Announcement

Collapse
No announcement yet.

Mo and all-time 1500-10K scoring

Collapse

Unconfigured Ad Widget

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #16
    Re: Mo and all-time 1500-10K scoring

    It's OK with me that these guys likely don't pay much attention to IAAF scoring tables and comparisons based on them. I'm guessing that they do care about running as fast as they can (given whatever preparation/competition limitations come their way) at whatever distances they race(d), and these scoring table-based comparisons are still interesting to bring together a range of performers and performances.

    To the comments on Komen's 10,000m lifetime best: Similar comments could be made about some of the others, I think. I'm sure Gebrselassie had a faster 1500m in him, & that's probably true of Bekele, too. Across such a range, no athlete is going to show whatever might have been his "eldrickian" best :wink: at every distance. The commitment to being the best 5k/10k runner in history means that attention could not be paid to being one's absolute best at 1500m, and Komen's focus at shorter distances meant he never showed his maximum potential at 10km (and on the USA level, the same could be said of Bob Kennedy).

    To aaronk's post: Your methodology makes about as much sense as adding together 1 million Indian rupees, 1 million US dollars, & 1 million Euros, and then saying they are equal to 3 million of some other currency. There has to be some strategy of conversion and comparison. Otherwise, you are basically saying -- to take an athletics example broader than the 1500m-10k, so as to highlight the point -- 60 seconds of David Rudisha's 800m WR is equivalent to 60 seconds of whichever men's WR/WB marathon you prefer. That 60 seconds of Rudisha's WR is ~60% of his WR performance, whereas 60 seconds of the WR/WB marathon is less than 1% of that performance. Those two 60-second segments can't be worth the same in these sorts of comparisons, but your aggregation of times method does just that. Some voices crying in the wilderness are prophets, and some are just, well, voices in the wilderness. Come in from the wilderness...listen to some of the voices here (not mine, but others).

    Finally, I do have a question about the statistical wizardry that resulted in the total scores listed for Bekele, Geb, Komen, Farah, et al.: I don't consult the IAAF scoring tables much, so I thought that such consultation would be relatively straightforward, requiring primarily reading and addition skills. However, using the IAAF scoring tables (2011 revised edition) on the IAAF site, I do not arrive at the same numbers as the stat wizards. Once again, I have been humbled by the most ordinary things. I have re-checked my reading and math skills, and feel they are appropriate to this task. What obvious thing am I missing? A new edition of the scoring tables? Initiation into the Stat Wizards Club? Perhaps some kind person can provide some explanation or a breakdown of the specific #s for any of the top 4 on this list? Thank you.

    Comment


    • #17
      Re: Mo and all-time 1500-10K scoring

      You're probably looking at the multi-event tables, not the "scoring tables," which are two different animals. This is a frequently made mistake.

      Comment


      • #18
        Re: Mo and all-time 1500-10K scoring

        Originally posted by gh
        You're probably looking at the multi-event tables, not the "scoring tables," which are two different animals. This is a frequently made mistake.
        I'm using the scoring tables (not the multi tables) and also get different numbers :?

        Bekele 5066
        Geb 5064
        Komen 5001
        Mo 4956

        What gives?

        Comment


        • #19
          Re: Mo and all-time 1500-10K scoring

          Thank you for the reply, gh. I looked again, and am sure I am using the "scoring" rather than the multi-event tables. I think this means I have failed to advance past lesson 1 of Stat Wizard 101. Oh well.

          Perhaps some other patient contributor can post the #s that comprise, e.g., Bekele's 5089, so I can learn from my mistakes.

          Comment


          • #20
            Re: Mo and all-time 1500-10K scoring

            Originally posted by LopenUupunut
            Originally posted by gh
            You're probably looking at the multi-event tables, not the "scoring tables," which are two different animals. This is a frequently made mistake.
            I'm using the scoring tables (not the multi tables) and also get different numbers :?

            Bekele 5066
            Geb 5064
            Komen 5001
            Mo 4956


            What gives?
            My #s are about the same as yours. Slight differences probably from my not knowing whether to go 1 # up or down when a mark is between those on the tables. In any case, it's clear to me that you and I are working with the same tables, in basically the same way.

            Comment


            • #21
              Re: Mo and all-time 1500-10K scoring

              Odd that the thread does not have the values in them.

              1. Kenenisa Bekele 3:32.35/7:25.79/12:37.35/26:17.73—5089
              2. Haile Gebrselassie 3:31.76/7:25.09/12:39.36/26:22.75—5085
              3. Daniel Komen 3:29.46/7:20.67/12:39.74/27:38.32—5021
              4. Mo Farah 3:28.81/7:34.47/12:53.11/26:46.57—4976
              5. Eliud Kipchoge 3:50.40y/7:27.66/12:46.53/26:49.02—4973
              6. Salah Hissou 3:33.95/7:28.93/12:50.80/26:38.08—4960

              The differentials are very consistent The five points per event are more than rounding up or down on a point would lead to

              Bekele 5066 + 23
              Geb 5064 +21
              Komen 5001 +20
              Mo 4956 =20

              Comment


              • #22
                Re: Mo and all-time 1500-10K scoring

                Do the multi-event scoring tables have the 3K, 5K, and 10K in them? Wouldn't think so, but maybe there's some "Ultra-deca" out there....400, long jump with ankle weights, shot with the 32 lb ball, HJ with weight vest, 1500, 400H, etc, finishing with a 10K.

                Comment


                • #23
                  Re: Mo and all-time 1500-10K scoring

                  While I don't agree with what you're all saying vis a vis giving points for times, I do have a strong point to make about this.

                  Let's take the top 2 (Bekele & Geb) and Komen.
                  I don't know how many points are allocated for each distance, so I'm going to go event by event.

                  1500
                  Komen is 2.30 secs ahead of Geb, and 2.89 ahead of Bekele.
                  If this were a Dec 1500, and the times were 4:20 range, how many points would Komen gain on the other two?

                  3000
                  Komen is 4.42 secs ahead of Geb and 5.12 ahead of Bekele.
                  So far, he leads two events, total of 6.72 secs ahead of Geb and 8.01 ahead of Bekele.

                  5000
                  Here, Komen LOSES some seconds.....2.39 to Bekele, and 0.38 to Geb.
                  Again, if this were the 3rd event in the Dec, Komen would lose a few points off his lead.

                  10000
                  Now, here's where the MAJOR difference occurs!
                  Komen loses 75.57 seconds to Geb, and even more to Bekele, a massive 80.59 secs!!
                  In an actual 10K race, at the approx pace they're running, Komen would finish about 500 meters behind Geb, and maybe 540 meters or so behind Bekele!!

                  Keep in mind that in the points scheme listed, Komen winds up just 64 points behind Geb and 68 points behind Bekele.
                  After 3 events, Komen would be slightly ahead.
                  But then, after a disastrous 10K, he's just 60+ points behind???

                  At the end of the 5K, Komen was AHEAD by 6.34 secs from Geb, and 5.62 from Bekele.
                  But then, after the 4th event, he's now BEHIND by 69.23 to Geb, and 74.97 from Bekele!!

                  I can maybe understand those 3 being quite similar over the 1500/3K/5K trio of events, and that the points given to them would be quite close.
                  But how can you lose just a few points when you're 500 meters behind in the 10K....and when your overall TIME loss is well over a minute??

                  Comment


                  • #24
                    Re: Mo and all-time 1500-10K scoring

                    Originally posted by DrJay
                    Do the multi-event scoring tables have the 3K, 5K, and 10K in them? Wouldn't think so, but maybe there's some "Ultra-deca" out there....400, long jump with ankle weights, shot with the 32 lb ball, HJ with weight vest, 1500, 400H, etc, finishing with a 10K.
                    Detmer got the 'double deca' which included the 10,000, 5000, 3000SC, 400h, 200, HT, TJ, 800, and ??? and ? So, they definitely do have the tables. What was amazing was that his deca sub-score was something like 7700 for this two-day event.

                    Comment


                    • #25
                      Re: Mo and all-time 1500-10K scoring

                      Originally posted by 26mi235
                      Odd that the thread does not have the values in them.

                      1. Kenenisa Bekele 3:32.35/7:25.79/12:37.35/26:17.73—5089
                      2. Haile Gebrselassie 3:31.76/7:25.09/12:39.36/26:22.75—5085
                      3. Daniel Komen 3:29.46/7:20.67/12:39.74/27:38.32—5021
                      4. Mo Farah 3:28.81/7:34.47/12:53.11/26:46.57—4976
                      5. Eliud Kipchoge 3:50.40y/7:27.66/12:46.53/26:49.02—4973
                      6. Salah Hissou 3:33.95/7:28.93/12:50.80/26:38.08—4960

                      The differentials are very consistent The five points per event are more than rounding up or down on a point would lead to

                      Bekele 5066 + 23
                      Geb 5064 +21
                      Komen 5001 +20
                      Mo 4956 =20
                      26mi235 -- I take this to mean you got these #s from the tables, too? If so, your #s also about the same as the #s LopenUupunut calculated, and what I came up with, too. This keeps me wondering about where the originally reported #s came from.

                      Comment


                      • #26
                        Re: Mo and all-time 1500-10K scoring

                        No, I did not do the calculations. The four at the bottom are copied from above, the top 6 are copied from the front page link. I also just divided the total differentials by the number of events. The implication seems to be that at least one of the events had a scoring change that shifted all scores up. Since there have not been a lot of great times lately in all four the the events they might have lifted up the scoring for all of them or a subset. This is JUST a guess.

                        Comment


                        • #27
                          Re: Mo and all-time 1500-10K scoring

                          Re: Using Times to "Score"
                          In speed skating, the allround championships are scored based on the average times for 500m for each race, and the skaters are ranked by the lowest sum of average times. This is known as the samalog scoring system. For example, if a skater had the following performances for the 4 distances (averages for 500m follow in parentheses):
                          500m 36.0 (36.0)
                          1500m 1:46.5 (35.5 = 1:46.5 / 3)
                          5000m 6:25.0 (38.5 = 6:25.0 / 10)
                          10000m 13:26.0 (40.3 = 13:26.0 / 20)
                          Total Score: 150.3 = 36.0 + 35.5 + 38.5 + 40.3

                          By taking the average for a 500m distance for each race, this takes into account the times relative to the distances. This assumes a linear relationship, which is probably similar to the IAAF scoring tables.
                          Here's how the top scorers compared using samalog system (using 500m as the standard unit of measure):
                          1. Kenenisa Bekele 3:32.35/7:25.79/12:37.35/26:17.73—5089 4:59.7
                          2. Haile Gebrselassie 3:31.76/7:25.09/12:39.36/26:22.75—5085 4:59.8
                          3. Daniel Komen 3:29.46/7:20.67/12:39.74/27:38.32—5021 5:02.2
                          4. Mo Farah 3:28.81/7:34.47/12:53.11/26:46.57—4976 5:03.0
                          5. Eliud Kipchoge 3:50.40y/7:27.66/12:46.53/26:49.02—4973 5:03.3
                          6. Salah Hissou 3:33.95/7:28.93/12:50.80/26:38.08—4960 5:03.1
                          As you can see, the results, from a rank order perspective, are almost exactly the same. The relative point differential is also very similar. This shows that both scoring systems are based on similar assumptions about performances relative to the distances.

                          Comment


                          • #28
                            Re: Mo and all-time 1500-10K scoring

                            The T&FN annual rankings issue for 2012 has a faster 3000m time for Farah listed (something like 7:31.9i+)... I presume it was an interim mark in a 2mile indoor race...

                            Comment


                            • #29
                              Re: Mo and all-time 1500-10K scoring

                              Originally posted by AS
                              The T&FN annual rankings issue for 2012 has a faster 3000m time for Farah listed (something like 7:31.9i+)... I presume it was an interim mark in a 2mile indoor race...
                              Yes, it was 7:31.92i+ and reasonably sure it was the intermediate mark in his 8:08.07 indoor 2 mile (Birmingham 18 February 2012).

                              Also, appreciate Samurai Runner's contribution, just above -- I found that very interesting.

                              (And, by the way, I still can't figure out where the IAAF total numbers -- 5089, etc. -- came from in the original news that prompted this thread. I think 26mi235's guess is probably correct about revised tables.)

                              Comment

                              Working...
                              X