Announcement

Collapse
No announcement yet.

Something is rotten in the state of statistics

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

  • Something is rotten in the state of statistics

    In the statistics http://www.flightradar24.com/share-statistics the top four feeders have the same stats for Up-time and Max. Range and differ only in Avg. range.
    Ranking.JPG

    So why is F-EIWF1 only in 3rd place when their Avg. Range is higher than F-LBSF1 in 1st and F-KVCB1 in 2nd place?

    And actually, how does 2x695 + 2x235 + 249 add up to 10,891 anyway?

    Stats are only interesting if they are accurate.

  • #2
    I've nearly given up on the stats already.

    The user stats are also flawed showing the sum of all receivers. Surely it would be better to average all the users feeds.

    I also don't understand how the overall score is calculated.

    Comment


    • #3
      I was wondering my own stats.

      Based on statistics-website, the max. range for my receiver is 185nm and the uptime 686h.

      Based on my feed data in FR24 Account-website, the max. range been 191nm and the uptime is 100%.
      I can see only the stats for one week and of course the receiver might been down before, but I have been following last few days and the uptime is just going down at staistics-site. And at the same time at the Account site the stats shows 100%.

      Screenshot from statistics-site:
      Screenshot_2015-11-08_at_8.38.52_PM.png

      Screenshot from my account:
      Screenshot_2015-11-08_at_8.29.56_PM.png

      Comment


      • #4
        So why is F-EIWF1 only in 3rd place when their Avg. Range is higher than F-LBSF1 in 1st and F-KVCB1 in 2nd place?
        I think it's a rounding error. Al max ranges appear to be rounded to the nearest 5 NM.
        F-KDAG1

        Comment


        • #5
          Originally posted by Kpin View Post
          In the statistics http://www.flightradar24.com/share-statistics the top four feeders have the same stats for Up-time and Max. Range and differ only in Avg. range.
          [ATTACH=CONFIG]6652[/ATTACH]

          So why is F-EIWF1 only in 3rd place when their Avg. Range is higher than F-LBSF1 in 1st and F-KVCB1 in 2nd place?

          And actually, how does 2x695 + 2x235 + 249 add up to 10,891 anyway?

          Stats are only interesting if they are accurate.

          Hi,

          Thanks for reporting this.. We are looking into it now.

          Comment


          • #6
            Originally posted by Milliam View Post
            I've nearly given up on the stats already.

            The user stats are also flawed showing the sum of all receivers. Surely it would be better to average all the users feeds.

            I also don't understand how the overall score is calculated.
            Hi Milliam,

            If we were to show average of all the users feed, a user with 5 good installations and one bad would be punished for that. What we want is to encourage users to have as many receivers as possible.

            /Olov

            Comment


            • #7
              Originally posted by Olov View Post
              Hi,

              Thanks for reporting this.. We are looking into it now.
              Tack, Olov.
              And while you are at it. How come nobody seems to reach 30x24=720 hours of uptime? 695 seems to be max.

              Comment


              • #8
                Originally posted by Kpin View Post
                In the statistics http://www.flightradar24.com/share-statistics the top four feeders have the same stats for Up-time and Max. Range and differ only in Avg. range.
                [ATTACH=CONFIG]6652[/ATTACH]

                So why is F-EIWF1 only in 3rd place when their Avg. Range is higher than F-LBSF1 in 1st and F-KVCB1 in 2nd place?

                And actually, how does 2x695 + 2x235 + 249 add up to 10,891 anyway?

                Stats are only interesting if they are accurate.
                Hi,

                Score is calculated by formula: 2 x Uptime + 2 x Max Range + Avg range summed for the last 30 days

                Which means, for each day we calculate daily score: daily_score = 2 x Uptime_daily + 2 x Max_range_daily + Avg_range_daily
                And after that, we sum all daily_score for last 30 days.

                Comment


                • #9
                  Originally posted by Kpin View Post
                  Tack, Olov.
                  And while you are at it. How come nobody seems to reach 30x24=720 hours of uptime? 695 seems to be max.
                  That is a confirmed bug now, will be fixed in the next release. Thanks again for reporting.

                  Comment

                  Working...
                  X