Kara Goucher Vs. Deena Kastor in NYC 2014 Marathon

According to your source (

http://arrs.net/TB_Mara.htm

), ARRS lists the following Race Time Bias' (RTB):

Paris..... -90.5 sec

Berlin.... -81.0 sec

NYCM... +82.8 sec

Based on Deena's Lifetime Bests (1:07:34/2:19:36), her HM/Marathon differential is HM x 2 + 4.5 min..

Based on her Philly Half in warm/humid conditions (68+ with a dew point of 63), I believe that puts her performance "on par" with the hills and potential wind in NYC-- no time adjustment necessary. Take 1:09:36 x 2 + 4.5 = 2:23:42, if she nails it, no injuries, and nothing crazy about the weather. A less than perfect marathon, I say she runs no slower than 2-4 min. off of this (2:25:42-27).

Sounds crazy, but I wouldn't put it past her to have a gutsy performance (a la Meb) in NYC.

fun with stats...?You mean ignore the stats altogether? lol

fun with stats wrote:

...Philly Half in warm/humid conditions (68+ with a dew point of 63), I believe that puts...performance "on par" with the hills and potential wind in NYC-- no time adjustment necessary....

I just want a tight butt shot of Kara in those tight, tight shorts??? Is that too much to ask?

But Deena will finish in front of Kara. Kara is there for Oiselle, to PROMOTE, and nothing else. Get a clue people.

There was a great one from Osaka, "butt" someone removed it:(:(:(:(:(

boomtown rascals wrote:

fun with stats...?

You mean ignore the stats altogether? lol

fun with stats wrote:

...Philly Half in warm/humid conditions (68+ with a dew point of 63), I believe that puts...performance "on par" with the hills and potential wind in NYC-- no time adjustment necessary....

?

For those that would like to try a NYCM prediction for Kastor, based on a half-marathon time-extrapolation, we can reasonably do so using historical information that actually includes and accounts for the context of the actual NYCM course bias.

2006 was the last year (I think; update me if this is incorrect) Kastor ran NYCM in a time of 2:27:54. That same year, Kastor ran her half PB of 1:07:34 on a flat Berlin Half course (similarly flat to Philly). Thus, biasing for Kastor's actual efforts on the hilly, challenging (and notoriously the slowest of the Marathon Majors) NYCM course, we see Kastor runs NYCM at 218.8% slower (i.e., 147.9/67.6) than her fast flat half marathons (that are run in the same year she runs NYCM). Therefore, this method projects Kastor to run the 2014 NYCM in 2:32:17 (i.e., 2.188 x 69.6min), based off her 2014 Philly Half.

Now granted, one may claim the Berlin Half in 2006 was not as hot/humid as the Philly Half in 2014. But remember, was pre-childbirth, when Kastor was eight years younger.

I like Deena Kastor; tremendous talent.

Cheerio.

Again, I put the hot and humid Philly performance on par with the hilly and potentially windy NYC course- no "course bias" needed. I say she's going to have a Meb-like performance, barring anything crazy happens.

Okay, you say the warm/humid conditions in Philly should eliminate the NYCM time bias. Subtracting the 160-170sec NYCM time bias (taken from ARSS.net site) from the projection we just arrived at, 2:32:17 (which was based on the ratio of Kastor's historical NYCM to fast Half times) pegs Kastor to go 2:29-2:30 at NYCM, which is what we projected originally in this thread.

Sciatica Road wrote:

According to your source (

http://arrs.net/TB_Mara.htm

),

No, that's not actually the source I linked. That is a different one you linked.

Sounds right as NYC isn't a super fast course. Who is McMillian and why does everyone use his tables? Our coach always doubled our half and added 5 min.

Sciatica Road wrote:

Okay, you say the warm/humid conditions in Philly should eliminate the NYCM time bias. Subtracting the 160-170sec NYCM time bias (taken from ARSS.net site) from the projection we just arrived at, 2:32:17 (which was based on the ratio of Kastor's historical NYCM to fast Half times) pegs Kastor to go 2:29-2:30 at NYCM, which is what we projected originally in this thread.

I don't think you understand what I'm saying? You don't need to adjust anything-- simply take her HM x 2 + 4.5 min. Even if you played it safe by another 2-4 min., she'd run 2:25-28. No way does she go 2:29-32 unless it's hellish conditions or her gut blows up.

I agree (see above). The only calculation is 2xhalf + 5 or in your case 4.5.

Bleu wrote:

Who is McMillian

From his site:

"Greg McMillan is one of the best and smartest distance-running coaches in America."

---Amby Burfoot / Runner's World Magazine

why does everyone use his tables?

It's convenient and is comparable to other predictors. This link provides four predictions side-by-side:

Our coach always doubled our half and added 5 min.

That is just a very general rule-of-thumb.

Gilbert and Daniels mathematically described the extrapolated data from lab measurements on runners (ref. "Oxygen Power", though personally have never read it; note "Daniels Running Formula" does not actually contain the formula.).

The Mcmillain Calculator may (?) be Mcmillain's own tweak of extrapolated data from his measurements:

fun with stats wrote:

I don't think you understand what I'm saying?

My understanding is that you are being optimistically biased in your estimation.

You don't need to adjust anything-- simply take her HM x 2 + 4.5 min.

The least you could have done was a percentage extrapolation (as I demonstrated) rather than a simple summation of some amount time.

Even if you played it safe by another 2-4 min., she'd run 2:25-28. No way does she go 2:29-32 unless it's hellish conditions or her gut blows up.

As I demonstrated in this thread, I came up with 2:29-2:30 using two different methods.

Deena's not running 2:24 but I agree she's finishing ahead of Kara.

What, the mail order coach who lost his training group? Are you kidding me?

Sciatica Road wrote:

fun with stats wrote:

I don't think you understand what I'm saying?

My understanding is that you are being optimistically biased in your estimation.

You don't need to adjust anything-- simply take her HM x 2 + 4.5 min.

The least you could have done was a percentage extrapolation (as I demonstrated) rather than a simple summation of some amount time.

Even if you played it safe by another 2-4 min., she'd run 2:25-28. No way does she go 2:29-32 unless it's hellish conditions or her gut blows up.

As I demonstrated in this thread, I came up with 2:29-2:30 using two different methods.

Why does it matter if you use minutes or a percentage? Minutes or a percentage are both linear methods. The "best of the best" marathoners can hit their recent HM x 2 + 4-6 min. (or even faster, ~Desiree, Paula)-- Deena is one of them. You have her pegged to hit HM x 2 + 10+ min. instead of 4-6 min, and I say no way. I think she has something to prove and is going to nail it.

fun with stats wrote:

[quote]Sciatica Road wrote:

Why does it matter if you use minutes or a percentage? Minutes or a percentage are both linear methods.

IF the minute offset were to remain a constant for all runners, regardless of ability, then it would be a linear definition. But it doesn't; it increases as the runners get slower, so the resultant function is non- linear.