Stories
Slash Boxes
Comments
NOTE: use Perl; is on undef hiatus. You can read content, but you can't post it. More info will be forthcoming forthcomingly.

All the Perl that's Practical to Extract and Report

use Perl Log In

Log In

[ Create a new account ]

Robrt (1414)

Robrt
  (email not shown publicly)

robert at perl dot org

Journal of Robrt (1414)

Friday December 05, 2003
12:00 AM

Meaningless Statistics

[ #16180 ]

UPS delivers about 13 million packages and documents a day. If 1/100 of a percent (.0001) are lost, thats 1300 packages a day. Which is a non-trivial number if the package is mine.

I wonder what the actual "lost or delayed" statistic is.

The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
 Full
 Abbreviated
 Hidden
More | Login | Reply
Loading... please wait.
  • If the statistic is accurate, it will almost never be your package that is lost.

    Tolerance of some problems in complicated systems is only reasonable. These are humans and there is a cost to perfection. Would you rather pay ten times as much and for them to lose only 1 package in 100,000? Would you perhaps prefer to pay 100 times as much and them to lose 1 package in a 1,000,000?

    The actual cost and tolerance for error is determined in the market. People will pay for the service level they can tolerate.

    • Would you rather pay ten times as much and for them to lose only 1 package in 100,000?
      Heh, a gourmet parcel service...I can see it now... :)
      --

      ------------------------------
      You are what you think.
  • If they lose 1/100th of a percent, then their reliabiliity rate is 4 sigma: .9999, which is pretty damn good.

    If the did manage to improve tenfold and reach 5 sigma, or one lost package per hundred thousand packages, that would be truly amazing. It's hard to imagine UPS doing much better than that.

    But they'd still lose or misplace 130 packages a day, one of which could be yours, which still wouldn't make you feel any better.

    For example, at Pearson Int'l Airport in Toronto, they lose, misplace or delay

    • For example, at Pearson Int'l Airport in Toronto, they lose, misplace or delay about 300% of the luggage that moves through the airport.
      I presume that's hyperbole, because to actually accomplish that, they'd need to lose the luggage of two other nearby airports as well. Entirely.

      Hard to imagine that.

      --
      • Randal L. Schwartz
      • Stonehenge
      • No, that statistic is entirely correct. They lose a lot of luggage at Pearson. They built the airport on top of a singularity or something.
        • > No, that statistic is entirely correct.

          Somebody is pulling 7600% leg here. They cannot lose more than 100%. Period.
          • Three separate things: lose, misplace, delay. If they lose, misplace, AND delay each piece of luggage, then it can be 300%!
    • For example, at Pearson Int'l Airport in Toronto, they lose, misplace or delay about 300% of the luggage that moves through the airport.
      That's an impressive statistic. Do they go to other airports to get luggage to lose? That might explain why so many airports claim very good loss rates, yet I always seem to lose my luggage there - Pearson Int'l is losing my luggage.

      :) Was that supposed to be 0.3% lost?

      • Nope. Not 0.3%. 300%.

        It's certainly an impressive feat. If you even think about your luggage going through Pearson, it gets lost. Even if you're just on a puddle jumper from Heathrow to Gatwick.

        And all of those cases of people losing luggage through Denver Int'l Airport? Macroscopic tunnelling. Really.

    • You can't base their reliability on the number of packages lost alone. Since they 'guarantee' on time delivery, you'd also have to factor in late deliveries.

      As I said, it's all based on what people are willing to tolerate. If an airline lost 1 in 10,000 passengers , that would be considered completely unacceptable, however, people seem to tolerate being late due to airline delays on 1 or even 2 in 3 trips.