Popov says quit whining about the LZR

The numbers are interesting, but I think that time progressions -- of WRs, of the 10th-ranked time, or whatever -- is probably the best way to get at this sort of thing. One reason is that raw numbers ignores the fact that some events have been added along the way (eg the non-free 50s) and that SCM records have become a little bigger deal. But even with time progressions, it is still difficult (though not impossible) to make causal connections. That's just the nature of this kind of study. One of the articles that Lindsay was probably referring to on Swimnews is www.swimnews.com/.../displayStory.jhtml Here is an excerpt: I took 450 performances of swimmers wearing the LZR Racer, including world record setters down to people finishing in the top 20 of their event. More than 400 of them were clustered in an approximate range from 1.6% to 2.3%. I called a professor and friend who spends his life looking at probabilities. I put the small test results to him. Statistically significant? "Without a shadow of a doubt ... if you have that kind of result in a medical experiment, you'd be looking at 'case proven.'" Setting aside the lack of specifics (the sort of thing that drives me bonkers...what was the comparison that was done? LZR vs previous years' performance? PRs with and without the LZR? LZR vs unshaved times?), what the stats guy was saying is that there is little likelihood that the difference -- whatever it was -- is due to random variation. But the mention of a medical experiment is misleading. There is a big difference between a typical medical experiment and the results here because other variables are not being controlled, so there is still the problem of attribution. In other words, what the stats guy is saying is "the difference is real" and what the author is saying is "the LZR caused the difference," and those are not at all the same. It is based on the critical assumption that the only real difference between the two sets of numbers is the suit. Since the other variables weren't controlled, you are at risk when you make that assumption ("correlation does not imply causation" and all that). This is a well known issue and is covered in any beginning text on epidemiology; google "Bradford-Hill criteria" if you are interested in knowing more. But that's just me as a scientist (you probably know the drill: accept the null hypothesis -- "The LZR does not improve performance" -- until rigorous/definitive peer-reviewed proof shows otherwise). Most people might want that level of proof for, say, product safety or drug effectiveness or to disprove an established scientific theorem. As a consumer, the bar need not be so high. Then it is about costs/risks and benefits (science can help quantify the latter, of course). This is why, despite lack of complete certainty, Schubert states that an elite swimmer would have to be crazy to give up what could potentially be a decisive advantage. If my livelihood depended on it, I would almost certainly wear the suit. As a masters swimmer, I think it is much more complicated. Fortress made the comment that only the overly-principled or under-funded would not get the suit, but I beg to differ. Maybe I am wrong, but I am not at all sure that my enjoyment of the sport would be greatly enhanced by, roughly, an additional $550 per year even if I gained an additional 1-2% on my times (and I am not sure that I would). The cost isn't just monetary. I read the thread about the FS Pro and the trials and tribulations in putting one on, the risk of ripping it, keeping it dry -- all things that may be worse with the LZR -- and all I think is, "thank God I don't go through all that before my races." The ONLY thing the suit potentially has going for it is speed, in ALL other respects it is worse than its predecessors. My enjoyment of the sport also depends on some balance. Travel to meets, practice time, etc: my wife puts up with this. I am sure she and I would appreciate the money for a longer/better vacation more than I would appreciate a LZR. Now if a technical suit existed that was comfortable, durable, reasonably priced AND was faster than briefs/jammers, you can sign me up for that! We'll probably get there at some point.

The numbers are interesting, but I think that time progressions -- of WRs, of the 10th-ranked time, or whatever -- is probably the best way to get at this sort of thing. One reason is that raw numbers ignores the fact that some events have been added along the way (eg the non-free 50s) and that SCM records have become a little bigger deal. But even with time progressions, it is still difficult (though not impossible) to make causal connections. That's just the nature of this kind of study. One of the articles that Lindsay was probably referring to on Swimnews is www.swimnews.com/.../displayStory.jhtml Here is an excerpt: I took 450 performances of swimmers wearing the LZR Racer, including world record setters down to people finishing in the top 20 of their event. More than 400 of them were clustered in an approximate range from 1.6% to 2.3%. I called a professor and friend who spends his life looking at probabilities. I put the small test results to him. Statistically significant? "Without a shadow of a doubt ... if you have that kind of result in a medical experiment, you'd be looking at 'case proven.'" Setting aside the lack of specifics (the sort of thing that drives me bonkers...what was the comparison that was done? LZR vs previous years' performance? PRs with and without the LZR? LZR vs unshaved times?), what the stats guy was saying is that there is little likelihood that the difference -- whatever it was -- is due to random variation. But the mention of a medical experiment is misleading. There is a big difference between a typical medical experiment and the results here because other variables are not being controlled, so there is still the problem of attribution. In other words, what the stats guy is saying is "the difference is real" and what the author is saying is "the LZR caused the difference," and those are not at all the same. It is based on the critical assumption that the only real difference between the two sets of numbers is the suit. Since the other variables weren't controlled, you are at risk when you make that assumption ("correlation does not imply causation" and all that). This is a well known issue and is covered in any beginning text on epidemiology; google "Bradford-Hill criteria" if you are interested in knowing more. But that's just me as a scientist (you probably know the drill: accept the null hypothesis -- "The LZR does not improve performance" -- until rigorous/definitive peer-reviewed proof shows otherwise). Most people might want that level of proof for, say, product safety or drug effectiveness or to disprove an established scientific theorem. As a consumer, the bar need not be so high. Then it is about costs/risks and benefits (science can help quantify the latter, of course). This is why, despite lack of complete certainty, Schubert states that an elite swimmer would have to be crazy to give up what could potentially be a decisive advantage. If my livelihood depended on it, I would almost certainly wear the suit. As a masters swimmer, I think it is much more complicated. Fortress made the comment that only the overly-principled or under-funded would not get the suit, but I beg to differ. Maybe I am wrong, but I am not at all sure that my enjoyment of the sport would be greatly enhanced by, roughly, an additional $550 per year even if I gained an additional 1-2% on my times (and I am not sure that I would). The cost isn't just monetary. I read the thread about the FS Pro and the trials and tribulations in putting one on, the risk of ripping it, keeping it dry -- all things that may be worse with the LZR -- and all I think is, "thank God I don't go through all that before my races." The ONLY thing the suit potentially has going for it is speed, in ALL other respects it is worse than its predecessors. My enjoyment of the sport also depends on some balance. Travel to meets, practice time, etc: my wife puts up with this. I am sure she and I would appreciate the money for a longer/better vacation more than I would appreciate a LZR. Now if a technical suit existed that was comfortable, durable, reasonably priced AND was faster than briefs/jammers, you can sign me up for that! We'll probably get there at some point.