Womens Oly Trials Marathon Standards

Link wrote:

Therefore, the A and B standards for the trials ought to be the A and B standards for the games.

2. I think that there's a lot of negative publicity that comes from runners finishing in the top three at the trials but not in possession of the Olympic standard.

Yet if a B standard runner wins (or is second or third behind runner(s) with the A standard) and doesn't run faster than the A standard in the process then the same issue will persist. Your reasoning actually dictates that the A standard for the games be the only standard for the OT. I don't have a problem with this concept, just wanted to point that out. Could runners who don't already have the A standard run the necessary time at the OT? Certainly, but at the same time why not insist that said ability be proven prior to the OT? Shouldn't be a big deal for those who are ready to compete on the world stage.

You're right. The exception would be if we didn't have three who had made the A standard. But, otherwise, yes. The games A standard ought to be the trials standard.

Please, stop calling it the "Olys." It's called the Olympics for a reason, so use it's full name. Would it pain you that much? Saying Oly in place of Olympic just sounds ridiculous.

If that is your biggest beef take a hike! I respect the Olympics and it's name as much as any athlete. Between Letsrun grammer correction and petty let's get pissed at my verbage, it might be nice to do something better than pick on that!

"If you think that 2:47 for a woman is comparable to a 2:22 for a man, you're terribly misinformed. Even if there were almost as many men qualifying with a sub-2:22 as there were women qualifying with a sub-2:47, it doesn't follow that the two standards are equally soft."

Why am I misinformed? It actually does follow, as a matter of fact, that if the same number qualify, then the two standards are equally soft, at least in the statistical definition. I am using a comparison of the top runners from two very different sets of bell curves, those of men and women elite runners. The women's curve has a much more positively skewed distribution, or a much longer tail, if you will, that covers a much wider range of marathon times.

"You've totally missed the point of my statement that a 2:40 standard might increase not only the number of sub-2:40 women, but perhaps also the number of sub-2:30 women. In general, when you increase the number of competitors at one level, you also get an increase in the number of competitors at a somewhat higher level, although the effects are obviously somewhat more attentuated. Whether and to what extent a 2:40 standard might generate more sub-2:30 competitors among U.S. women, I have no idea."

I don't think I have missed the point, because you just re-stated it exactly as I interpreted it. Your statement is total conjecture: if you increase the number of 2:40, it does not necessarily (or logically) mean that you will increase the number of 2:30 women. Should we set the standard at 2:20 and hope that we get more women to run sub-2:10s? Ridiculous, yes, but so is your statement. You should have stuck to your last statement: "I have no idea." That is the only logical thing that you've said.

Lastly, for someone who got his underwear in a wad over having his motivations questioned and being labeled a hater, I ask you to consider why you attempted to "guess" the motivations of the 2:19-2:29 men in their argument to keep a B standard.

And, by the way, it's not HE. It's SHE. Why do you assume that an anonymous poster is male. You seem to make a lot of incorrect assumptions.

As for your last point: I didn't assume that you were male. In fact, I realized that you could be male or female. I used the word "he" because, in the English language, it has traditionally been an acceptable pronoun for a person whose gender is unknown. Perhaps you disapprove. If so, feel free to call your local lexicographer to complain. (I also realized after I wrote the message that I could have reworded it pretty easily to make it more gender-neutral, but I'm not aware of any way to edit messages that have already been posted.)

As for the other points:

"It actually does follow, as a matter of fact, that if the same number qualify, then the two standards are equally soft, at least in the statistical definition."

A statistical definition of "soft"? I'd love to see that, especially the part that defines "softness" by the number of people on one side of a vertical line on any distribution curve. I think most of us understand the general considerations that go into judging a standard to be soft, and while one can reasonably debate comparisons of various standards, I don't think that there's much debate among knowledgeable observers that a 2:22 for a male runner is a tougher standard than a 2:47 for a female runner. I also don't think I'm going to get you to concede that, and you're not going to get me to concede otherwise, so I don't see much value in discussing that further.

"if you increase the number of 2:40, it does not necessarily (or logically) mean that you will increase the number of 2:30 women."

Of course not. I never said otherwise. It's a purely empirical point as to whether a more stringent qualifying standard will lead to more performances that exceed that qualifying standard or some superior level of performance. And it would obviously not be correct to say that any more stringent qualifying standard would produce more performances at a superior level of performance.

"Lastly, for someone who got his underwear in a wad over having his motivations questioned and being labeled a hater, I ask you to consider why you attempted to 'guess' the motivations of the 2:19-2:29 men in their argument to keep a B standard."

What I said was that 2:19-2:29 men were "hardly a disinterested group." That was a correct statement. But if a 2:25 marathoner joined the discussion to argue for maintaining a more lenient standard, I believe that the most civil and fruitful response would deal with the merits of his arguments rather than calling him a "hater" (of faster runners or people who set high standards, I guess) or "angry" or "bitter" or whatever other psychological state I might choose to ascribe to him. I don't think that it would be a stretch to guess that you're angry at me, and I suppose I could label you a "hater," or accuse you of getting "your underwear in a wad," or speculate about how fast you run a marathon and tell you to "grow a pair" and train harder, but I believe it's more civil and fruitful to discuss the merits of your arguments than to talk about your motives for making them and perhaps believing them.

Avocados Number wrote:

I don't think that there's much debate among knowledgeable observers that a 2:22 for a male runner is a tougher standard than a 2:47 for a female runner.

You could have just stopped after the third word in the above quote if all you're willing to do is monotonously repeat your view while continuing to fail to explain why you think that way, you condescending prick.

airbear wrote:

"If you think that 2:47 for a woman is comparable to a 2:22 for a man, you're terribly misinformed. Even if there were almost as many men qualifying with a sub-2:22 as there were women qualifying with a sub-2:47, it doesn't follow that the two standards are equally soft."

Why am I misinformed? It actually does follow, as a matter of fact, that if the same number qualify, then the two standards are equally soft, at least in the statistical definition. I am using a comparison of the top runners from two very different sets of bell curves, those of men and women elite runners. The women's curve has a much more positively skewed distribution, or a much longer tail, if you will, that covers a much wider range of marathon times.

No, you're completely wrong. The total number of female competitors *IS* important. We can argue about the shapes of the curves, but I think we all agree that there are more male competitors at the upper level, for whatever reason.

So why does the number of competitors make a difference? Say you have a distribution f(x), and double the number of competitors, the new distribution will simply be 2*f(x). To get the same number of qualifiers from the distributions f(x) and 2*f(x), you have to sum/integrate the two distributions over different ranges. This means there will have to be different cutoffs.

In other words, more competitive populations will have relatively faster qualifying standards, which makes intuitive sense.

Most people think it's more fair to compare world records than to compare say the 100th fastest times, since most believe that the world records tend to approach some limit of human achievement, and improvements to them generally become smaller over time. Although it would be cool to see how the distributions match up.

I'm not saying that I think there should be fewer female runners in the trials, but I don't want to take away from the accomplishments of the more competitive males either.

"And it would obviously not be correct to say that any more stringent qualifying standard would produce more performances at a superior level of performance."

Yet you said exactly this. You did. Look back at your post.

"I think most of us understand the general considerations that go into judging a standard to be soft."

Who is this "us" of which you speak? I don't think "we" do use the same criteria for judging.

I am still waiting for you to provide an empirical reason (because this is what you stated that the argument should focus on, and I agree) as to why 2:47 for women is "softer" than 2:22 for men. You simply state this and then say that most of "us," if we "knowledgeable observers," would agree. What makes you, and not me, a knowledgeable observer?

I apologize for outright calling you a hater. However, your not-so-subtle inference that those who do not agree with you are not knowledgeable observers is also an insult.

gorg wrote:

Most people think it's more fair to compare world records than to compare say the 100th fastest times, since most believe that the world records tend to approach some limit of human achievement, and improvements to them generally become smaller over time. Although it would be cool to see how the distributions match up.

I wouldn't know what "most people think" but I do know that due to physiological and societal factors, there will pretty much always be more men grouped within x% of their WR or AR than there will be women grouped within x% of their WR or AR. Men have a much greater distribution of narrow-hipped, light ectomorphs among the general population than women do, to point out one major factor. Lingering paternalistic attitudes, which vary by culture and region but still have an influence in nearly all of them, have served to limit opportunity and nurturing of women's athletic ambitions while they have served to enable and encourage it among men would be another relevant factor.

From another angle, the US stats for 2008 have 45 women running 2:43:19 or better and 35 men running 2:22 or better. However, in 2007 the stats showed a radically different picture, almost the opposite. I see MLDR and WLDR having similar goals but due to differing variables need to take differing approaches to those goals. There is little use in smugly slapping the "soft" (let alone "extremely soft") label on a standard. It is what it is and the committee making the determination had its reasons for going with it. I think those who have at least a decent understanding of the factors that have impact, including the two I've mentioned in the above paragraph, wouldn't really consider it an unreasonable standard in either direction.

While the men have it noticeably better and while things have improved for women significantly even just over the past 4-8 years, the health of competitive marathoning in the US has quite a ways to go yet before it could be considered as robust as it was for men back in the '70s and '80s. Keflezighi and Culpepper could step away and Hall and Ritzenhein could step in to carry the torch and Rupp and McDougal and/or Bauhs could step in after them. With Kastor on the wane and De Reuck all but done and Rhines disinterested in the marathon there is sort of a void left. Sure Goucher has potential, though it is still mostly potential and she is starting to get up there in age, and Flanagan has huge upside, but again that is still just potential, yet little else clearly coming to fruition via the NCAA pipeline. Would Koll be good in the marathon or will she end up being another Leigh Daniel? It's simply indicative of the disparity in depth between the genders.

And just for the record: women, I'm pulling for you at least as much as I am for the men.

Avocados Number wrote:

I don't think...

Stop right there. As a researcher, I don't give a crap what you or anyone else "thinks". I want to see objective evidence to support 2:47 being softer than 2:22.

Lets talk statistics. To linearly compare two populations, they must follow a normal distribution and exhibit equality of variance. As someone already mentioned, the women's side is positively skewed and thus doesn't follow a normal distribution. This hasn't changed the past 7 years, even with improving depth. As such, it is invalid to say that a men's standard "x" minutes/percentage from the men's world record = women's standard "x" minutes/percentage from the women's world record (which is also an outlier). The Mercier comparison is the "only" valid method for comparing the genders, which "normalizes" the two populations by using the rank-ordermethod.

According to the Mercier comparison, 2:47 for a woman = 2:21:22 for a man, so the 2008 men's standard was softer than the women's.

I completely agree with this post. I made the same point in a similar thread that was started before the 2008 Trials.

I have a couple of points to make about the improper use of statistics here. Many of the comparisons between the men's and women's standard make the statistical assumption that both men's and women's marathon performances follow a standard normal distribution (the bell curve), which is quite obviously not the case. As many of you have so subtly and delicately pointed out, there is not the depth at the highest levels of the elite women's field that there is for the men. All assumptions as to why this is the case aside (ranging from valid points that women don't have the same marathon history--first Olympic marathon in 1984--to the inherent laziness of women), it is, empirically, the case.

Because of this positively skewed distribution on the women's side, comparisons between the men's and women's A and B standards cannot follow either a straight linear comparison (i.e. men's standards x or y minutes slower than the world record, so therefore the same should be true for women), or a slightly better but still incorrect use of percentage differences (i.e. men's standards x and y % slower than the WR, so therefore the same should be true for men). The best way to EQUATE the two standards is to simply see how many athletes of each gender qualify for each standard. If more men than women make the A standard, then to EQUATE the two one must either make the men's standard tougher or the women's standard "softer." If more women make the B standard (in 2008, the total fields were not that far off in number: about 140 vs. 166, I think I read in one of the posts), then one would have to drop the women's standard slightly to result in equal field sizes.

That being said, the whole questions seems odd. Why compare the men's and women's standards at all? A better question is: what is the goal of running the Olympic Marathon Trials? The men may have a different answer than the women. Possible reasons are: 1)to choose the Olympic team, 2)to set high standards for elite status, 3)to make money for any number of parties, or 4)to encourage participation at the highest levels. Depending on the answer to that questions, the field sizes will vary.

One final point: why are we comparing the women's standards by using the men's standards as the "gold standard," meaning that the men's standards are the proper way to set time cut-offs for the Trials. Why are we not comparing the men's standards to the women's standards?

airbear, I bestow on you one of my Gold Stars! Your take on the situation & reply to Avocado's # warms my heart.

know more than you do wrote:

According to the Mercier comparison, 2:47 for a woman = 2:21:22 for a man, so the 2008 men's standard was softer than the women's.

The Mercier calculator that you're referring to was based on distributions of men's and women's times in the late 1990s. Even if that were a reasonable methodology for determining how difficult a particular time is to achieve -- a wildly improbable assumption -- the calculator is horribly out of date. You can see that for yourself by plugging various times into the calculator. Do you really think that a 2:20 marathon for a woman is the equivalent of a 2:03:09 for a man?

If you're looking for a more sophisticated and current set of equivalency tables, why not look at the IAAF's own scoring tables for 2008? A woman's 2:47:00 gets the same points as a man's 2:26:47. A man's 2:22 gets the same points as woman's 2:40:18, and a man's 2:19 gets the same points as a woman's 2:36:06. You can quibble with those comparisons if you want, but they're obviously much closer to reality and more authoritative than the very crude and outdated Mercier calculator that you've cited.

The interesting issue isn't whether 2:47 is a softer standard than 2:22. The interesting issue is how to determine optimal qualifying standards. For me, one of the biggest factors, which I haven't discussed in any detail here, is that of "branding." Put very roughly, an Olympic trials standard can be set so low -- whether that's 2:47 or 3:47 -- that it ceases to have any meaningful connection to the Olympics or the selection of the U.S. Olympic team. It simply becomes, at most, a kind of "gold star" that is handed out by the USATF to hundreds or thousands or tens of thousands of people every four years. There may be good reasons for handing out such gold stars, and even having races that are limited to people who have satisfied the standards for a gold star, but I wouldn't confuse that with the process by which the U.S. Olympic team is selected.

The Olympic brand is strong and well-established. In fact, the USOC and the IOC have gone to great lengths to maintain the exclusivity of the Olympic brand. (Within the U.S., the USOC and the IOC have generally enforced their brand against infringement under federal trademark law (the Lanham Act) and, more recently, the Amateur Sports Act of 1978.) Extending the Olympic brand to four-hour marathoners, or three-hour marathoners, or (perhaps) 2:47 marathoners, tends to weaken the brand, even if it has certain beneficial effects. My own sense -- others may disagree -- is that for trials purposes, the Olympic brand should be set quite high, in keeping with the public perception that the Olympics reflect the highest levels of sports achievement, and in keeping with the expectation that the purpose of something called the Olympic trials is to select the creme de la creme. But if it's easier to qualify for the Olympic trials than it is to qualify for the NCAA championships or achieve All-America status, for example, the Olympic brand is probably weakened. For some people, that's a reasonable price to pay, and for others, the whole concept of Olympic exclusivity is offensive to begin with. I'm not in either of those groups.

Avocados Number wrote:

The Olympic brand is strong and well-established. In fact, the USOC and the IOC have gone to great lengths to maintain the exclusivity of the Olympic brand.

This is total hogwash. The Olympic Trials is NOT the Olympics. Are you ignorant enough to think that the word "Olympic" is that powerful and prestigious? Someone better tell my 13 year old nephew who is playing OLYMPIC DEVELOPMENT SOCCER". Do you know how prestigious that is? It is available for any kid in our city that has enough money to pay for the travel. It has nothing to do with developing Olympians. It is just a way to promote their sport.

Avocados Number wrote:

If you're looking for a more sophisticated and current set of equivalency tables, why not look at the IAAF's own scoring tables for 2008? A woman's 2:47:00 gets the same points as a man's 2:26:47. A man's 2:22 gets the same points as woman's 2:40:18, and a man's 2:19 gets the same points as a woman's 2:36:06. You can quibble with those comparisons if you want, but they're obviously much closer to reality and more authoritative than the very crude and outdated Mercier calculator that you've cited.

Obviously? How so?

The Olympic brand is strong and well-established. In fact, the USOC and the IOC have gone to great lengths to maintain the exclusivity of the Olympic brand. (Within the U.S., the USOC and the IOC have generally enforced their brand against infringement under federal trademark law (the Lanham Act) and, more recently, the Amateur Sports Act of 1978.)

You mean like how USATF uses the term for their oh-so-exclusive Junior Olympic competitions?

Avocados Number wrote:

The Mercier calculator that you're referring to was based on distributions of men's and women's times in the late 1990s.... If you're looking for a more sophisticated and current set of equivalency tables, why not look at the IAAF's own scoring tables for 2008?

The Mercier calculator was updated in 2005 and based on the world rankings from 2001-2004.

"The Mercier 2005 scoring tables were determined using the Mercier-Beauregard method (see here) and the world rankings for 2001, 2002, 2003 and 2004. The IAAF 2005 scores were taken from the scoring tables used for the IAAF World Rankings (see here). We should note that the Mercier 2005 scoring tables are the only tables developed for comparing performances between men and women."

From the IAAF scoring tables:

"Due to obvious biological differences, we do not propose to fully compare men’s and women’s performances. Thus, the system contains scoring tables for men’s and women’s events."

know more than you do wrote:

From the IAAF scoring tables:

"Due to obvious biological differences, we do not propose to fully compare men’s and women’s performances. Thus, the system contains scoring tables for men’s and women’s events."

http://www.iaaf.org/mm/Document/Competitions/TechnicalArea/ScoringOutdoor2008_742.pdf

Yes, there are separate scoring tables for the men and the women, because the equivalent performances across events are different for men and women. For example, a 2:19 marathon for a man is considered the equivalent of a 28:46 road 10K for a man, whereas a 2:19 marathon for a woman is considered the equivalent of a 30:21 road 10K for a woman. But the points awarded for the men and women are similarly distributed, so a point total of 1070, for example, reflects approximately the same level of performance within gender for both men and women. Your Mercier calculator acknowledges as much when it provides the 2005 IAAF equivalents, although Mercier comes up with different equivalent performances. As for the accuracy of your somewhat updated (but still out-of-date) Mercier calculator, it now shows a woman's 2:20 marathon to be the equivalent of a 2:02:59.5 marathon for a man. Its updated figures, in other words, produce calculations even more inaccurate than its figures from the late 1990s did, reflecting its inferior methodology.

You don't have to look at the men's table to get an idea of how soft the women's marathon trials standard is. A women's 2:46 marathon is considered the equivalent of a women's 10,000 of 35:17, which is very far from the Olympic trials standard (automatic or provisional) in that event. By contrast, a man's 2:19 marathon is considered the equivalent of a man's 10,000 of 28:41, which I believe is better than the provisional qualifying time for the 2008 trials, and within 15 seconds or so of the automatic qualifying time. Thus, the men's marathon qualifying time, which was very soft for a trials qualifier at 2:22, has been brought back within range of qualifying times for shorter distances, while the woman's marathon qualifying time remains very far from the range of qualifying times in shorter distances. That is what I meant when I said that the women's marathon qualifying standard remains "exceptionally soft" when compared with qualifying standards in other athletics events, for both men and women. One can certain argue that an exceptionally soft standard for the women's marathon is appropriate for some reason, as others have done, but denying that it is exceptionally soft when compared with other qualifying standards isn't credible, as the IAAF tables indicate.

According to the 2005 IAAF scoring tables, 2:19 male marathon = 29:38. This is backed by the 2005 Mercier calculator with a similarly equivalent 29:57 (or 35:18 for a female). For the 2008 IAAF tables to suggest a 2:19 is "now" equivalent to a minute faster for 10K reflects a flawed statistical model. There is no denying that 2:19:00 is a more stringent standard than 2:46:00 (~2:43:48, according to the valid, but yes, somewhat out-of-date Mercier calculator). However, I'm not buying that 2:47 is softer than 2:22 "because you think so", nor is 2:19 "within qualifying range of qualifying times for shorter distances".