2024 Advanced Training and Racing Thread (still competitive jerks) (Read 270 times)

wcrunner2


Are we there, yet?

     

    Trying to follow this argument about AG (which btw I have never paid any attention to at all). I think JMac here is arguing that the talent pool is much deeper for the 20-40 age range, since that age range has the most runners. I don't know this for sure but I would speculate that among the runners between 20 and 40, a higher percentage of them are "competitive" in the sense of constantly trying to set new PRs and beat peers.

     

    But the AG is not dependent on the depth of the pool. The AG is computed as age std/race time.  The depth of the talent pool may, though after 20+ years of compiling data, I doubt it, affect the age graded time as the age factor is dependent on sufficient data to determine how much performance degradation is due to age and not other factors.  If the age std for the mile is 4:00 and I run 5:00, then my AG = 80% (4:00/5:00 as a percent).  My age equivalent time would be 4:39. The age factor of .93 is dependent on size and validity of the results.  Since training levels would be most comparable with the high end performers, I doubt that results from slower runners would have much of an effect on the determination of the age factor and none on the age std.

     2024 Races:

          03/09 - Livingston Oval Ultra 6-Hour, 22.88 miles

          05/11 - D3 50K, 9:11:09
          06/17 - 6 Days in the Dome 12-Hour, 35.82 miles
          10/12 - Hainesport 12-Hour

     

     

         

    mmerkle


       

      But the AG is not dependent on the depth of the pool. The AG is computed as age std/race time.  The depth of the talent pool may, though after 20+ years of compiling data, I doubt it, affect the age graded time as the age factor is dependent on sufficient data to determine how much performance degradation is due to age and not other factors.  If the age std for the mile is 4:00 and I run 5:00, then my AG = 80% (4:00/5:00 as a percent).  My age equivalent time would be 4:39. The age factor of .93 is dependent on size and validity of the results.  Since training levels would be most comparable with the high end performers, I doubt that results from slower runners would have much of an effect on the determination of the age factor and none on the age std.

       

      Ok so I think I'm with you on the age factor not depending on the talent pool. Where I need clarification perhaps, is with the percentage that comes from comparing your time to the WR for that distance and age group. If one age range has more people competing and pushing that boundary, it seems to me that WR is going to be more difficult (or rather COULD BE more difficult), relative to that age group, than the WR for the group that has less people competing.

      Marky_Mark_17


        Ehh, I’m old and a Data Scientist and tend to agree with JMac. If you peak in your 30s chances are you won’t be running much or at all in your 50s. Way fewer people peak in their 50s so if you do, you’ll be way up the AG compared to what you could have done in your 30s.

         

        How about that, didn’t even need to drop the data scientist label to explain the point. But I did anyway.

         

        Except wcrunner pointed out the AG % is calculated based on your relativity to the fastest time for the age group.  If that is true, it doesn't matter how many people are in the AG, or how many are at their peak, what matters is just the fastest time.

         

        There is an argument that says with fewer people running in older AG's, somehow the record is less valid, but I doubt you could suggest that makes a lot of difference.  And interestingly, the Sally Gibbs example comes up again because she is setting older AG records having set none when she was younger.

        3,000m: 9:07.7 (Nov-21) | 5,000m: 15:39 (Dec-19) | 10,000m: 32:34 (Mar-20)  

        10km: 33:15 (Sep-19) | HM: 1:09:41 (May-21)* | FM: 2:41:41 (Oct-20)

        * Net downhill course

        Last race: Runway5 / National 5k Champs, 16:22, National Masters AG Champ!

        Up next: Still working on that...

        "CONSISTENCY IS KING"

        JMac11


        RIP Milkman

          Maybe we should just clarify something here, and if I’m wrong on this, I would backtrack on this argument. My understanding is the percentages are set off the world record in that age group, by gender, by distance, etc. So, put another way, 100% AG means you are the world record holder for that age group in whatever the distance is.

           

          Assuming that is true, the depth of the pool is absolutely relevant. I think some of you guys are mixing up sample sizing when it comes to averages vs. deep tail events. Put another way, you may only need approximately 1000 runners within the age group to be 95% sure that the average time of that pool of runners is equal to the average time of that age group in the broader population. However, you would need a group many, many times that size in order to be sure that the fastest time you recorded in that group is actually the fastest time within the entire population.

           

          So my entire point is, in let’s say the age 55-59 bracket, you are dealing with a much smaller sample size of runners, meaning you are much less likely to get the true representative of the world record holder, meaning that an 80% AG of the world record may only really be 75% if you get the right pool of people.

           

          And sorry, this is very technical statistical stuff here, but it’s why the age-grading doesn’t work well comparing against ages. You are much more likely to have the fastest guy in the world actually running when he’s 25 than you are to find the fastest guy in the world running when he’s 55. To bring this back to running: I’ll stick with Kipchoge given the tragedy around Kiptum, but it’s highly likely that Kipchoge would break the vast majority of AG records if he kept running. Or, put another way, if he started somehow running only at 30 and then realized his peak much later. But because of incentives, etc. that is almost never going to happen that the world’s fastest man in general is going to be running that late in his life. They will realize talent earlier in life and maximize their value when they are 25-35, not 45-55.

          5K: 16:37 (11/20)  |  10K: 34:49 (10/19)  |  HM: 1:14:57 (5/22)  |  FM: 2:36:31 (12/19) 

           

           

          JMac11


          RIP Milkman


            The next time I go to NYC will be at the funeral of JMac. I will be 89yo and during the trip I will try and find a 5k race and win the 85-90yo AG award in honour of him. Respect. 🙏

             

            I still think the greatest insult was when RP said he’ll help Cal start CIM when he’s older by pushing him in his wheelchair down all of the hills. Still laugh when I think about that.

            5K: 16:37 (11/20)  |  10K: 34:49 (10/19)  |  HM: 1:14:57 (5/22)  |  FM: 2:36:31 (12/19) 

             

             

            JMac11


            RIP Milkman

              Good luck with the Masters program that sounds like a good step! he rubbed me the wrong way because he started insulting older folks as stupid but then made a fool of himself wrongly using math jargon. Sample size, skewness and standard errors are not the issue here. Anyways I will leave it here - agree the older AGs are probably biased and make the moose mug easier to achieve; had never put much thinking into it but that makes sense!

               

              Your reading comprehension skills continue to amaze me. Maybe I should age grade your posts though, will make a lot more sense then.

              5K: 16:37 (11/20)  |  10K: 34:49 (10/19)  |  HM: 1:14:57 (5/22)  |  FM: 2:36:31 (12/19) 

               

               

              JMac11


              RIP Milkman

                 

                There is an argument that says with fewer people running in older AG's, somehow the record is less valid, but I doubt you could suggest that makes a lot of difference.  And interestingly, the Sally Gibbs example comes up again because she is setting older AG records having set none when she was younger.

                 

                I think that’s actually a perfect example right? How many AG holders set records when they were younger? I imagine it’s very few.

                 

                The record is absolutely valid. As you said earlier, you can only race the people that show up. You would never tell someone with the AG 65 year old record that their record is somehow tainted. But, with AG, they could argue they are “as good” as Kiptum’s record, which is the point I’ve always made around AG as being crazy. No AG record holder would do that. But there have been posters on here who somehow think because they run an 80% AG at older ages, that they can convert those times to what it means at younger ages and directly compare themselves to those times as being “better.” That’s where I’ve drawn the line.

                5K: 16:37 (11/20)  |  10K: 34:49 (10/19)  |  HM: 1:14:57 (5/22)  |  FM: 2:36:31 (12/19) 

                 

                 

                SteveChCh


                Hot Weather Complainer

                  I've never seen more people arguing while basically agreeing with each other before

                  5km: 18:34 11/23 │ 10km: 39:10 8/23 │ HM: 1:26:48 9/23 │ M: 3:29:54 6/24

                   

                  2024 Races:

                  Motorway Half Marathon February 25, 2024 1:29:55

                  Christchurch Half-Marathon April 21, 2024 1:27:34

                  Selwyn Marathon June 2, 2024 DNF

                  Wellington Marathon June 23, 2024 3:29:54

                  Hagley Park Run July 20, 2024

                  Dunedin Half Marathon September 15, 2024

                  Timaru Ten October 26, 2024

                  JMac11


                  RIP Milkman

                    Are you new to the Internet?

                     

                    I've never seen more people arguing while basically agreeing with each other before

                    5K: 16:37 (11/20)  |  10K: 34:49 (10/19)  |  HM: 1:14:57 (5/22)  |  FM: 2:36:31 (12/19) 

                     

                     

                    Marky_Mark_17


                       But there have been posters on here who somehow think because they run an 80% AG at older ages, that they can convert those times to what it means at younger ages and directly compare themselves to those times as being “better.” That’s where I’ve drawn the line.

                       

                      Yeah I don't think you can extrapolate that at all.  I see those AG things as an output.  That's it.  You ran a 75%, or an 85%, or whatever it happens to be, at a particular age.  In no world do I think you can sensibly extrapolate times for yourself at a future or past age based on that particular percentile (and nor do I personally think there is any point doing so).

                       

                      But, to your point on technical statistical stuff, I would argue that probably doesn't make a lot of difference as long as the 100% time is pulled from a large enough sample (even if not nearly as large as the open records).  It would have to be off by a large margin before it made any meaningful difference.

                      3,000m: 9:07.7 (Nov-21) | 5,000m: 15:39 (Dec-19) | 10,000m: 32:34 (Mar-20)  

                      10km: 33:15 (Sep-19) | HM: 1:09:41 (May-21)* | FM: 2:41:41 (Oct-20)

                      * Net downhill course

                      Last race: Runway5 / National 5k Champs, 16:22, National Masters AG Champ!

                      Up next: Still working on that...

                      "CONSISTENCY IS KING"

                      wcrunner2


                      Are we there, yet?

                        Playing around with the AG calculator, I just discovered something that puts into question some of my arguments.  Interestingly it may also undercut all of JMac's arguments as well.  I tried to compare the age standard at 10K in 5 year increments beginning with 40 with the world best for each age group.  The age std that the calculator uses is faster in every case and is directly computed using the age factor.  So it seems the creators and maintainers of the age grading tables put more stock in their statistical analysis of results from which they derive the age factor than in the actual age group records.

                         2024 Races:

                              03/09 - Livingston Oval Ultra 6-Hour, 22.88 miles

                              05/11 - D3 50K, 9:11:09
                              06/17 - 6 Days in the Dome 12-Hour, 35.82 miles
                              10/12 - Hainesport 12-Hour

                         

                         

                             

                        Mikkey


                        Mmmm Bop

                           

                          I still think the greatest insult was when RP said he’ll help Cal start CIM when he’s older by pushing him in his wheelchair down all of the hills. Still laugh when I think about that.

                           

                          I must’ve missed that one. 😂😂😂

                           

                          But when will it happen, this year? next year?

                          5k - 17:53 (4/19)   10k - 37:53 (11/18)   Half - 1:23:18 (4/19)   Full - 2:50:43 (4/19)

                             

                            Your reading comprehension skills continue to amaze me. Maybe I should age grade your posts though, will make a lot more sense then.

                             

                            At first you attacked old folks for their lack of math, then for being slow, now its reading comprehension. You were probably a bully in school, trying to desperately find a leverage to use on your opponents. Ageism at its best.

                             

                            Going back to your mathematical argument,which in my opinion is wrong: the record holder is a single point, not the average in the tail (which makes me believe you are involved in options pricing, or insurance costing, where these concepts are frequently used).

                             

                            Population of runners is not to be misunderstood with world population. I think we should move on from this, there is no point. Have fund calculating tail averages.

                            HM: 1:47 (9/20) I FM: 3:53:11 (9/23)

                             

                            2024 Goals: run a FM & HM + stay healthy!

                            shouldbedeleted


                              ....

                              CommanderKeen


                              Cobra Commander Keen

                                Mulberries are the greatest of all berries. Fight me!

                                5k: 17:58 11/22 │ 10k: 37:55 9/21 │ HM: 1:23:22 4/22 │ M: 2:56:05 12/22

                                 

                                Upcoming Races:

                                 

                                November 2  - Crossroads Marathon