Wednesday (New Years) Highlights

Have fun with those resolutions. (Which are so often summed up via Spike Lee’s enjoiner “Do the Right Thing”).

  1. Well, in January, it seems I’ll be following the rich and famous (that is to say having a my colon scoped). My results will not likely reported by ABC News.
  2. Without cause. Oh, goody.
  3. Good news presented as bad.
  4. On the hand, perhaps “if true” needs to be appended to the above.
  5. Father Time is a T. Rex.
  6. Who is surprised?
  7. What a perfect landing looks like.
  8. I’d share some of the other links, but my guess none of my readers would be interested in things like this.

Posting in general seems to have slowed. Have a good New Years day!

One response to “Wednesday (New Years) Highlights

  1. Re #4

    In terms of data analysis, imagine you had your daily bank balance from the day you were a teenager with your first bank account to today…what would we be looking at.

    Well clearly you’d have short term variation. Probably with a high around your payday and a low around the day before the next payday.

    You’d have seasonal variation. Perhaps highs around the time of your tax refund or bonuses (if you had jobs that had yearly bonuses). Lows perhaps right after Christmas or your yearly vacation.

    You’d have a long term trend. You probably started your life with little income and have been gradually increasing ever since.

    Then you might have one shot random events. A huge windfall from that trip to Atlantic City 15 years ago, when you sold your house, when someone stole your debit card and wiped out your account and it took a week to get the bank to give you the money back etc.

    If we’re talking about global warming, we are talking about a data set that is as complicated as the above. If your average bank balance started setting ‘more record lows’ than ‘record highs’ for a period of time, would that demonstrate your yearly income was not still increasing? Assuming I only could look at your daily bank balances and not your paystubs or tax returns, would that factoid really say much to me? I don’t think so.

    But what if figuring out your income wasn’t our goal. Suppose we are trying to pull a con. Perhaps trying to convince a rich relative of yours to write you into her will, we want to present her with ‘facts’ that are technically true but nonetheless present a picture of someone whose income has not been increasing on a regular basis? Well we might then take the data set and start mining it.

    “Look Aunt Helen, the # of days that Mark’s balance has fallen by more than 50% from the previous days has gone up over the last 8 years. Look, the average balance on alternate Fridays has fallen over the last 5 years! Look the # of record low balance days is more than the # of record high balance days over the last 15 years!”

    All these may be true facts but the picture they are being enlisted to serve, that Mark is suffering financial hardship and Helen should give him a bigger portion in her will, is false.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>