Randomness
Aug. 15th, 2009 01:22 pm![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
rhythmaningA few weeks ago, XKCD comic strip had a very funny cartoon illustrating the randomness of the Microsoft file copy dialog – the box that tells you how long you’ll have to wait to copy all your files over.
Yesterday, I was copying all my photo files over to a new hard drive – my old hard drive having got so full of music and pictures that I had only 3G of the original, so huge I could never possibly fill it, 120G. (This PC is over five years old – I had neither MPn player nor digial camera back then.) I was copying 59G – so I knew it would take a long, long time. Hours.
I set it running, keeping an eye on progress. The dialog box told me how long I would have to wait. It won’t surprise you that this seemed pretty random. After the copy process had been running about an hour or so and just before I set out to leave my flat, leaving the PC running, I decided to write down the predicted time: over about ten minutes (sorry, I wasn’t being so scientific – I should have timed it!), I recorded 78 readings; I didn’t catch them all – sometimes it changed so quickly that I couldn’t write it down fast enough. The readings weren’t evenly spaced, either. Then I stuck the times into Excel. Here is the graph it drew:
It looks pretty random. The mean was 85.5 minutes, the median was 83.5 minutes; the maximum predicted time was 140 minutes, the minimum was 57 minutes and the standard deviation of the data is 21.83 (I can’t remember what the units for SD… sorry!).
I don’t know what the actual time was, since I was out of my flat when it finished.
But that really is a poor process of prediction: the variability from one moment to the next is huge. My guess is that the estimation doesn’t include folders until it actually gets to them. It would make more sense not to show any information at all, since one can’t actually use the information it provides in any meaningful way.
Except create plots, and remember a rather amusing cartoon…
Yesterday, I was copying all my photo files over to a new hard drive – my old hard drive having got so full of music and pictures that I had only 3G of the original, so huge I could never possibly fill it, 120G. (This PC is over five years old – I had neither MPn player nor digial camera back then.) I was copying 59G – so I knew it would take a long, long time. Hours.
I set it running, keeping an eye on progress. The dialog box told me how long I would have to wait. It won’t surprise you that this seemed pretty random. After the copy process had been running about an hour or so and just before I set out to leave my flat, leaving the PC running, I decided to write down the predicted time: over about ten minutes (sorry, I wasn’t being so scientific – I should have timed it!), I recorded 78 readings; I didn’t catch them all – sometimes it changed so quickly that I couldn’t write it down fast enough. The readings weren’t evenly spaced, either. Then I stuck the times into Excel. Here is the graph it drew:
It looks pretty random. The mean was 85.5 minutes, the median was 83.5 minutes; the maximum predicted time was 140 minutes, the minimum was 57 minutes and the standard deviation of the data is 21.83 (I can’t remember what the units for SD… sorry!).
I don’t know what the actual time was, since I was out of my flat when it finished.
But that really is a poor process of prediction: the variability from one moment to the next is huge. My guess is that the estimation doesn’t include folders until it actually gets to them. It would make more sense not to show any information at all, since one can’t actually use the information it provides in any meaningful way.
Except create plots, and remember a rather amusing cartoon…
![[identity profile]](https://www.dreamwidth.org/img/silk/identity/openid.png){kind=small}
no subject
Date: 2009-08-16 10:06 am (UTC)