Although the people who frequent the forums are not necessarily a representative slice of the USMS membership, I have been moderately surprised at the positive response to your table. I might have predicted the opposite, since some people might not want to be reminded of their age-group days or don't want to think of masters being as "serious" as age-group swimming.
Chris - can you explain your rating calculator like you would to a second grader please? Say, for instance you have an 85% rating on an event, what does that mean to those of us with liberal arts educations?
So the next question is, what is the "ideal" record? Ultimately it is a number (derived from WRs and ARs) that is supposed to describe the effect of age on performance in a given event.
If you plot each record time (in sec) against the median age of the age group of the record, you get data that trend upward. I then "fit" a function to those data to describe the trend. The fitted function, which averages out the effects of "soft" and "insanely fast" records, describe the "ideal" record for a given age. Here is a link to an example; the line describes how the "ideal record" depends on age.
If you make the assumption that the records are, generally, equally hard across all events and across genders, then the rating allows you to compare performances in different events, courses, ages and genders. It can even allow you to do time conversions between courses with an age-related component to it (I have done this for myself; the conversions are pretty consistent with other converters I've seen out there).
So you can start trash-talking anyone you want, not just men in your own age group. Which ultimately is the point.
(Actually, mostly the point is for personal goal setting. As age sets in and we get slower, we can possibly take some comfort if our ratings remain the same or improve; it means we are beating the aging curve of record-holders.)
The ratings on the VA site use the 2007 (pre-LZR) records. At some point soon I would like to offer two versions: one using the most recent pre-LZR records, and one using the current records. So if we lose the suits -- or someone elects to swim without them, if they remain legal -- you can make comparison between the two types of swims.
Last edited by Chris Stevenson; December 15th, 2009 at 05:57 PM.
To Mr. Stevenson and Mr. Stuart,
What do you think the odds are that the two best and most active, mathematically proficient, swim-data curve fitters would both be from the south and have surnames beginning with St?
Let me quickly reference yet another swim-math motivator I have personally found useful, especially before discovering St. Chris's formula.
St. Chris, for his part, knows of this other formula, but it is possible that St. Steve does not. It's an age regression swimming time calculator that Phil Arcuni posted a number of years back on this forum, that is, in the pre-speed suit era. Phil included the ability to also check the "Finnish formula" option, which is basically the same thing as the American version, but invented by a race of people who live on reindeer milk and cell phone dividends much of the year, and thus arguably more accurate.
One thing the Phil Arcuni and Chris Stevenson approaches have in common is that they are adjustable by each year of a swimmer's age, not just the 5 year age group. Hence, a fellow such as myself, in the dotage of his 57th year, can still "improve" over his 55 year old performances despite the absolute times getting worse.
I do think that the next major need for motivation, at least for me, will be coming to grips with what I suspect will be a relatively huge uptick in times in the coming post-speed-suit era.
Would either of your fellows care to turn your keen mathematical minds to the job of predicting "equivalent times" with and without such suits--and do so before we need to wait several years for new data to filter in upon which your new curves can be fitted and hung?
Or, barring that, what do you predict, generally speaking, will be an average time change before and after B70s go to jammers? Obviously, for St. Chris and his hairless ilk, the differences have not been too great. But for us lumpkin types, what seems reasonable? Would 2 seconds per 100 be too much of an expected increase in times--or too little?
At last year's Colony Zones championships, for instance, I swam a B70 aided 52.86 in the 100 free. At this year's meet, all other things being equal, should I be delighted if I can break 55--or should I lapse into catatonia?
I found both the motivational times chart and Chris's calculator to be interesting and helpful. In both I end up strongest in fly, breast and IM's,
100 to 200 best distance. I don't train the mega yardage I did in college (thank goodness!) and I am weaker now in the long free stuff compared to some of the women I swim against. And I'm not a drop dead sprinter - just not quite enough speed.
It's all food for thought and I appreciate the work that went into creating the charts and the calculator.
For the men's 100 free, for example, the actual time was 0.7 sec faster (with the suit) than predicted; for men's 100 fly it was 1.0 sec.
But keep in mind that this analysis lumps the effects of all the suits of that time (B70s, LZRs, etc). Also this is for elite athletes who, um, have a different body profile and stroke efficiency than the typical masters swimmer. It is also for LCM and the effects might be different on the shorter courses.
Any predictions that I or someone else can come up with will, at best, be good on the average. But I have heard too many stories of drastic time drops to dismiss them entirely, so the effects may also be spectacularly individual. One (skinny) woman I know had more or less hit a recent plateau in an event, donned a B70 and promptly dropped about 6 seconds in a 100, to set a new national record.
How to account for such a drastic difference? At the Boston meet, Mike Ross theorized that the suit could reposition the body in a way that greatly increased efficiency for some people (but maybe not have much effect on others). In that scenario, it is even possible that with proper training the improvement could be maintained even after the suits become less buoyant or less compressing. George Parks mentioned something similar in a recent post.
All this hand waving mostly means is that you're on your own. Right now the swim ratings are based on 2007 records (ie, pre-LZR). I'll add one based on the most recent records, so you can use them to get a ballpark estimate of the effect (eg, determine the post-LZR rating of 52.86 and then use it to calculate an equivalent time with the pre-LZR rating curves).
Out of curiosity, I once plotted a rough graph of 50SCM record times (from USMS rulebook) correlating to age. I just wanted to see the typical sprinting speed decline with age. Haven't compared it to a distance event yet but I assume the curve would shift right.
Last edited by __steve__; November 12th, 2011 at 08:34 PM.
Thanks for the reference, Chris. One question: is it correct to assume these predictions were based on improvements from earlier generation body suits--like the Fastskin Pro, FSII, etc.?
If so, then finding that the 100 LCM freestyle times improved by .7 seconds more than anticipated does not really tell the story of how these guys would have swum in jammers, right? It only says the top 8 improved, on average, by .7 seconds more than historical time drop norms would have predicted, with these historical norms based, at least in recent years, on swims done in all the textile body suits in wide use since the Aquablade made its debut in 1996. According to Speedo, this same year, at the Atlanta Olympics, 76% of all swimming medals were won in SPEEDO, most wearing SPEEDO Aquablade."
I guess my basic question is probably unanswerable for now--how much of a time increase is likely to occur when going from B70 to jammers, not going from B70 to something very much like a zipper-less Fastskin Pro (as the women are probably going to be allowed to use).
As far as it being very individualized, I agree with this supposition but only so far. I would think that for the outliers on either end, the suit change could make either drastic or negligible differences in individual performance. But I still maintain that for the bulk of swimmers, there has got to be a reasonably accurate conversion factor--even if this is not likely to be discovered till enough actual times start filtering in to fit curves to the new data.
PS my friend and swimming coach Bill White, who is 39, and I both swam the 400 SCY IM at a local Y meet a few weeks back. I just used your grading formula and found--to my delight--that Bill's 4:34.12 at age 39 scored exactly the same as my 5:06.45 at age 57. We both scored 89.3. I don't think this is necessarily very motivating for Bill, but it certainly is for me!
This doesn't necessarily mean older technology wasn't at all effective. It might mean is that the advancement in suit technology was fairly steady -- and incorporated in the model -- until the latest jump (ie, invention of the LZR, B70, etc).
That's where you lose me ....If you make the assumption that the records are, generally, equally hard across all events and across genders
The assumption being incorrect basically affects the ability to use the rating as a basis to compare performances in different events or between genders. For example, I can pick two very impressive swims from the NEM swim meet.
Mike Ross, age 41, 50 back: 25.53, rated at 104.6
Karlyn Pipes-Neilsen, age 47, 200 fly: 2:23.18, rated at 104.2
Consider how difficult a comparison this is: different genders, different ages, different events. The rating system gives a slight nod to Mike's swim. But that assumes that the records for women's 200 SCM fly are -- as a whole -- equally as tough as the men's 50 SCM back records. If we wish to use the rating in this manner, you are basically forced to make that assumption. So the accuracy of the statement, "Mike's swim is a little more impressive than Karlyn's" is only as good as that assumption is true.
If you also compare across courses (eg, to Mike's 100 SCY back from last year was 48.49, rated at 106.9), you have to make a further assumption that SCM records are equally as hard as SCY records. This is a dubious assumption in my mind, though I think the SCM records have gotten much tougher over the last 5-10 years. But if you want to make comparisons across courses, you really are stuck with making an assumption of this type.
Finally, all of those ratings use 2007 (pre-LZR) records to evaluate swims done in LZR-era suits, which tends to inflate those ratings and make it difficult to compare to older swims.
But if you use the ratings in a more limited sense, such as personal goal-setting, fewer assumptions need to be made. If my rating in a given event improves over the years, it means that I am withstanding the rigors of aging better than the record-holders in that event. That's not a bad thing...
Last edited by Chris Stevenson; December 15th, 2009 at 11:17 PM.
It's just my opinion - and I understand the math, but for the Men - all the strokes are valued way way too high compared to Freestyle. Have not looked at the women's records.
You mentioned Mike's swim from last year" " to Mike's 100 SCY back from last year was 48.49, rated at 106.9"
--- in order to get that same rating in Freestyle (same age group), one would have to swim a 42.60 in a 100 Free. Now, a 48.49 is silly fast, but it can't compare to a 42.6. There always exceptions, but for the most part the Free records are much closer to the Open world records than in the strokes.
Look at the 200s in my age group - Free record is about 7 seconds above the NCAA record -- Backstroke is 13 / Breast 12 sec / Fly 11 / IM 11
I set an IM record and almost a back record - I am not that good of an IMer.
Probably those with a merely healthy dose of competitiveness just skipped this thread, or are biting their tongue. Not that internet forum-dwellers are usually known for saying nothing when they have nothing nice to say. What a polite group this must be...