try to add it later today). Is it not acceptable that a perfect Bekele could run 1-minute faster?

Regarding Bekele, I've also suspected that Bekele had similar ability to what you say (3:28, 7:20, 12:33, 25:55, 57:40) in 2003/2004.

Using 12:37.5 and 26:17.5, I get:

1500 3:31.37

1-mile 3:47.72

3000 7:21.06

5000 12:37.5

10000 26:17.5

half-mara 57:49.08

marathon 120:02.72

However, if I find his real ability

,

12:37~>12:32.71

26:17~>26:07.73

Then I get,

1500 3:29.62

1-mile 3:45.86

3000 7:17.81

2-mile 7:51.64

5000 12:32.7

10000 26:07.73

half-mara 57:31.18

marathon 119:30.49

The pacing algorithm isn't complete yet, so here's another possibility,

12:37~>12:31

26:17~>26:07.73

Then I get,

1500 3:28.29

1-mile 3:44.48

3000 7:16.06

2-mile 7:49.87

5000 12:31.0

10000 26:07.73

half-mara 57:37.35

marathon 119:56.13

This suggests that a perfect Bekele would be 90-120 seconds faster than Kimetto in a marathon. Do you find this to be reasonable? In other words, do you think a perfect Bekele is ~21.3-28.4 seconds ([90-120]*10k/marathon) faster than a perfect Kimetto in a 10000m?

PS - Forget those MIT guys, west coast is the best coast ;p

Dear Renato,

sorry, I did not see your second long post earlier. So here are some comments:

> What I say, is that there are too many casula and individual factors that can't be reduced to a mathematic formula.

This is true, but it does not mean that a formula or algorithm cannot make good predictions.

When dropping a ball from the tower of Pisa, a mathematical model cannot take into account (for practical reasons) the movement of every dust particle and air molecule, or whether a bird will catch the ball, then will be struck by lightning, and drop the ball on the experimenter's head.

Nevertheless, one can predict rather precisely when the ball will hit the ground. In most cases, that is. And even if the last thing happens, one would not say that it refutes of the theory of gravitation.

In sports, there is just more uncertainty.

And I would argue that the degree of this uncertainty is not of a degree that would make predictions impossible.

After all, even Riegel's formula (while it has its limitations) is not that bad, and the Purdy scheme works quite well.

(there is empirical proof of this)

> This is the case of every Elite : the research. for having some statistical validity, must be done with omogeneus subjects.

That is in principle true, though maybe not in a way as one may think.

The subjects need not be homogenous in the sense that they are all the same - all one needs is that they come from a pool comparable to the subjects you would like to make a statement about.

In most cases, the subjects for which predictions are made are different from the ones you have already seen - but this it not per se problematic, since if they were the same, it would not be prediction...

>1. I just stumbled upon this thread - why was I not invited to this discussion?

1. Sorry, I might have sent your formal invitation to the wrong address. Is there, perchance, a person with a name like yours but without the underscores?

>2. Franz, beszelsz magyarul??

2. persze, hogy is ne, ezzel a névvel?

(but let's converse in English, otherwise the other people will assume that we are swearing behind their backs)

Renato Canova wrote:

Kenenisa told me he was not in good health during his first WR. After his 5000m had cough and fever, and the situation was by far not the same of the WR in 5000m.

About the second record (Brussels 2005), instead, it was really the shape his problem, and when he explained it, I understood why, in Helsinki, he didn't impress me, and I thought Moses could also win, if his position was number 3 at the bell.

Kenenisa clearly told me his best shape for 10000m was when he won his first WCh title in Paris 2003 (second 5000m in 12'57"...), and never he reached the same level on the distance in the next seasons.

Renato, your information is quite interesting. However, it does strike me as a bit surprising that Kenenisa would have attempted the 10,000m record in 2004 if he had a fever and cough subsequent to the 5000m record, and therefore, by definition, within a week or so prior to the 10,000m record. I would concur that Kenenisa was not looking as dominant in Helsinki in 2005 as he had in some other championships.