Comparing Times With Others

[old] John Rupp · Post by **[old] John Rupp** » June 15th, 2004, 3:08 am

As many others, I have long had the interest of comparing my times with other athletes and competitors. For example, as a runner I compared my time to the world records by inverting the times. My 440 yard time of 51.7 was about 85% of the world record at the time. Doing the same for the other events allowed me to compare the relative closeness, set my objectives to get closer, and even compete with myself in one event against another event the next time. I wondered why the same simple formula was not used for for the decathlon, and out of curiosity did my own calculations for how it affected the scores. Now there are many who have the same curiosity in rowing, to compare times between the different age groups and weight classes for women and for men. The Nonathlon does an excellent job of this, and keeps improving it's method. A problem with doing such comparisons, is the great multitude of events. First off all, the records need to be determined for ALL the events, in every class and division, and then somehow make an equivalent comparison of these. The difficulty is compounded in that the 2k is by far the most popular competitive venue and international event. Thus other events do not have the same history and competitive nature. Although some divisions do have excellent times in certain events, such as Freed's distance times, other divisions do not. Thus comparisons between divisions and events is very difficult, if not totally impossible to calculate. One way would be to take the *best* time percentage in one division, and apply the same percentage to all the other divisions, for the given event. However, I have discovered an even better way, that equalizes ALL events and all divisions to one standard. This standard is the world record in the 2k for that division, and that age. For example, the women's LWT world record 2k is 8:05.4 for age 60 and 8:54.6 for age 70. By using the power learning curve formula, a Predicted Age Time can be calculated for each age. I originally called this predicted world record" (PWR) but "predicted age time" is more appropriate. The Predicted Age Times, rounded to the nearest 10th of a second, are as follows: STEP #1 - determining the PAT pace for each age: 60 2:01.4 / 61 2:02.7 / 62 2:03.9 / 63 2:05.1 / 64 2:06.4 65 2:07.2 / 66 2:08.8 / 67 2:10.0 / 68 2:11.2 / 69 2:12.4 70 2:13.6 STEP #2 - relating the WR pace to the rower's times for each event: My times come out like this --- PAT 1:40.8 --- total points 8840 / 100 = 88.40 500 ---- 1000 --- 2000 ---- 5000 ---- 6000 ---- 10000 ---- 21097 ---- 42195 -- 30:00 -- 60:00 1:39.6 - 3:35.6 - 7:24.0 - 18:57.4 - 22:49.9 - 38:44.7 - 1:22:48.9 - 2:58:46.2 - 7743 - 15285 1:39.6 - 1:47.8 - 1:51.0 -- 1:53.7 -- 1:54.2 --- 1:56.2 ---- 1:57.8 ---- 2:07.1 - 1:56.2 - 1:57.8 988 ----- 935 ----- 908 ----- 887 ------ 883 ----- 867 ------- 856 ------- 793 ---- 867 ----- 856 *Note: Math corrected in last line above. Here is a comparison with the 1st two rowers in the Nonathlon. For the purpose of direct comparison between age groups and events, the PAT method counts all 10 events and no bonus. 500 ---- 1000 --- 2000 ---- 5000 ---- 6000 ---- 10000 ---- 21097 ---- 42195 -- 30:00 -- 60:00 1 Ole Granny 9470 68 F H -- 2:07.9 PAT pace -- 8599 points / 100 = 85.99 (9 events) 1:53.4 - 4:05.4 - 8:30.3 - 22:32.9 - 27:43.2 - 46:32.9 - 1:39:31.0 - 0:00:00.0 - 6494 - 12761 1:53.4 - 2:02.7 - 2:07.6 -- 2:15.3 -- 2:18.6 -- 2:19.6 ----- 2:36.8 ------ 0 ------ 2:18.6 - 2:21.1 1128 --- 1042 --- 1002 ---- 945 ------ 922 ----- 916 -------- 816 ------- 0 ------- 922 ----- 906 2 DavidMiller 9259 Nonathlon points 17 M L --- 1:35.9 WR pace --- 9040 / 100 = 90.4 1:28.8 - 3:07.3 - 6:39.3 - 17:42.5 - 21:31.8 - 37:22.0 - 1:22:39.3 - 2:47:43.1 - 8194 - 15619 1:28.8 - 1:33.6 - 1:39.8 -- 1:46.2 -- 1:47.6 --- 1:52.1 ---- 1:57.5 ---- 1:59.2 -- 1:49.8 - 1:55.2 1080 --- 1025 --- 961 ------ 903 ----- 891 ------ 855 ------- 816 -------- 804 ----- 873 ---- 832 Notice the PAT method gives more than 100 points to PAT's at less than 2k, and less than 100 points for PAT's at events longer than 2k. This works out rather evenly, as the difference for 500 meters is greater than for the half marathon etc, thus also equalizing over weight divisions. I hope this PAT method is useful to others. Let me know if you like it. Thanks.

[old] John Rupp · Post by **[old] John Rupp** » June 15th, 2004, 3:20 am

I'm quite impressed with Ol' Granny's performances. I'd thought the reason to be the Nonathlon skewing the times in favor of certain groups. However, Ol' Granny's 9 event total of 85.99, with the PAT method above, is at the rate of 95.44 for 10 events. I first mistakenly used the lwt PAT, but have corrected the percentages/

[old] PaulH · Post by **[old] PaulH** » June 15th, 2004, 10:06 am

Understandably I've looked at this from a number of angles already! I agree with everything you said John (which may be a first!) My only concern here is the use of the power learning curve. I'm unclear as to where the extra points come from (one point is 2k, what is the other?) Even with two points, fitting a curve is only a guess, doesn't necessarily match reality anyway, and may ignore variances we know exist between groups (e.g. lightweight have a shallower curve then heavyweights). Having said all that, I think your method is at least as valid as the way I'm calculating the Nonathlon scores, particularly as you say because the 2k is the only event we know to be heavily contested. Cheers, Paul

[old] John Rupp · Post by **[old] John Rupp** » June 15th, 2004, 11:40 am

Cheers, Paul, Curiousity got the better of me last night. Fortunately I was able to stop before it got *too* late. <table border='0' align='center' width='95%' cellpadding='3' cellspacing='1'><tr><td class='genmed'>QUOTE </td></tr><tr><td class='quote'>I'm unclear as to where the extra points come from (one point is 2k, what is the other?)  </td></tr></table> All Predicted Age Times (PAT's) are predicted from WR times already established from the 2k. The WR time is taken as being either the lowest, or the highest, age in the bracket. For example the 50-59 time is taken as age 50 and the 13-18 time is taken to be age 18. Calculating the PAT's with the power learning curve formula is very simply done. For example, the women's LWT record for age 60-69 is 2:01.4 and the 70-79 record is 2:13.6 pace. The men's LWT record is 1:48.3 pace for 12 and under and 1:34.0 pace for ages 13 to 18. Thus: 60 2:01.4 / 61 2:02.7 / 62 2:03.9 / 63 2:05.1 / 64 2:06.4 65 2:07.2 / 66 2:08.8 / 67 2:10.0 / 68 2:11.2 / 69 2:12.4 70 2:13.6 12 1:48.3 / 13 1:45.3 / 14 1:42.6 / 15 1:40.2 / 16 1:38.0 / 17 1:35.9 / 18 1:34.0 All the other event percentages are then easily calculated based on the PAT for that age, weight, and gender division.

[old] PaulH · Post by **[old] PaulH** » June 15th, 2004, 12:15 pm

I didn't quite get that last point - Assuming I have a valid 2k PWR for an age (there's some variability there, but your method gives a good approximation I think) how do I know what the 5k PWR should be?

[old] John Rupp · Post by **[old] John Rupp** » June 15th, 2004, 2:40 pm

Paul, That's the beauty of this. None of the other event Predicted Age Times (PAT) need to be known, as they are *all* based on the 2k. For example, let's say given the PAT for 50's LWT is 1:37.0 and some else age 50 rows a 5k at 1:48.0 pace. This would be 1:37.0 / 1:48.0 = 898 points (or 89.8 etc). Also, this is regardless of the event! If it was a 10k (or an hour, or a marathon, or 500 meters) at 1:48.0 pace it would still be 898 points -- although 898 points for a 10k might end up being in first place!, as very few others will end up with as much as 898 points for a 10k.

[old] PaulH · Post by **[old] PaulH** » June 15th, 2004, 3:14 pm

So you would calculate one PWR for each of the four major groups (Lwt/Hwt x M/F)? Doesn't that give an advantage to long distance rowers? Cheers, Paul

[old] Ralph Earle · Post by **[old] Ralph Earle** » June 15th, 2004, 3:32 pm

You've given this a lot of careful thought, John, and at first reading it seemed an elegant solution. But then I thought that giving someone the same number of points for the same pace, no matter the distance would create a strong incentive to improve one's times at the shorter distances and neglect the longer ones. After all, although you personally could row both a 500m and a 5000m @2:00.0 for about 900 points, I assume you would much rather erg 500m than 5000m. But on further reflection, I realized that your PWR method gives the same reward to a one-second improvement in pace regardless of the distance, which gives every one the same age/weight/gender the same incentive to better their weakest event(s). The only suggestion I would make is, rather than estimate the age-specific WRs from the age-group ones, that PaulH and Cran use the fastest 2K race times in the rankings and continue to smooth them to remove the undue influence of a few extraordinarily fast times. Well done, John!

[old] John Rupp · Post by **[old] John Rupp** » June 15th, 2004, 5:34 pm

Paul, There is a Predicted Age Time (PAT) for every age, for each of the 4 divisions. Ralph, Yes, the PAT's *are* smoothed, from the current 2k world records. The 2 reference points for each age group, are the current world records on each side of that age group. Each year in between these 2 reference points is smoothed, by the power learning curve, to determine the PAT's for each year (each age), in each division. For example, the open LWT (note correction) women's 60-69 record is 2:01.4 pace, and the age 70 (to 79) record is 2:13.6 pace. These are the current 2k world records, in bold type below, and the PMR's are calculated in between them. 60 2:01.4 / 61 2:02.7 / 62 2:03.9 / 63 2:05.1 / 64 2:06.4 65 2:07.2 / 66 2:08.8 / 67 2:10.0 / 68 2:11.2 / 69 2:12.4 70 2:13.6 The men's LWT record for 12 (and under) is 1:48.3 pace, and the age (13 to) 18 record is 1:34.0 pace. These are the current 2k world records, in bold type below, and the PAT's are calculated in between them. 12 1:48.3 / 13 1:45.3 / 14 1:42.6 / 15 1:40.2 / 16 1:38.0 / 17 1:35.9 / 18 1:34.0 Records within each division are calculated from the current 2k world records on each side of that division, rather than using the entire world record base, in order to predict the intermediate PAT's based on *what has been done*, rather than *what might be done*. As the 2k records are improved, the intermediate PAT's can likewise be adjusted. But this doesn't happen that often. The PAT's are calculated in each division by the 2k world records on either side of that division, and each division is likewise calculated, seperately from any of the other divisions.

[old] John Rupp · Post by **[old] John Rupp** » June 15th, 2004, 6:31 pm

Current 2k World Records, for determining Predicted Age Times (PAT) for each age. Age ----- w lwt ----- w hwt ----- m lwt ----- m hwt 12 ----- 2:05.8 ----- 1:52.6 ----- 1:48.3 ----- 1:48.3 18 ----- 1:46.8 ----- 1:38.5 ----- 1:34.0 ----- 1:26.8 29 ----- 1:45.0 ----- 1:37.2 ----- 1:30.6 ----- 1:24.6 30 ----- 1:44.2 ----- 1:37.2 ----- 1:31.6 ----- 1:24.2 40 ----- 1:48.2 ----- 1:42.0 ----- 1:34.6 ----- 1:28.0 50 ----- 1:51.3 ----- 1:46.6 ----- 1:37.0 ----- 1:31.9 60 ----- 2:01.3 ----- 1:56.8 ----- 1:41.6 ----- 1:35.9 70 ----- 2:13.6 ----- 2:10.6 ----- 1:48.5 ----- 1:45.6 80 ----- 2:31.4 ----- 2:31.4 ----- 2:03.4 ----- 2:00.9 90 ----- 3:01.9 ----- 3:01.9 ----- 2:21.4 ----- 2:21.4

[old] PaulH · Post by **[old] PaulH** » June 16th, 2004, 9:26 am

I think I've got it! Two questions then. First, why do you assume 29 and 30 as the two 'representative' ages for those groups. The fastest people in the world are not, to my knowledge, 29.5, they seem to be about 33. So shouldn't you choose 29,33 for those two categories. Second, as I mentioned before, doesn't this give an advantage to lightweights? They're generally able to keep a greater percentage of their 2k rates at longer distances than heavyweights, but the heavyweights only get two events (1k, 500m) to compensate on the shorter end? Great system btw! Cheers, Paul

[old] John Rupp · Post by **[old] John Rupp** » June 16th, 2004, 2:42 pm

Women Predicted Age Times based on current 2k World Records age - 10 ---- 11 ---- 12 ---- 13 ---- 14 ---- 15 ---- 16 ---- 17 ---- 18 ---- 19 lwt 2:15.4 2:10.3 2:05.8 2:01.8 1:58.2 1:55.0 1:52.0 1:49.3 1:46.8 1:44.5 hwt 1:59.6 1:55.9 1:52.6 1:49.7 1:47.0 1:44.6 1:42.4 1:40.4 1:38.5 1:37.2 age - 20 ---- 21 ---- 22 ---- 23 ---- 24 ---- 25 ---- 26 ---- 27 ---- 28 ---- 29 lwt 1:44.2 1:44.2 1:44.2 1:44.2 1:44.2 1:44.2 1:44.2 1:44.2 1:44.2 1:44.2 hwt 1:37.2 1:37.2 1:37.2 1:37.2 1:37.2 1:37.2 1:37.2 1:37.2 1:37.2 1:37.2 age - 30 ---- 31 ---- 32 ---- 33 ---- 34 ---- 35 ---- 36 ---- 37 ---- 38 ---- 39 lwt 1:44.2 1:44.6 1:45.1 1:45.5 1:45.9 1:47.3 1:46.7 1:47.1 1:47.5 1:47.8 hwt 1:37.2 1:37.2 1:37.6 1:38.2 1:38.8 1:39.3 1:39.9 1:40.4 1:41.0 1:41.5 age - 40 ---- 41 ---- 42 ---- 43 ---- 44 ---- 45 ---- 46 ---- 47 ---- 48 ---- 49 lwt 1:48.2 1:48.5 1:48.9 1:49.2 1:49.5 1:49.8 1:50.1 1:50.4 1:50.7 1:51.0 hwt 1:42.0 1:42.5 1:43.0 1:43.5 1:44.0 1:44.4 1:44.9 1:45.3 1:45.7 1:46.2 age - 50 ---- 51 ---- 52 ---- 53 ---- 54 ---- 55 ---- 56 ---- 57 ---- 58 ---- 59 lwt 1:51.3 1:52.4 1:53.4 1:54.4 1:55.5 1:56.5 1:57.5 1:58.5 1:59.5 2:00.4 hwt 1:46.6 1:47.7 1:48.7 1:49.8 1:50.8 1:51.8 1:52.8 1:53.8 1:54.8 1:55.8 age - 60 ---- 61 ---- 62 ---- 63 ---- 64 ---- 65 ---- 66 ---- 67 ---- 68 ---- 69 lwt 2:01.4 2:02.7 2:03.9 2:05.1 2:06.4 2:07.2 2:08.8 2:10.0 2:11.2 2:12.4 hwt 1:56.8 1:58.2 1:59.6 2:01.0 2:02.4 2:03.8 2:05.1 2:06.5 2:07.9 2:09.2 age - 70 ---- 71 ---- 72 ---- 73 ---- 74 ---- 75 ---- 76 ---- 77 ---- 78 ---- 79 lwt 2:13.6 2:15.4 2:17.2 2:19.0 2:20.7 2:22.5 2:24.3 2:26.1 2:27.9 2:29.6 hwt 2:10.6 2:12.7 2:14.7 2:16.8 2:18.9 2:21.0 2:23.0 2:25.1 2:27.2 2:29.3 age - 80 ---- 81 ---- 82 ---- 83 ---- 84 ---- 85 ---- 86 ---- 87 ---- 88 ---- 89 l/h 2:31.4 2:34.4 2:37.3 2:40.3 2:43.4 2:46.4 2:49.5 2:52.5 2:55.6 2:58.8 age - 90 ---- 91 ---- 92 ---- 93 ---- 94 ---- 95 ---- 96 ---- 97 ---- 98 ---- 99 l/h 3:01.9 3:05.1 3:08.2 3:11.4 3:14.7 3:07.9 3:21.1 3:24.4 3:27.7 3:31.0

[old] John Rupp · Post by **[old] John Rupp** » June 16th, 2004, 3:05 pm

Paul, Thanks. <table border='0' align='center' width='95%' cellpadding='3' cellspacing='1'><tr><td class='genmed'>QUOTE </td></tr><tr><td class='quote'>Two questions then. First, why do you assume 29 and 30 as the two 'representative' ages for those groups. The fastest people in the world are not, to my knowledge, 29.5, they seem to be about 33. So shouldn't you choose 29,33 for those two categories.  </td></tr></table> The 29 and 30 were just the official "given" world records. After reading your message I also expanded the curve out from the given 12 and 18, and down from the given 40 and 50. These gave WR paces at just over age 19, and just under age 33. Based on this, I adjusted the Predicted Age Times (PAT) to be identical from age 20 through 33, and then used age 33 as the reference point to age 40. Historically, world records have been set in athletic events between the ages of 19 and 38, but most commonly in the late 20's. If memory serves, the average age of Olympic Champions in athletics is age 28. It would also be possible to "average" the two adjustments between ages 18 and 30, i.e, 18 to 24 and 24 to 30. The same could be done from ages 30 to 33 and 33 to 40. In any case, the Predicted Age Times (PAT) listed here are already quite accurate. <table border='0' align='center' width='95%' cellpadding='3' cellspacing='1'><tr><td class='genmed'>QUOTE </td></tr><tr><td class='quote'>Second, as I mentioned before, doesn't this give an advantage to lightweights? They're generally able to keep a greater percentage of their 2k rates at longer distances than heavyweights, but the heavyweights only get two events (1k, 500m) to compensate on the shorter end?  </td></tr></table> Ol' Granny has around 95.5 (considering 10 events), which is 95.5 percent of WR 2k pace for *all* of the ten events, and she is heavyweight. That is certainly an excellent percentage, regardless of the weight class. Dwayne Adams is HWT and does very well at the distances. Even Graham Benton, who has around a 130 percent for the 500 meters, does very well in the distance events. My feeling is that it will come out very evenly. Also it would likely encourage sprint specialists to improve their distance times, and visa versa.

[old] John Rupp · Post by **[old] John Rupp** » June 17th, 2004, 4:56 pm

Men Predicted Age Times based on current 2k World Records age - 10 ---- 11 ---- 12 ---- 13 ---- 14 ---- 15 ---- 16 ---- 17 ---- 18 ---- 19 lwt 1:55.4 1:51.7 1:48.3 1:45.3 1:42.6 1:40.2 1:38.0 1:35.9 1:34.0 1:32.2 hwt 1:55.4 1:50.2 1:45.6 1:41.6 1:38.0 1:34.8 1:31.9 1:29.2 1:26.8 1:24.6 age - 20 ---- 21 ---- 22 ---- 23 ---- 24 ---- 25 ---- 26 ---- 27 ---- 28 ---- 29 lwt 1:30.6 1:30.6 1:30.6 1:30.6 1:30.6 1:30.6 1:30.6 1:30.6 1:30.6 1:30.6 hwt 1:24.2 1:24.2 1:24.2 1:24.2 1:24.2 1:24.2 1:24.2 1:24.2 1:24.2 1:24.2 age - 30 ---- 31 ---- 32 ---- 33 ---- 34 ---- 35 ---- 36 ---- 37 ---- 38 ---- 39 lwt 1:30.6 1:31.0 1:31.5 1:31.9 1:32.3 1:32.7 1:33.1 1:33.5 1:33.9 1:34.2 hwt 1:24.2 1:24.2 1:24.2 1:24.7 1:25.2 1:25.7 1:26.2 1:26.7 1:27.1 1:27.6 age - 40 ---- 41 ---- 42 ---- 43 ---- 44 ---- 45 ---- 46 ---- 47 ---- 48 ---- 49 lwt 1:34.6 1:34.9 1:35.1 1:35.4 1:35.6 1:35.9 1:36.1 1:36.3 1:36.6 1:36.8 hwt 1:28.0 1:28.2 1:28.8 1:29.2 1:29.6 1:30.0 1:30.4 1:30.9 1:31.2 1:31.5 age - 50 ---- 51 ---- 52 ---- 53 ---- 54 ---- 55 ---- 56 ---- 57 ---- 58 ---- 59 lwt 1:37.0 1:37.5 1:38.0 1:38.4 1:38.9 1:39.4 1:39.8 1:40.3 1:40.7 1:41.1 hwt 1:31.9 1:32.3 1:32.7 1:33.2 1:33.6 1:34.0 1:34.4 1:34.8 1:35.1 1:35.5 age - 60 ---- 61 ---- 62 ---- 63 ---- 64 ---- 65 ---- 66 ---- 67 ---- 68 ---- 69 lwt 1:41.6 1:42.3 1:43.0 1:43.7 1:44.4 1:45.1 1:45.8 1:46.5 1:47.2 1:47.8 hwt 1:35.9 1:36.9 1:37.9 1:38.9 1:39.8 1:40.8 1:41.8 1:42.7 1:43.7 1:44.7 age - 70 ---- 71 ---- 72 ---- 73 ---- 74 ---- 75 ---- 76 ---- 77 ---- 78 ---- 79 lwt 1:48.5 1:50.0 1:51.5 1:53.0 1:54.5 1:56.0 1:57.4 1:58.9 2:00.4 2:01.9 hwt 1:45.6 1:47.1 1:48.7 1:50.2 1:51.7 1:53.2 1:54.8 1:56.3 1:57.8 1:59.4 age - 80 ---- 81 ---- 82 ---- 83 ---- 84 ---- 85 ---- 86 ---- 87 ---- 88 ---- 89 lwt 2:03.4 2:05.2 2:07.0 2:08.8 2:10.6 2:12.4 2:14.2 2:16.0 2:17.8 2:19.6 hwt 2:00.9 2:02.9 2:04.9 2:07.0 2:09.0 2:11.1 2:13.1 2:15.2 2:17.2 2:19.3 age - 90 ---- 91 ---- 92 ---- 93 ---- 94 ---- 95 ---- 96 ---- 97 ---- 98 ---- 99 l/h 2:21.4 2:23.2 2:25.0 2:26.9 2:28.7 2:30.5 2:32.4 2:34.2 2:36.0 2:37.9

[old] John Rupp · Post by **[old] John Rupp** » June 18th, 2004, 9:12 pm

The charts are now complete!!! A few notes: #1- Current world records were used for all reference points; #2- Records were rounded to the nearest tenth, or if a choice, to the nearest even tenth; #3- The predicted age times are based on what has been done, i.e. current records, rather than what could be done; #4- Thus some categories, like some records, are stronger or weaker than others; #5- These charts are called the Predicted Age Times (PAT), instead of the original "predicted world records" (PWR); #6- The same Predicted Age Times (PAT) were used from age 20 through age 32, based on taking curves from ages 12 and 18 through 20, and curves from ages 40 and 50 through age 32, and these curves vary slightly between categories; #7- Men's times, based on the age 12 and 18 curves, are equalized at age 10; #8- Women's times are not equalized at age 10; #9- The charts reflect different rates of maturation and levels of decline; #10- Thus each age/weight/gender class is slightly different from the others; #11- Where the lightweight record is also the open class record, it is used for both categories. A few errors have been found and corrected. Please bring any others to my attention. Cheers.

Concept2 Forum

Comparing Times With Others

Competitions

Competitions

Competitions

Competitions

Competitions

Competitions

Competitions

Competitions

Competitions

Competitions

Competitions

Competitions

Competitions

Competitions

Competitions