Stupid Measurements

There are: The World Almanac and Book of Facts, 2001 reports that a foot is 12 inches exactly, and a rod is 16.5 feet exactly. It makes no mention of cubits at all, I checked quite thoroughly.

I have a book called Schott's Original Miscellany by a chap called Ben Schott, published November 2002. It's full of excellent things like this. I particularly like the full wording of the Riot Act... The reason you "Read someone the Riot Act" is because quite often people would forget the "God Save the King" bit from the end and case against the rioters would be thrown out because the Riot Act had not been correctly served :-) You can see the book on Amazon at --MattStephenson?

Why is an Acre not square?? My calculator tells me that an acre is 43,560 square feet (or 4,840 square yards). However, the square root of 43,560 is 208.71032557111... and the square root of 4,840 is 69.570108523704...

On the other hand, a square mile is exactly 640 acres! So a square (mile) area comprises an integral number of non-square rectangular shapes (acres) each having an integral number of square feet and yard equivalents?? But the acres themselves are not square?? Hello??

An acre is one chain by ten chains, that's 22 yards by 220 yards. Historically, it's the strip of land that could be ploughed in a day by one man (and his horse), and it's long and thin because they packed into a field well. Hence the square mile (1,760 yards x 1,760 yards) can be divided into 80 x 8 = 640 one-acre rectangles of 22 yards x 220 yards each (i.e., 1 chain x 1 furlong each; "furlong" derives from "furrow long").

Ah, so a square mile, being 80 chains long on a side, divides nicely up into a grid that's 80 x 22 yards one way and 8 x 220 yards the other way, and 80 x 8 is very neatly 640 (acres). I'd never noticed the relationship between chains and miles.

Most of the U.S. is divided up into "quarter sections" (1/2 mile by 1/2 mile, 40 chains by 40 chains, an area of 160 acres). As far as I can tell, the U.S. government expected individual farmers to occupy a 1/2 by 1/2 mile square in Kansas, Oklahoma, and I assume other states as well. Naturally we have 4 "quarter sections" in a section (1 square mile by 1 square mile). Then we have 36 sections in a "township" (a square, 6 miles by 6 miles, roughly 0.1 degrees of latitude). A lot of roads were built on section lines, so we often have main roads exactly 1 mile apart, and twisted tangled "residential streets" inside each 1 mile by 1 mile square.

The roads going East-West are straight and follow lines of latitude. The roads going North-South follow lines of longitude. Most roads going North-South have a strange little jog every _??__ miles where it intersects the East-West section line road. This is because if you drive 60 miles north, 60 miles west, 60 miles south, you'll have to drive about 61 miles east to return to your starting point. Somewhere in the middle of that "square" is an east-west street that has 61 individual roads (one at each section line) going south from some east-west road, and 60 individual roads going north (one at each section line). It wouldn't help any to use the metric system; this is an inherent problem with trying to tile the globe with square tiles. (See HexGridSphere for other ways of tiling the sphere).

And from "Good Omens" by TerryPratchett and NeilGaiman:

Note for young people and Americans: [...] the original British monetary system:

Two farthings = One Ha'penny. Two ha'pennies = One Penny. Three pennies = A Thruppenny Bit. Two Thruppences = A Sixpence. Two Sixpences = One Shilling, or Bob. Two Bob = A Florin. One Florin and one Sixpence = Half a Crown. Four Half Crowns = Ten Bob Note. Two Ten Bob Notes = One Pound (or 240 pennies). One Pound and One Shilling = One Guinea.

The British resisted decimalized currency for a long time because they thought it was too complicated.

What you're used to always seems simpler. Think about making the change to decimal time!

The most peculiar units of mensuration I ever encountered were in the tin can business. Tinplate, the stuff of which tin cans are made (it's mostly steel) is sold in "packages" -- a package is 112 sheets -- and priced by the basis box, which is a unit of area equal to 112 sheets each 14 x 20 inches. The size of an actual sheet of tinplate determines its basis box ratio, which is the area of the sheet in square inches divided by 280. Thickness or gauge is specified by basis weight, which is the weight of a basis box in that thickness; an example is the once-familiar three-pound coffee can, which had a 107-weight body and 95-weight ends. Thickness of tin coating is also expressed in pounds per basis box, per side, as the coating on each side may differ.

This leads us into the economics of pricing tin, which is generally quoted in units of Malay dollars per picul. A picul is generally 133-1/3 lbs., except for the royal Thai picul, which is twice that big...

When my dad was a Caltech student in the 1940's, one especially annoying and tormenting physics professor made the mistake of letting students select any units they wished for answering questions on the final. As an interesting coincidence, every paper used a different system of measurement. Dad especially remembers FurlongsPerFortnight as one of them.


This system was originally from England. The US has abandoned mils, cubits, rods and to some degree sea miles, but the descendants of the inventors of the complete system think WE'RE the crazy ones! Because we gave it up, while you struggle on!

We certainly haven't given up mils entirely. I have no idea how widespread use is, but I know that when I was ordering capillary tubing for a physics lab, all diameters were in mils. -- AdamBerger

And rods are still used for canoe portage distances.

My dictionary gives geographical or NauticalMile one minute of longitude measured along the equator - 6082.66 feet: in British practice, Admiralty measured mile, 6080 feet; international NauticalMile - official unit in U.S. since 1954 - 6076.1033 feet. I've read elsewhere that a knot (one nautical mile per hour) is usually 6080 feet per hour. (now actually 1852 metres per hour)

What is so stupid about these measurements? They result from natural unit sizes appropriate to what is being measured. What is stupid is force fitting everything into some decimal based system of measurement due to someone's sense of conformity.

It isn't about conformity. The SI is simply a far superior system of measurement. Now I agree with other posts here that much of the benefit is lost when you are talking about small scale guesswork, and not actual measurement. On the other hand, I suspect that your "natural" simply means "what I grew up with". Have you ever talked to kids who grew up in an SI country? They are just as comfortable with metric units as you and I may be with Imperial. In most cases, there is *nothing* to choose between the two : a mile is no more natural than a km, an inch no more natural than a cm. There are some units that are more 'unnatural' but it isn't a big deal. The problem with Imperial is that if you grow up using it, then you want to keep using it when you are an engineer/architect/whatever - Imperial is inferior for these applications, and we are stuck with it due to momentum.

Throwing a rope in the water, with a board attached to the end, and then counting the number of "knots" in the rope after a fixed period of time is natural; very natural. OK, now we have "better" ways, but knots were very natural. ;-> Better? I dunno. It's very satisfying to sail without reliance on complex gadgets and external resources. Measuring nautical miles using the latitude scale at the side of the chart is very convenient. Measuring depth by fathoms (6 feet, approximately the spread of a man's arms) is very convenient. It's the SimplestThingThatCouldPossiblyWork.

Then there are the really StupidMeasurements, like MIPS and Megahertz. What is stupid about these? The measurements are sensible, they are sometimes used is very silly ways. Compare to the above, which are examples from a brain-damaged system of measurement (imperial)

There's nothing brain-damaged about the Imperial/US Customary system. It is a very sophisticated system, well suited for its purpose. That purpose being everyday use by people performing craft activities without special measuring equipment. It is pretty hopeless for automated scientific computation, but that's a relatively recent human activity. For almost all it's history the Imperial/Customary system was very flexible and effective.

A pound of most everyday materials is a couple of hands-full. It's easy to divide a pound into ounces (avoirdupois) by eye since there are a power of 2 ounces per pound (and drams per ounce, as it happens) Kilos are too large, grams too small (and you need a balance to do the subdivision). A pint of most everyday fluids is about as much as you want in one go. Aside: the full-dress Imperial system of volume measurement runs in binary, so has the same convenient subdivision properties as pounds and ounces. Yards, rods, chains and furlongs relate well to areas measured in acres. And so on.

I agree, brain-damaged is too harsh. The proper description should have been 'obsolete'. SI does at least as good a job of all the day to day stuff... if you find kilos too big for a material, 10g or 100g is probably just fine - and dead easy to subdivide by eye (ok, slightly harder than repeated division by two, but not noticeably). I find a litre a more convenient 'one-go' amount than a pint. Etc. etc., the point being that there is no real advantage (other than familiarity) to imperial. The real issue is that it is silly to have competing standards these days, and it would be inane to standardize on anything other than SI at this point. It's a huge waste of energy, time, and effort to maintain the infrastructure around multiple sets of fasteners, architectural standards, etc. *That* is the brain-damage!

(And it leads to software errors.)

You can down a full litre in one go? Man, I'm impressed. ;-)

It is *hot* here. I probably drink about 3 litres of water/day :) Otoh, I see where you are coming from ... but that is what 1/2 litres are for!

The problems with the imperial system go deeper than odd ratios. When Canada was still using imperial measurements, I am told you would every now and then run into a man who would drive to the US to buy gas, because it was that much cheaper. Only it wasn't really - the gallons were different sizes. Depending on where you were, there were three kinds of mile and four kinds of feet. A metre is a metre everywhere. Isn't that worth the change? -- this points to the historical reason for the change to metric as previously (18th century france) not just different countries but different towns would have feet of different length, pounds of different weight etc. what the actual sizes are doesnt matter its the fact that you can rely on them being the same standard everywhere that is. imperial units are also fine unless you need to interface with some one using metric, or imperial units of a slighlty different size, see also -- JamesKeogh

there is no real advantage to imperial.

It is far easier to estimate and imply a degree of accuracy in imperial measures. If a say someone is 6 feet tall, it immediately expresses an approximate size, i.e. +/- 1/2 foot. If I use meters, I need to start adding in fractional digits, while if I use centimeters, I need to use quantities of hundreds. In both cases, it is not apparent how accurate the estimate is intended to be.

That is ridiculous. You just claimed that it was a problem to use fractional approximation, after using a fractional approximation. You, like many of us, are simply comfortable with feet. They are no more natural than metres, and have significant disadvantages. Not long from now, with any luck, your descendants will find feet to be a historical oddity like furlong is to us. Most of these systemic changes require a generation or two to die off before they are well seated; history shows this again and again

The desired precision in an estimated height does make centimetres too small and metres too big. Of course, if you really cared you could use decimetres. But not all units are going to be absolutely convenient for all purposes, and that's not an excuse to use different units each time. Does anyone remember that pounds of feathers and gold are not the same weight? That's undoubtedly more convenient for the people in question; it's also stupid.

Why is it stupid to use the more convenient measure? Why would it be more intelligent to use a less convenient measure?

If you don't see why it's stupid to have the weight of a pound depend on what you are weighing, I can't explain it to you.

This looks like a blooming flame war...

No, the stupidest measurement is SLOC--SourceLinesOfCode?. I kept telling them that they should count source lines of code not written. And then they wouldn't let me add a test suite because it illegitimately added to the SLOC count. -- EricJablow

I suspect that one customer used to measure SLOC, judging by their CodingStandard. Whenever possible, it specifies vertical whitespace. -- AnonymousDonor

Hard drives have usually been promoted by advertising the average seek time, and CD-ROM drives are measured against their maximum sustained data transfer rate. A more meaningful description of these devices capabilities would be provided by the reverse, MST for HD and ST for CD-ROM drives.

Speaking of hard drives, I was dismayed to find my hard drive advertised as 30G to only actually be 28G. (Which reminds me that Windows Explorer considers 1000 bytes to be a kilobyte.)

they are probably following this

Networks are usually rated only in terms of capacity (bandwidth), ignoring how quickly they respond to a request (latency). How usable is a 1 Gb/sec connection that takes five seconds to begin each file transfer?

More on latency vs. bandwidth:

"It's the Latency, Stupid" :

In the engineering world, it's now common to buy lengths of pipe in metric, but the diametres are still imperial!

When I design a printed wiring board (PCB) in the U.S., the width of the printed wires is specified in mils (1000 mils to the inch). The thickness of the printed wires is "ounces" (actually "ounces of copper per square foot"). Most PCBs (in 2001) have printed wires with a thickness of "1/2 ounce" (17 um). "1 ounce" (0.035 mm) is also very common. I'm starting to see a few places specify thickness in "grams of copper per m^2". Really. I couldn't make up something this silly. -- DavidCary
In the UK, timber (lumber) used to be sold in feet. Then we went metric and sold it in metres. But not any old meters, you can't just buy 2 metres. Oh no! It has to be in multiples of 0.3, 0.3m being a kind of 'metric foot'. So instead of buying 8 foot, or 12 foot, we now buy 2.4m or 3.6m - completely different ;-) - StephenHutchinson

In New Zealand, in the gelcoat and paint industries, people use the units 'mils' and 'microns' to describe film thicknesses, although I find that because of the confusion caused by 'mils', (sounds like the abbreviation of millimetres), that when speaking in imperial, it is better to speak of 'thou' which is equivalent to mils (1000th of an inch). - Oliver Jackson-Hill

Compare to: WhimsicalUnitsOfMeasurement

View edit of January 24, 2005 or FindPage with title or text search