Sub 2 hour Marathon - Do the Math?

No one has or will run under 2:08 without EPO.

Bad Wigins wrote:

Jebron Lames wrote:

It does if u use the date in that format. In MS excel it converts the date to a # of days I think is the unit based on starting at 1/1/1900 or whatever it is, you also have to convert your answer (y) to hh:mm:ss from days.

ok so sometime around 2000 the # of days was 36,500.

y = .357(36,500)^-.119 = 0.102 days

That's 2.448 hours or 2:26. Almost 20 percent off.

Around 1970 x was 25,550.

y = .357(25,550)^-.119 = .107 days = 2.56 hours = 2:33

Your formulas aren't withstanding scrutiny.

See the second response in this thread, I originally posted the wrong formula, it's .3057. I don't have the file in front of me right now but I recall doing similar testing and being around single digit % off.

That would be fairly accurate for 2000, off by about 4 percent in 1970, and 6 percent in 1940. What's 6 percent amount to in years?

Solving for x:

y^8.40 = .3057^8.40/x

x = (.3057/y)^8.4

then for two values, y1 and y2 = 1.06y1,

x1 = 1/y1^8.4 x .3057^8.4

x2 = 1/y1^8.4 x (.3057/1.06)^8.4

and x2/x1 = (1/1.06)^8.4 = 0.61

If y2 = .94y1, then

x1 = 1/y1^8.4 x .3057^8.4

x2 = 1/y1^8.4 x (.3057/.94)^8.4

and x2/x1 = (1/.94)^8.4 = 1.63.

So an error margin of +/- 6 percent for y corresponds to + 63/-39 percent for x. Your prediction would be in a confidence interval of 100 years.

Unless you find a better approximation than a power function. You could use polynomials. But then again, there's no reason to suppose the marathon WR will continue to follow any pattern at all. For example, PED's could be stamped out for good and there might be no more sub-2:07's ever again.

ASDL;JSL;ADJ wrote:

No one has or will run under 2:08 without EPO.

How does a drug which thickens the blood and lowers heart rate deliver more oxygen?

Do you know if you have the right functional form? If not, there are no valid statistics on goodness of fit and the regression coefficients are of unknown property. Do you select the dates a priori or after fitting the data in some manner; the most common mistake is to look at the data from breaks and then use those breaks to define the periods for the statistical fit. Were the earlier marks made on the same sort of courses, and if not, have you made adjustments for how much the times would differ?

This is another typical stupidity from somebody not knowing anything what he's speaking about.

In the all-time lists, at the end of 2012 2:08 was the performance no. 245.... and at the beginning of this year already 3 new athletes joined this big group.

Did everybody take EPO ? Or it's better to think that 2:08 is today a normal performance, possible for a lot of good athletes (it's enough to be able running 10000m in 28:40, if there is specific mental and physiological attitude for the full marathon, of course with a correct type of training), without being great champions ?

Typical example of limited minds, producing mental limits that don't allows athletes in performing at their best.

Renato Canova wrote:

Typical example of limited minds, producing mental limits that don't allows athletes in performing at their best.

This, exactly.

As i said before, there are only 2 regressions I did. One based on all data points from 1908 to the most recent record. There was no picking/elimination of specific dates to better fit the curve. The second regression was the latest 10 years just to see how it looked since the marathon has been increasing in popularity for younger athletes. In that case I just picked those recent points from 2002 to now. No fancy adjustments for courses etc. I just wanted to look at data and logically a curve that decays closer and closer to a single y value is what you would expect... like an asymptote. There are so many variables in running a marathon I almost feel like it is more rare for improvements to halt vs. continue as time goes on.

But that's not useful for predicting x at a particular y near the asymptote, because delta x is much greater than delta y.

ukathleticscoach wrote:

' 6/25/2028 and 10/6/2051'

That's really narrowed it down

How narrow of a window do you expect it to be? Sometimes world records last for decades. The 200m world record was set in 2009. If it weren't for one man's existence (Bolt), it would still be 1996, 13 years different from what it is. You can't narrow them down much.

Maybe somebody should develop a function for Rudisha's 800 time or Bekele's 10k time. I'm sure those weren't supposed to happen either.

To say it will never happen is very close minded. One observation? Selective breeding. You may notice more and more professional athletes are marrying other professional athletes. That is setting up for some very dominant future runners down the road.

In the past 11 years the WR has dropped exactly 2 minutes.

The 11 years before that? 2 minutes 36 seconds.

There isn't a consistent pattern in reduction progression either before that.

In my opinion: within 25 more years and we'll see sub 2 hours.

The formulas you people are coming up with would probably only apply if it was one athlete you were looking at. Also the training methodology has to remain the same and the marathon course has to be the same and so many other things like incentives remaining constant. This is false assumption. We have two athletes improving their one-yr old 2:04 PRs by more than a minute. E. Mutai improved in 2011 by more than a minute to run 2:04:40. There is G. Mutai and Tsegay Kebede. Then there is Moses Mosop debuting the distance in 2:03:06. Two other athletes debuting with 2:04:23(A. Abshiro) and 2:04:16(D. Kimetto). With all these statistics, you cannot help but believe that this generation is capable of anything including sub 2hrs. The hunger to run faster times has never been like it is now. In the next five years alot will happen including a number of women running in the 2:16 range and possible 2:15.

How does a drug which thickens the blood and lowers heart rate deliver more oxygen?[/quote]

Long answer: the erythropotein chemical soaks up oxygen, the more epo, the more oxygen you get (i'm not a scientist but that's the way i understand it)

Short answer: Lance Armstrong.

ASDL;JSL;ADJ wrote:

How does a drug which thickens the blood and lowers heart rate deliver more oxygen?

Long answer: the erythropotein chemical soaks up oxygen, the more epo, the more oxygen you get (i'm not a scientist but that's the way i understand it)

Short answer: Lance Armstrong.[/quote]

The erythropotein chemical soaks up oxygen?????

Yes, you're not a scientist, or even an intelligent person at all.

The erythropotein chemical soaks up oxygen?????

Yes, you're not a scientist, or even an intelligent person at all.[/quote]))))))))))))))))))))))))))))

Not a scientist? Correct.

Not an intelligent person? Incorrect.

So you're saying that the extra amount of EPO does nothing?

Try an answer, don't expose yourself as a stooge for druggies.

Synthetic EPO increases red blood cell count. How does having more red blood cells than normal increase oxygen delivery to the muscles when it thickens the blood and lowers heart rate?