Shannon Rowbury entered in Brussels 5000m, will she break the AR?

Frinlay wrote:

A lot of people are failing to realize the 50% guy was joking.

Yep, this is an old Letsrun joke. Responding to the 50/50 guy grants one a badge of idiocy.

Good luck to Shannon. Look forward to seeing how hard she can push herself at 5k.

If the pace is honest, she will 100% break the AR in the 5000m. 3:57 1500m shape will mean AR 5k pace will feel like a jog. She hadn't broken 4 when she ran 14:48.

zero chance !!!!!

hate you shannon wrote:

zero chance !!!!!

I understand being a fanboy, having favorites, and those you root against, but you are pitiful and certainly don't add anything interesting to the discussion. Have a nice day.

Rowbury will break the AR this Friday. She is in great shape and since Ayana is in the field and wants to break the world record, the pace will be fast -- from the gun. This should be an exciting race to watch.

Brave or foolhardy wrote:

As for what the probability "really" is, I think it's like David Friedman said about earthquake predictions: in many situations, "probabilities are a distraction." They give a false sense of precision and accuracy. They don't tell us more than "She could break it, but I don't think she will."

You might find the book Superforecasters interesting. One of the distinguishing traits of those who were better forecasters is that they gave more fine-grained forecasts. The average person has basically three mental settings: it will happen, it won't happen, or I'm not sure. Military intelligence analysts have to fill out reports based on a seven-point scale (somewhat likely, very likely, etc.). But the best forecasters express their predictions as a percentage, and adjust it in small increments as new information is obtained.

And yes, it's possible to assess the accuracy of percentage probability forecasts of discrete yes-or-no events -- as long as you're assessing a body of many forecasts rather than a single isolated forecast. You use something called a Brier Score:

jungleroy59 wrote:

hate you shannon wrote:

zero chance !!!!!

I understand being a fanboy, having favorites, and those you root against, but you are pitiful and certainly don't add anything interesting to the discussion. Have a nice day.

Ouch, yo! Man. Franzie sends his regards.

I'd say she's got a chance, but peak 1500 shape doesn't always translate to break 5K PR in less than two weeks. That said if she doesn't get it, bet on it in 2017. Or Simpson. Or Huddle. That record will be 14:36 or better within a year.

jungleroy59 wrote:

with PRs of 3:57 and "only" 1:59, I'd say her chance at the 5000m AR is about 50/50. I would wager on her.

Some estimates:

3:57 with 1:59 -> 14:40

3:57 with 1:57 -> 15:04

3:55 with 1:59 -> 14:20

3:55 with 1:57 -> 14:45

It might be interesting to see her splits from her 3:57. Does anyone know them? Her true shape might be faster and would help give a better estimate on her 5000 shape.

hold the phone wrote:You might find the book Superforecasters interesting. One of the distinguishing traits of those who were better forecasters is that they gave more fine-grained forecasts. The average person has basically three mental settings: it will happen, it won't happen, or I'm not sure. Military intelligence analysts have to fill out reports based on a seven-point scale (somewhat likely, very likely, etc.)

Using words of estimative probability, Rowbury's chance of breaking the AR is "unlikely."

As a US track fan, I'd love to see an AR and Rowbury probably has the capability on the right day, but it would take better conditions than are likely at Brussels. It has been a long season for Rowbury and moving up from 1500 to 5000 makes it that much harder for her. If Ayana is trying to run around 14:10, the lack of pacers for Rowbury around 14:40 also hurts. If Rowbury has to run her own pace, it makes it hitting the AR that much harder.

stats.gangsta_the_real_1 wrote:

jungleroy59 wrote:

with PRs of 3:57 and "only" 1:59, I'd say her chance at the 5000m AR is about 50/50. I would wager on her.

Some estimates:

3:57 with 1:59 -> 14:40

3:57 with 1:57 -> 15:04

3:55 with 1:59 -> 14:20

3:55 with 1:57 -> 14:45

It might be interesting to see her splits from her 3:57. Does anyone know them? Her true shape might be faster and would help give a better estimate on her 5000 shape.

http://timescalculator.appspot.com/

I don't understand this. Are these estimates suggesting that if two runners are both in 3:57 1500 shape, then the one who can run 2 seconds faster at 800 meters should be 24 seconds faster over 5000?

The Bangkok sportsbook has her at 11-5/7 to 2-7/9

timmy runs for second wrote:

stats.gangsta_the_real_1 wrote:

Some estimates:

3:57 with 1:59 -> 14:40

3:57 with 1:57 -> 15:04

3:55 with 1:59 -> 14:20

3:55 with 1:57 -> 14:45

http://timescalculator.appspot.com/

I don't understand this. Are these estimates suggesting that if two runners are both in 3:57 1500 shape, then the one who can run 2 seconds faster at 800 meters should be 24 seconds faster over 5000?

No, for equal 1500m times, the person with the faster 800m will be SLOWER over 5000m. Or did I read your post incorrectly?

stats.gangsta_the_real_1 wrote:

timmy runs for second wrote:

I don't understand this. Are these estimates suggesting that if two runners are both in 3:57 1500 shape, then the one who can run 2 seconds faster at 800 meters should be 24 seconds faster over 5000?

No, for equal 1500m times, the person with the faster 800m will be SLOWER over 5000m. Or did I read your post incorrectly?

That is what I assumed, I just totally read your chart backwards.

Where do you guys watch these races. The only place I've found is youtube and it'll be titled "Shannon rowberry breaks AR" or "shannon rowberry gets devastated" or some other spoiler filled title

Jimmy21 wrote:

Where do you guys watch these races. The only place I've found is youtube and it'll be titled "Shannon rowberry breaks AR" or "shannon rowberry gets devastated" or some other spoiler filled title

This article lists lots of options. You might find some that are available even though the Olympics are over:

Jimmy21 wrote:

Where do you guys watch these races. The only place I've found is youtube and it'll be titled "Shannon rowberry breaks AR" or "shannon rowberry gets devastated" or some other spoiler filled title

PS If you end up looking through a list of European broadcasts, remember track is called "athletics" in the rest of the world.

She'll get the record if she tries. It's soft. The 5k record compares very unfavorably with times currently being run by African athletes, and both Rowbury and Simpson have 3K times that convert to faster than the AR. Huddle probably was capable of a faster 5K, but she's moved up in distance and doesn't show any sign of going back to better her time. I think the 5K requires a bit too much speed for Flanagan these days.

douglas burke wrote:

The 5000 looks softest to me, Rowbury, Simpson, Huddle and Flanagan are all capable of sub 14:40.

hold the phone wrote:

You might find the book Superforecasters interesting.

I'm a Tetlock fan, but haven't read this one yet. I'll put it on my list.

I think Tetlock and Freedman are making the same point. They both think forecasts are bad and can be improved. Freedman, who was a mathematical statistician, worried that people would treat statistical techniques as magic that could turn ignorance into exact knowledge. As an example, he looked at the statistical models underlying the USGS earthquake predictions for the Bay Area. He argued these models were so bad that any probabilities based on them would be worthless, and wouldn't add anything useful to more qualitative statements. Hence his claim that probabilities could be a distraction.

What is the chance of an earthquake?