Age-Grading for Masters Swimmers

I do think there has to be strong consideration for using the median time of the Top 10 or Top 50 in each event and each course. The reason is because, for the most part, the records are set by elite swimmers. Since I believe elite swimmers are genetically different than the rest of us, they will always have an advantage over the rest of us. Yes, there are exceptions. BUT, in general, it holds true. Paul Windrath As a swimmer in the age group dominated by the magnificent Rick Colella, I have no doubt you are correct about the nature of the elite! Nevertheless, Alan Rowson, the British fellow who does the age-rating for the British Decathalon, addressed this exact question regarding using WRs vs median Top 10 time: I started to look at age adjustment formulae as a result of feedback about the accuracy of the Finnish formulae. I explained it to a friend of mine who was a lecturer in Maths at Cambridge University. He said it would make a good maths challenge for the local schools to see who would come up with the best solution. Anyhow that failed to come to fruition. What he did say was it might be better to use median of top ten masters times rather than masters world records. I did derive formulae using both. The use of median was much more labour intensive to derive formulae and decisions as what to do in older age groups when there are not ten times recorded was an issue. The formulae produced by median and record only were not hugely different, so I took pragmatic view of using world records as easier in time effort to produce. As Alan says, the curves produced by WRs vs. median Top 10 times were not terribly different. Even when somebody like Rick smashes a WR (in the 400 SCM free, for instance, Rick's recent 4:27.91 destroyed Jim McConica's 4:45.88), such an "outlier point" is just one of many that the fitting function takes into account. Perhaps the best way to combine pragmatism with a nod to the fact that most of us aren't swimming gods would be to take the middle time each year in the FINA ALL Time Top 10 ranking. This hasn't yet been updated to reflect Rick's super swim, but if you check out this, archives.fina.org/.../tabs_SC_all.pdf you will find that the 400 SCM free for 65-69 year old men is as follows: 400 M. FREE 4:45.88 JIM MC CONICA USA 2015 4:52.89 RICK WALKER USA 2015 4:53.03 DAN KIRKLAND USA 2015 4:54.45 GLENN GRUBER USA 2014 4:54.84 J.CHATARD FRA 2016 4:55.56 TOM LANDIS USA 2007 4:55.97 DAVID QUIGGIN USA 2009 4:59.03 VITTORIO ERMIRIO ITA 2007 4:59.42 JIM CLEMMONS USA 2015 5:00.26 DON DAVIS USA 2009 When this year's SCM times are added in, Don Davis will be bumped off the Top10 and conceivably one or two others depending on who else in the age group might swim faster. If you use the 5th place time for the fitting function, even with Rick smashing the old record by nearly 20 seconds, the data point would not change much (i.e., right now J. Chatard is in 5th; after Rick's time is posted, Glenn Gruber will be in 5th. The difference between Chatard and Gruber is only .39 seconds.) Anyhow, this could be one way to ensure a more moderate change over the years. By picking the person in the 5th place spot, you wouldn't have to do any tedious averaging, etc. And with the exception of the very oldest age groups, odds are that there will be at least five swimmers in every All Time Top 10 tabulation. What do you think?

I do think there has to be strong consideration for using the median time of the Top 10 or Top 50 in each event and each course. The reason is because, for the most part, the records are set by elite swimmers. Since I believe elite swimmers are genetically different than the rest of us, they will always have an advantage over the rest of us. Yes, there are exceptions. BUT, in general, it holds true. Paul Windrath As a swimmer in the age group dominated by the magnificent Rick Colella, I have no doubt you are correct about the nature of the elite! Nevertheless, Alan Rowson, the British fellow who does the age-rating for the British Decathalon, addressed this exact question regarding using WRs vs median Top 10 time: I started to look at age adjustment formulae as a result of feedback about the accuracy of the Finnish formulae. I explained it to a friend of mine who was a lecturer in Maths at Cambridge University. He said it would make a good maths challenge for the local schools to see who would come up with the best solution. Anyhow that failed to come to fruition. What he did say was it might be better to use median of top ten masters times rather than masters world records. I did derive formulae using both. The use of median was much more labour intensive to derive formulae and decisions as what to do in older age groups when there are not ten times recorded was an issue. The formulae produced by median and record only were not hugely different, so I took pragmatic view of using world records as easier in time effort to produce. As Alan says, the curves produced by WRs vs. median Top 10 times were not terribly different. Even when somebody like Rick smashes a WR (in the 400 SCM free, for instance, Rick's recent 4:27.91 destroyed Jim McConica's 4:45.88), such an "outlier point" is just one of many that the fitting function takes into account. Perhaps the best way to combine pragmatism with a nod to the fact that most of us aren't swimming gods would be to take the middle time each year in the FINA ALL Time Top 10 ranking. This hasn't yet been updated to reflect Rick's super swim, but if you check out this, archives.fina.org/.../tabs_SC_all.pdf you will find that the 400 SCM free for 65-69 year old men is as follows: 400 M. FREE 4:45.88 JIM MC CONICA USA 2015 4:52.89 RICK WALKER USA 2015 4:53.03 DAN KIRKLAND USA 2015 4:54.45 GLENN GRUBER USA 2014 4:54.84 J.CHATARD FRA 2016 4:55.56 TOM LANDIS USA 2007 4:55.97 DAVID QUIGGIN USA 2009 4:59.03 VITTORIO ERMIRIO ITA 2007 4:59.42 JIM CLEMMONS USA 2015 5:00.26 DON DAVIS USA 2009 When this year's SCM times are added in, Don Davis will be bumped off the Top10 and conceivably one or two others depending on who else in the age group might swim faster. If you use the 5th place time for the fitting function, even with Rick smashing the old record by nearly 20 seconds, the data point would not change much (i.e., right now J. Chatard is in 5th; after Rick's time is posted, Glenn Gruber will be in 5th. The difference between Chatard and Gruber is only .39 seconds.) Anyhow, this could be one way to ensure a more moderate change over the years. By picking the person in the 5th place spot, you wouldn't have to do any tedious averaging, etc. And with the exception of the very oldest age groups, odds are that there will be at least five swimmers in every All Time Top 10 tabulation. What do you think?