As Is Customary

17 04 2016

Your Blogmeister’s Hotel Room

One of my prob/stat professors taught me, and everyone that ever took her course, I presume, that a very quick and dirty way to estimate the standard deviation of a data set is to subtract the bottom bookend from the top bookend and divide by six.  Because virtually all data points (99.74%, according to the normal distribution) are within three standard deviations of the median.

I finally figured out this is why we still hang on to miles per hour and Fahrenheit.  MPH works very well for ground level automobile travel, and F works well for mid-latitude temperate climate air temperature, because there are 100 gradations ranging from zero to one hundred between almost all of the real world practical possibilities.  Almost all of the time, the temperature is between 0 and 100 F, any deviation on either side is considered newsworthy and noteworthy.  As a bonus, Earth’s average temperature is 59 F, a little more than halfway up from the bottom end to the top end.  Almost all of the time while traveling in an automobile, you’re going some number above 0 MPH but you very rarely top 100 MPH.  To wit, the highest posted American speed limit is 85 MPH, close to the top bookend.

In the metric system, this neat centigradation does not present in either scenario.


Actions

Information

8 responses

17 04 2016
Alex the Goon

Metric is even worse for MPG calcs. 40 mpg = 16.5 km/l, which doesn’t sound too impressive. 20mpg = 8 km/l sounds even worse. I think the mileage weenies also use “liters per 100km”, which is just retarded.

17 04 2016
countenance

I had always assumed that metric fuel economy was KM per L. Except I found out that they totally turn the convention around. Instead of a variable distance given a fixed amount a fuel, it fixes the distance then tells us the variable volume of fuel used. Like you said, L/100km.

Yes, it seems retarded, but I think I know the practicality of it. Here, it’s a very long way between major cities, and gas, relative to other places in the world, isn’t expensive, so the crucial knowledge is to know how far you can go on your whole tank or your prorated portion, i.e. can I make it to Denver with what I have left in the tank while I just blew past Salina, Kansas? In Europe, distances between cities is far less, but gas is way more expensive, so the paramount knowledge is how much you’ll have to spend if you want to go from Dusseldorf to Rotterdam. I.e. 220 km between them, a 30 highway mileage car is 7.8 L/100km, 1.43 Euro per L gas price in Dusseldorf right now, so 2.2 x 7.8 x 1.43 means you’re plonking out 25 Euros for the one way trip.

17 04 2016
Alex the Goon

The other problem with L/100km is your result will be in the single digits, plus decimals. So you see guys bragging about their 5.5 L/100 rating. 5.5, 6.5, 3.3, who cares?

19 04 2016
Adoll

This extends to most of the imperial system, as they were derived organically and reasonably uniform without formal worldwide agreement, (an inch is your thumb, a foot is your foot, a yard is a step, a cup is a cup, a mile is 1000 paces by a roman legion). Whereas the metric system was cut from whole cloth and based on factors which only make sense to an esoteric french autist (The meter intended to be one-ten-millionth of the circumference of the earth, but the frenchmen initially calculating it fucked up, it was later changed to the wavelength of kypton-86 radiation. None of this is fathomable to the common man.)

As an engineer, I use a variety of units daily. I find people who push for the metric system tend to have no argument aside from being bad at fractions.

19 04 2016
countenance

The other impetus for the metric system is the bandwagon, that the rest of the world does it.

We know “mile” comes from “mille passum,” that is, “thousand steps.” The reason it became 5280 feet was another practical reason: Eight furlongs, a furlong being the average distance an average horse could pull a plow before stopping, 660 feet. 8 x 660 = 5280. So the new mile definition became practical for land measurement.

19 04 2016
Adoll

In the end, you are correct, they were all uniformly defined. Nothing wrong with that.

As for “the rest of the world uses it”, the rest of the world forces their populations to use it. You don’t need to don a black and yellow bowtie to think this is tyrannical:
http://www.telegraph.co.uk/news/uknews/3179333/Market-trader-in-shock-at-conviction-for-selling-fruit-and-veg-by-the-pound.html

The metric system is also in fact the official system of the US per the US federal government, the government just does not force the people to use any one system of measurement (freedom, fuck yeah).

(The day they all the rest of the world abandons their languages and adopts english will be the day I will almost consider using metric. Almost)

19 04 2016
countenance

Going back to my original thoughts, the reason why C sucks for mid-latitude temperate climate air temperature is that, in F, the six standard deviation relevant range is 0 to 100, Earth average is 59, and ideal comfort level is 70 (depending on the person). In C, the relevant range is -20 to 40 (rounding), Earth average is 15 and ideal comfort level is 21. Huh?

17 05 2016
Super Zips, Revisited | Countenance Blog

[…] (I do have a tweet into Charles Murray himself, hopefully he’ll know), but I do remember the Rule of Six when you need a quick and dirty standard deviation:  Top minus bottom divide by six.  This means […]




%d bloggers like this: