Could a fit Alan Webb break 3000M WR?

ventolin^3 wrote:

now imbecile, look up top 20 alltime 3ks steeple guys & look up what they ran for the 3k flat in the same year

if no figures for a particular guy, go to next in list & next after, until you get 20 inputs

then work out the mean & then give 95% confidence intervals

now idiot, get a clue, if you want to disprove the 25 - 30s suggestion, then show some hard mathematical proof if your peabrain allows it

Alternatively, we could look at the top performers ever over 3000m, take the top 20 who DID run a steeple that year, and look at the average differential. Bet it wouldn't be 25-30s.

It's hilarious how you're throwing out "big words" like "confidence intervals" when you clearly don't understand the first thing about statistical sampling. Pre-selecting the sample population based on the most extreme cases of a dependent variable that is known to be highly correlated with the independent variable is an EGREGIOUS violation of the non-biasedness assumption and totally invalidates any "confidence intervals" you may wish to construct. It's akin to trying to predict the height-weight ratio of very tall people, then using data from the world's heaviest 20 people. Though not as extreme, it is the same type of model error.

I think you're a bit confused on the term "mathematical proof" as well - I have two degrees in the field, and it's risible to suggest "statistical evidence" based on extremely biased data is anything near "mathematical proof."

kartelite wrote:

Alternatively, we could look at the top performers ever over 3000m, take the top 20 who DID run a steeple that year, and look at the average differential. Bet it wouldn't be 25-30s.

It's hilarious how you're throwing out "big words" like "confidence intervals" when you clearly don't understand the first thing about statistical sampling.

kartelite, you're making a fool of yourself. Eldrick certainly understands statistical sampling, and you don't understand steeplechase running.

Looking at the top performers at 3000m won't really tell you much at all, because the differentials will be artificially small due to the relatively few high level flat 3000 opportunities for most steeplechasers. Flat/hurdle differentials are fun factoid, but don't read too much into it them unless the numbers are very large. In that case, it almost always is a indication that the athletes best events are away from the beams and water event.

Got Wheels? wrote:

Malmo, just out of curiosity, what was your flat out 400m time during your prime years? A ball park figure would suffice! There's always a lot of talk about the elite's basic speed, and usually they are assumptions rather than facts so I was merely interested if you'd share your personal experience. Thank you.

5/28 Sat - PM 3 easy, 2 x 400 barriers (66, 65), 4 x 400 (62, 61, 60, 52.2), 3 x 200 (30), 6 easy

I figure I might have been good for 51.5 some day, maybe as fast as 51.0. In that workout we were hauling ass as soon as we hit the line.

A lot of you guys are making the erroneous assumption that when a guy like Bekele runs a 52 at the end of the 10k that it means he could run much faster in an all-out 400m. That's usually not so. Most people who think that way is becasue they've never been around elite runners and they let their fecund imasginations run wild. Good kickers are good becasue they are able to run closer to top-end at the end of a race than their peers are.

Marty Liquori is a good example. Marty's fastest 200 is about 25 flat, and he was certainly one of the best kickers in the world in his day. He could run sub 26 at the end of a race. When you look at a sub 26 at the end of ANY race is hauling ass. It looks impressive whoever is doing it. Years ago myself, Liquori and Steve Scoot did a little tune up workout before World Cup. Basically it was a ladder: 200, 400, 600, 400, 200 @ 60 pace. Nothing difficult, just trying to keep the legs from getting stale. On the last 200m Scott asked permission to let his legs run. Marty and I approved and when we hit it Scott was flying. He ran 23.5, while Marty and I ran about 25.5. We were both running all-out. At the finish we looked at each other with grins and said something like 'holy shit'.

Care to take a guess which of the three of us didn't have a kick on race day? I'll give you a hint: it wasn't Mitch Cumstein.

malmo wrote:

kartelite, you're making a fool of yourself. Eldrick certainly understands statistical sampling, and you don't understand steeplechase running.

Looking at the top performers at 3000m won't really tell you much at all, because the differentials will be artificially small due to the relatively few high level flat 3000 opportunities for most steeplechasers. Flat/hurdle differentials are fun factoid, but don't read too much into it them unless the numbers are very large. In that case, it almost always is a indication that the athletes best events are away from the beams and water event.

Malmo, my point went right over your head. It was that steeple specialists run relatively few high quality, flat 3000's so the differences are smaller than they should be. Which you just regurgitated while calling me a fool. Kinda like how on a recent thread you were insisting that the OP wasn't ambiguous with his statement and I was an idiot for thinking otherwise, and then it turns out you were completely wrong. I was implying it's just as foolish to look at the top flat 3000 runners with steeple times, who are more likely to be "casual" steeple runners (although most people at that level will only do it as a primary event), so the differential could just as easily be biased on the other end. If you had understood my very simple height-weight ratio you would have grasped that.

If you really think using biased data and then touting it as "mathematical proof" constitutes an understanding of statistical sampling, then this debate isn't even worth having.

ventolin^3 wrote:

Mr. Obvious wrote:

Why are you basing your line of fit on 400m and 800m speed for a guy who never ran the 400 and 800? I just don't get the methodology here. Why don't you base your inputs, I don't know, on races he actually ran?

because it reduces every runner to his basic DNA

you then go along the line & pick off their "best" distance - it may be 1500 or 3k or 5k or 10k or some even go all the way along like paula

see example of hicham & morceli - it is likely they had virtually identical 400 ability & only differed at potential 800 by ~ 0.25s ( i get 1'43.00 to 1'43.25 )

therefore, all the difference in 3'26 v 3'27, 4'44 v 4'47, 7'23 v 7'25, etc is reduced to the fundamental difference between them, in this case possibly only a meagre 0.25s difference in 800m extrapolated up

Except you don't know what the 400 and 800m for Hicham and Morceli are because they didn't run those events often and certainly their times are not representative of true ability. Where are you getting those times of 143.00 and 143.25? Where are you getting the 400 ability?

It seems clear to me that you are assigning them times based on their times at longer distances--You are guessing what Hicham can do based on what he ran in the 1500. This is not science. It is educated guessing. You may be right but that is not the issue. You are extrapolating those numbers from 1500 and 3000 times. Then you are using those numbers to extrapolate back the times you say are their maximum capabilities at 1500 and 3000. It is circular logic and reasoning.

I don't think your calculator is worthless at all but it seems to me much better to stick to interpolating between known points (by known points I mean actual races with actual times--not guesses and not times that have had times "adjusted" for wind, rabbit, and oxygen rich environment), not to make up inputs--GIGO and all of that, you know.

If you want to see the official PB of many champions, you cant understand their real strength. For example, Shaheen in 8 years (and 6 years undefeated in the World, with 27 following victories in 3000 steeple) ran 3000m flat once only, and not in shape (7'34" in Doha behind Kipchoge) cominh from typhous. His PB is 7'32" in 2009, when his shape was very far from what he was able to do during his best period (2003-2006). At his best, I suppose he could run under 7'25", may be 7'23", so the difference is between 28" and 30".

I can give you some evaluation about the best specialists I coached, during their best shape :

Christopher Koskei : 7'37" / 8'05" (28")

Richard Mateelong : 7'32" / 8'00" (28")

Julius Nyamu : 7'40" / 8'07" (27")

Wilson Boit Kipketer : 7'35" / 8' (with me, when he bettered the WR he was withou coach, after he was injured and came with me in 1999)

Elijah Chelimo : 7'43" / 8'10" (27")

David Chemweno : 7'40" / 8'09" (29")

Paul Kosgei : 7'35" / 8'07" (32") that's the reason because I moved him to longer distances, and he became World Champion in HM (2002) and bettered the WR of 25 km (2004), running HM in 59'07"

Ladies :

Dorcus Inzikuru : 8'46" official, but in 2005 I think 8'37" / 9'15" (but in that shape, with the situation of steeple of 2009, of sure she could run in 9'08") (41")

Ruth Bosibori : she ran her PB in the same period (August and September 2009) : 8'45" / 9'13" (28")

Durka Mana (Sudanese Record Holder) : 9'13" / 9'47" (34").

In my experience, the difference depends not so much on the technique, but on the ability in using for all the race a high percentage of the personal eccentric strength. Due to this fact, I used some specific training for increasing the eccentric strength, very important when the athletes lands after every barrier and especially in every water jump. I used few times technical exercises, for the simple reason that in Kenya (Iten) we don't have any barrier and any hurdle...

About 400m, I want to say an anedoct :

In the first edition of World Championships (Helsinki 1983) Alberto Cova won 10000m with the last lap in 53"2.

After one week, for fun but at his max speed, Alberto competed in a local race of 400m, as we were curious to see his real speed. He ran... 52"8!

So, we must understand that who is fast in his kick is not the fastest like speed, but who is able to use the highest percentage of his speed.

That's the reason because we speak about "sprint of the aerobic runners".

In other words, between people having the same level of specific endurance, of course the fastest goes to win. But this means that the first step is to reach the same level of endurance of the best athletes, and this is the right direction to work.

What are you doing with your speed, when you are not able to arrive together with the best runners at the end of the race ?

Renato Canova wrote:

About 400m, I want to say an anedoct :

In the first edition of World Championships (Helsinki 1983) Alberto Cova won 10000m with the last lap in 53"2.

After one week, for fun but at his max speed, Alberto competed in a local race of 400m, as we were curious to see his real speed. He ran... 52"8!

So, we must understand that who is fast in his kick is not the fastest like speed, but who is able to use the highest percentage of his speed.

That's the reason because we speak about "sprint of the aerobic runners".

In other words, between people having the same level of specific endurance, of course the fastest goes to win. But this means that the first step is to reach the same level of endurance of the best athletes, and this is the right direction to work.

Pretty much what I said before Renato, but I'd tend to disagree with the last paragraph to some degree. As you are aware, there are some athletes that just have an internal trigger in the last stages of a race. For them, taking it to an equal 'specific fitness' peer who possesses greater foot-speed is never a problem. My example of Jeff Wells and Suliman Nyambui in 1978.

the smartest letsrunner wrote:

Mr. Obvious wrote:

Why are you basing your line of fit on 400m and 800m speed for a guy who never ran the 400 and 800? I just don't get the methodology here. Why don't you base your inputs, I don't know, on races he actually ran?

GROUNDBREAKING!

(Well, for v^0, anyway...)

Ventolin's calculator would be better if he stuck to interpolation, not extrapolation. It would also be better if he used real data, instead of the "adjustments" he applies to everybody's time.

Ventolin^3--

Webb (+Willis +Verbizcas) is running the mile at New Balance indoor this weekend. Care to offer a prediction from your calculator on how each of them will do?

the smartest letsrunner wrote:

ventolin^3 wrote:

rule of thumb is that you add 25 - 30s to 3k flat time to get steeple time, depending on efficiency of technique

therefore a 7'25 ->7'50 - 7'55

well. wrote:

http://www.puma.com/running/athletes/mustafa-mohamed

I think the fact that his 3000 "PR" is INDOORS should tell you something about how often and seriously he's contested the distance outdoors.

You don´t know what you´re talking about. Mohamed is a ferocious racer. During a year he races everything from 30 km XC and HM all the way down to 1500m and he has run innumerable 3000m races. He also only has a 3,44 1500 PB from a highly competitive race. Basically his aother flat PB:s match his 3000 flat PB.

either you are stoopid with no lateral thinking ability or just trolling

there are virtually infinite 400/800 combos i couda used for them, & off course in hicham's & morceli's cases, infinite numbers will throw up ~ 3'26 or 3'27 respectively

however, repeated trial'n'error shows that for hicham ~ 47.75/1'43.00 was the unique combo which fitted best with what was possibly his 1500/3k or even 5k ( if he coud maintain perfect endurance ) ability at his peak

if you can't grasp this concept, i suggest you sign up for remedial classes

err...

yes stoopid

these guys didn't run 400 & in hicham's case no serious 800 ever, so what else are you going to use for guidance other than what they did run ???

the only info we have are some comments on a book-signing he did in oregon post-retirement when a poster here asked him his 400 ability & all was offered was

"i dunno, 48 maybe ?"

& what sounded like a drunken comment from his coach at same signing, when asked about 800 :

"i'm incredulous of hicham, 1'44 i think !"

just disingenous comments - look them up, they are in the archives here

gigo ?

you are a moron if you think that

clearly you didn't see hicham run his 7'23 or his 12'50 & the manner of those runs & how close or not he was from his limit in them - same for morceli's 7'25 & 13'03

renato

i'm not sure if you have any idea of shaheen's 400 or 800 ability at his peak ( but you are his coach )

i'm just throwing out some figures for your critique :

i think based on your comments & watching his 12'48 back in '03, is this line close to what you think he was capable off at his peak ?

51.50/1'48.25 ->

3'31.90

4'47.98

7'23.48

12'43.31

26'32.25

ventolin^3 wrote:the only info we have are some comments on a book-signing he did in oregon post-retirement when a poster here asked him his 400 ability & all was offered was

"i dunno, 48 maybe ?"

& what sounded like a drunken comment from his coach at same signing, when asked about 800 :

"i'm incredulous of hicham, 1'44 i think !"

here is actual post ( above was from faded memories ) :

Hicham said 49-50 seconds for 400m and 1:45 for 800m. I then asked his coach the same questions-he said he has run 48 for 400m and 1:44 high for 800m in training. He said he would do 4X800m at 1:48 each with 3min rest between them in training. He said he thought he could probably run 1:43 low at 800

conflicting info from athlete & coach, but they are from the horses' mouths

How in the world did a thread that considered Alan Webb running 7:20 get over 200 posts?

Rather than having any talk of when Alan Webb will finish 4th or worse in a championship race People started talking about runners that actually might come close to 7:20 or have been close.

Looking at what Shaheen was able to do in training (for example, 2 sets of 5x300m with 50" recovery and 2'50" between every set, average of the first 5 was 38"5, times of the second set 39" / 37"2 / 37"3 / 37"2 / 37"0, in 2006, during a special block that saw him, in the morning of the same day, running 4x1600m with 4'30" recovery in 3'56" / 3'59" / 3'58" / 3'57"), I can say he never reached his best, also during the WR. Infact, he was very much stronger in 2006, but there were external reasons (at first, very bad pace, cold and wind in Zurich) not allowing him in bettering his WR of two years before.

I can suppose that, in that period (July / August 2006), Shaheen could run 48"5 / 1'44"5 / 3'28" / 7'23" / 12'42" and 7'51" in steeple. He could also run about 26'40" in 10000m, looking at the fact that in 2005, one week after winning Golden Gala in Rome in 7'56", ran 35 km in altitude in 1 hr 51', arriving 5' before the Kenyan marathon runners of the Kenyan Team for WCh.

You can see if these estimated data can find a good confirmation from your calculator.

But, in any case, this is "virtual" athletic, and at the end of the game, in the all-time lists are the official performances remaining, only.

ventolin^3 wrote:

I don't think your calculator is worthless at all but it seems to me much better to stick to interpolating between known points (by known points I mean actual races with actual times--not guesses and not times that have had times "adjusted" for wind, rabbit, and oxygen rich environment), not to make up inputs--GIGO and all of that, you know.

gigo ?

you are a moron if you think that

Eldrick, I really think you need to heed his advice. Interpolations from real data are fine, but extrapolations are full of folly and error. If you insist on extrapolations then you MUST use the most conservative estimates for those data points.

Does that make sense to you?

Renato Canova wrote:I can suppose that, in that period (July / August 2006), Shaheen could run 48"5 / 1'44"5 / 3'28" / 7'23" / 12'42" and 7'51" in steeple.

are you serious ?

i did see him run his 3'33pb on tv & even if he wasn't at his peak that never looked like a 3'28er in sheep's clothing - a coupla secs better maybe one day, but 3'28 !?

also 48.5 speed isn't far off that of hicham ( from his coach )

well, your his coach & you know him best...

Thank you for your response Malmo, it seems the top guys are obviously fast but the talk of elites running 47 400's seems slightly OTT.