Announcement

Collapse
No announcement yet.

To GH (or other expert): Please explain validity of 1.08 ...

Collapse

Unconfigured Ad Widget

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

  • To GH (or other expert): Please explain validity of 1.08 ...

    ... conversion factor for 1500-mile and 3k-2M?

    I've always used it, trusted it and intuitively understood it, but I can't convince several folks that I'm debating with that it's valid, no matter what speed your running and how old or experienced you are. They're trying to use examples of HSers who've run 8:10 but haven't broken 9 and stuff like that.

    I argue that those personal stats are just coincidental, but the fact is if you run a 5-minute 1500, you'll get 5:24 for a mile with the same effort - all other things being equal and regardless of any other factors...

    Please help, you true scientific experts of track out there!

  • #2
    Re: To GH (or other expert): Please explain validity of 1.08

    Hmmm... first thing I'll say is that nobody is making the claim that this is a 100% accurate measure. 12 inches in a foot, yes, but 1.08 from 1500 to mile, no. So they need to know that no conversion factor involving race distances would ever be passed by the Dept. of Weights & Measures. I'm sure you realize that, but perhaps it's a preliminary step towards convincing your doubting friends.

    The next thing is the general theory of conversion factors. If two distances are very close together (like 400m and 440y--which is 402.34m) then you simply use a multiplier which is a ratio of one distance to other (400/402.34=0.99414023). So if we want to what a 45.00 time for 440y from the old days is worth as a 400m time, we multiply it by 0.9942 (gotta truncate those decimals somewhere). 45.00x0.9942=44.739, which is 44.74. The world's statisticians have no trouble running "combined" lists with converted marks like that. Same for 220/200, 880/800.

    But when you get up to the 1500/mile, not only are you talking about a relatively long gap between events (1500m vs 1609.35), you also know that there's a significant fatigue factor that has to be taken into account. So, a 1609.35/1500 gives you a sans-fatigue ratio of 1.0729. Multiply a 4:00.0 in the 1500 by that, for example, and you get 4:17.5 for a mile. Multiply 4:00.0 by the 1.08 "everyone" uses and you get 4:19.2. You can see how quickly what look like minor decimals have a major effect.

    So, where did the numbers come from that changed the multiplier from the 1.0729 to 1.08? Beats me! No, seriously, I actually had nothing to do with that work, even though I was the T&FN statistician at the time it was popularized. I believe T&FN founder Bert Nelson was one who was very instrumental in the research. But I do know that many years of data, and many races, were analyzed. I have also seen a graph which plots the 1500 WR against the mile WR through the years (conveniently, usually broken by the same people), and if you draw a line through the dots, it pretty much runs on 1.08. So, if you want "validity" that's all the evidence there is.

    But your post inherently had two questions (anybody still with me here?), since you mentioned the 3K/2M as well. Unfortunately, there virtually no relevant data on real races, because the same people didn't run them enough. So the use of 1.08 there makes what some consider a great leap of faith, that the fatigue factor is the same, even though it probably isn't. It's not remotely as reliable as at 1500/M, but it's the best anybody's come up with.

    As to your friends talking about HSers who run 8:10 but can't break 9:00, I would suggest that that's quite possibly just a sampling error. Unless they've run each of hte distances multiple times under similar conditions, you run into anomalies. Just as if you ran a 4:00 in the 1500 with regularity in a cool area, but the only time you ran the mile it was 90 degrees and windy and you ran 4:25 and said, "man, that 1.08 makes no sense at all."

    OK, getting carpal tunnel now...

    Comment


    • #3
      Re: To GH (or other expert): Please explain validity of 1.08

      The big problem is, of course, how fast are they finishing? In a slow tactical race, they are really at top sprint speed and in a rabbited stiff pace, they are all dying. That alone can be at least a second difference between the two. I have always seen the 1.08 as the least worst conversion factor, taking into consideration that virtually no one can run all all-out mile at an all-out 1500 pace.

      Comment


      • #4
        Re: To GH (or other expert): Please explain validity of 1.08

        No, that's a common misconception. The conversion factor isn't intended to tell you how fast they would have run a mile had they continued on from the 1500 to the mile. It tells you the "relative value" of the two times. If they were running a mile instead of the 1500, they would have started to kick that much later and and he dieing part of it would have kicked in later and they would still cross the line at the same speed.

        Think of it as the difference in distance between the two races as coming at the START of the race, not the end. The data that was used to generate the 1.08 compared actually 1500 races to actual mile races, not 1500s en route to miles. Make more sense now?

        Comment


        • #5
          Re: To GH (or other expert): Please explain validity of 1.08

          >The big problem is, of course, how fast are they
          finishing? In a slow tactical race, they are
          really at top sprint speed and in a rabbited
          stiff pace, they are all dying. That alone can be
          at least a second difference between the two. I
          have always seen the 1.08 as the least worst
          conversion factor, taking into consideration that
          virtually no one can run all all-out mile at an
          all-out 1500 pace.<

          This is somewhat imponderable. If it's a slow, tactical race, the last 109 meters of the race will be fast whether the race is a mile or 1500m. You could argue that in such a race, the conversion should assume an additional 109 at the slower pace, since presumably the kick will start n meters from the finish, regardless of whether it's a mile or a 1,500. But that may not be a good assumption and in any event, there's no way a single conversion factor could account for all of that.

          I don't know who came up with the 8 percent factor, but it's a nice round number and it gets you close enough.

          There's another approach you can take and that is to use the IAAF scoring tables (the so-called Hungarian tables) to ascertain the point equivalent for a given mark. But that requires owning a book and carrying it around with you. For most purposes, 8 percent works just fine. It's what I use and I think most people do for most purposes.

          Comment


          • #6
            Re: To GH (or other expert): Please explain validity of 1.08

            Obviously, while I was typing my last message (and taking a few phone calls), gh was posting his response to the same effect. It's nice to know we're on the same page.

            Comment


            • #7
              Re: To GH (or other expert): Please explain validity of 1.08

              Garry: Thanks for the detailed explanation. I WAS hoping it was a little more scientific , but I understand what you said and appreciate it.

              BTW, although it has slipped out of relevance since the 70s (except for some HS CC kids), is the "accepted" 3M/5k (and 6M/10k) conversion factor 1.036? That's what I seem to remember from the past.

              Thanks again.

              Comment


              • #8
                Re: To GH (or other expert): Please explain validity of 1.08

                yeah, gh, you burst my bubble. All these years I thought there was some real statistical rationale behind it, and now it turns out that it's a WAG! Not that this is the answer to the universe, but is there no better way?

                Comment


                • #9
                  Re: To GH (or other expert): Please explain validity of 1.08

                  >But your post inherently
                  >had two questions (anybody still with me here?),
                  >since you mentioned the 3K/2M as well.
                  >Unfortunately, there virtually no relevant data
                  >on real races, because the same people didn't run
                  >them enough. So the use of 1.08 there makes what
                  >some consider a great leap of faith, that the
                  >fatigue factor is the same, even though it
                  >probably isn't. It's not remotely as reliable as
                  >at 1500/M, but it's the best anybody's come up
                  >with.

                  So thus, would it be in the realm of possibility to assume the fatigue factor would be less relevant to a strength runner--one whose specialty is in distances longer than a 3km/2M--than someone whose specialty in a middle distance race? It would seem that the lactate-generating finish would be the same for both hypothetical runners but the bulk of the "cruising" pace would mean less fatigue-wise.

                  I ask this because many years back I had a teammate in the 10km. who ran a late season Bowerman style blow-out 1500m. in a small scale meet in 3:58 and the local "experts" were telling the guy it was only worth 4:19 at best because of the almighty fatigue-factor (these guys were jealous jerks anyway). Fortunately, I had the entire race on videotape and simply marked the splits both ways, start to finish and from the finish backwards (as I had seen reported in T&FN!). This showed that the guy ran his first 800m. in 2:08+ and his second 800m.--from 700m. to the finish--in 2:04+, making a 4:12-4:14 for 1600m. I didn't know the ballpark conversions at the time, but I did not believe it was 6-7 seconds, as the "experts" were claiming. I told the guy he ran no worse than 4:16+. This blow-out race had been done for his confidence, too, with the Conference meet up the next week-end. Knowing he ran a PR instead of a high school PR matching time did matter that next week. He was pumped and got a second in the 10km. and third in a blanket finish 5km.
                  I have always wondered how relevant the fatigue factors was to someone stepping down in distance...?

                  Comment


                  • #10
                    Re: To GH (or other expert): Please explain validity of 1.08

                    Fatigue factors from 5k to 10k, or from half-marathon to marathon, are relatively stable from one person to another. Much research has been done on the subject. If a fatigue factor in highly predictable when the distance is increased by 100%, then it stands to reason that it would be even more predictable when the distance is increased by 7.3%.

                    Of course, none of these predictors are perfect. But if you want to combine lists, they're as good as you're going to get.

                    Comment


                    • #11
                      Re: To GH (or other expert): Please explain validity of 1.08

                      I ask this because many years
                      >back I had a teammate in the 10km. who ran a late
                      >season Bowerman style blow-out 1500m. in a small
                      >scale meet in 3:58 and the local "experts" were
                      >telling the guy it was only worth 4:19 at best
                      >because of the almighty fatigue-factor (these
                      >guys were jealous jerks anyway). Fortunately, I
                      >had the entire race on videotape and simply
                      >marked the splits both ways, start to finish and
                      >from the finish backwards (as I had seen reported
                      >in T&FN!). This showed that the guy ran his first
                      >800m. in 2:08+ and his second 800m.--from 700m.
                      >to the finish--in 2:04+, making a 4:12-4:14 for
                      >1600m. I didn't know the ballpark conversions at
                      >the time, but I did not believe it was 6-7
                      >seconds, as the "experts" were claiming. I told
                      >the guy he ran no worse than 4:16+.

                      The splits don't look right: if he ran 2:08 for 0-800 and 2:04 for 700-1500 and finished in 3:58, the 100 in the middle (700-800) would have to take just 14 seconds. That's pretty much impossible, considering the average for the race was only just inside 16 secs. The whole thing just doesn't add up. BTW, according to the Hungarian tables, 3:58.0 is equivalent to around 4:16.8.
                      Było smaszno, a jaszmije smukwijne...

                      Comment

                      Working...
                      X