Metric versus US Customary/Imperial for Everyday Use

[Derek]

This is for Metric units versus US customary or Imperial units (I know they're not the same, but they're close enough) for everyday use (not scientific use).

I'm sure I'm in the minority here, but I argue that customary/imperial units are better for everyday use than metric units, for two main reasons:

First, these units are derived from hundreds or even thousands of years of practical tradition. This means that they have been chosen through a sort of natural selection to be convenient to use. Metric, on the other hand, was created by revolutionaries who thought that the faces on trading cards needed to be replaced (and who tried to decimalize time as well, more on that later). This makes metric units rather arbitrary. While they start off alright (the meter is a pretty good middle-distance unit), after converting to different units (mainly through properties of water), they end up being off. The gram is too small, leading to the kilogram being used more often. For example, Fahrenheit gets a lot of ridicule from my experience for its arbitrary boiling and freezing points of water, but its very convenient for measuring daily temperatures, with most falling within the range of 0 (cold) to 100 (hot). The equivalent range for celcius (which of course is based on the boiling and freezing points of water) on the other hand is something like -15 to 40.

The second reason, a derivative of the first, is that customary units divide better than metric. Metric is base 10, which is a terrible base. Most customary units use either base 12 (feet, time) or some binary base (volumes, weights). I say this is a derivative of the first because I'm sure these bases were chosen due to their practicability. Division by two, three, four, etc. is more useful than division by five and ten, so units that allow division by three or multiple divisions by two have become dominant.

I think the best example of this is our measurements of time, the 60 minute (5*12) hour and 24 hour (2*12) day, a system so useful that it resisted metrication and can be traced all the way back to the Babylonians. We can easily divide our hours and days into pretty much any useful fraction. You can't do this with metric measurements. The 60 minute hour also provides a good example of the frequency of dividing by 3 versus 5, since it allows both. When was the last time you measured something in fifths of an hour (12 minutes)? When was the last time you used thirds of an hour (20 minutes)? For me, I use 20 minute intervals as a reference much more often.

The obvious downside of customary units is their inconsistency. In particular, the reason metric originally become so common was that it unified a previously chaotic system of measurements in which neighboring towns might use different units. But there is no reason we couldn't have instead unified around a base 12 system of units based around pre-existing traditional units.


Ok, flame away!

[EvanED]

I find it hard to give input, except one thing: temperature. I really will stand up for Fahrenheit. As you say, it's quite convenient. One of my favorite /. comments is "Fahrenheit is a wonderfully human temperature scale: 0 is too damn cold, 100 is too damn hot." And while individual preference will shift around a bit exactly where the "too damn hot", "too damn cold", and "just right" numbers are, I definitely think that Fahrenheit is a lot closer to covering the "natural" range with convenient numbers than Celsius is.

In terms of everyday life, "0 is too damn cold, 100 is too damn hot" is way way way more useful than "0 is where water freezes and 100 is where water boils". When I'm making pasta, I don't stick a thermometer in and go "oh, it's 90 C now, only 10 more to go"; when I hear the water boiling, it's boiling.

[Kyubey]

Metric obviously. Though I may be biased because I am a physics student and a Canadian.

[enk]

I don't agree with the Fahrenheit too cold/hot idea. Zero degrees Celcius makes more sense: it's the threshold where rain turns to snow (Edit: and water on the roads to ice, more importantly) which is pretty relevant for daily life. The difference between 0 and -1 in Fahrenheit is simply the difference between "too damn cold" and "a tiny bit colder than too damn cold", a distinction which imo does not warrant the most basic threshold in the scale. 20°C is considered standard indoor temperature, at least in semi-cold countries like mine. The gradient from hot to very hot to too hot could just as well go 25°C, 30°C, 35°C, .... as it could go 80°F, 90°F, 100°F, .... The point being that the boundary between sub-100 and super-100 is not a significant threshold in daily life. Whether 100° is used for the boiling point of water or horse body temperature doesn't matter much in everyday life, but I do prefer the former

I'll admit that the inch is a practical unit, at least for hand sized items. For sizes smaller than ~1cm/.4" or bigger than 50cm/20", metric still wins for me.

[phlip]

I'll refer to inches or feet in extremely vague terms, 'cause it's useful to have a short name for those sorts of length... but whenever I need more precision than "a couple of inches" I use centimetres.

As for the claim that imperial units are "more convenient": 1760 yards to the mile. 2 yards to the fathom, 11 fathoms to the Gunter chain, 10 Gunter chains to the furlong, 8 furlongs to the statute mile. Simple. Also simple. And yes, I know not all those are US Customary units, but still, even the ones that are, are pretty weird. People quote the factors of 12 that show up in places (like inches/feet) as being useful because you can easily divide by 2, 3 or 4... and while that can be handy, you can then point to other places where no factors of 3 are available. And you instead have a factor of 11, which is rarely useful (seriously, why 1760?).

It's more a thing with natual languages as a whole... sure, you end up naturally selecting stuff which is useful, and can get a bunch of situational usefulness (nice middle-ground everyday units, like the foot), but end up with a lot of inconsistency, as each standard is naturally evolved completely separately... so you get 12*3*1760 inches to the mile, but 16*100*20 ounces to the ton, and 8*2*2*4 fl.oz to the gallon (where are your useful factors of 3 now?). Metric may miss out on a couple of "useful" named lengths, but at least it has consistency.

Celsius/Fahrenheit is really a separate debate... a lot of the arguments for both sides don't apply to that one. For one, Celsius isn't really a metric unit... "kilocelsius" doesn't really make much sense.

[Fedechiar]

As an Italian, I'd have to say metric; base 10 is really easy once you get the hang of it (you just have to move the decimal point, and it's much easier with square meters and such: as a mile is 5280 feet, calculating the number of square feet in a square mile is much harder than just saying 1 km^2=10^6 m^2) and you don't have to convert when doing physics (except with Celsius/Kelvin, but that's a simple addition).

[AvatarIII]

i say metric, but i'm happy to use imperial in some cases, normally generalizations and speeds/long distances, i might say something is about an inch long, or that i am about 6ft high, and speeds in ingrained because speed and distances on road signs is the last thing in britian that is still officially measured in mph, probably because it would be a bitch to change.

i prefer metric because it is more precise, a gram is much smaller than an ounce, a centimeter smaller than an inch, and then it's easy to break up metric numbers because you just say six point four for example, instead of six inches and four sixteenths for example,

as for the fact that 12s are easy to divide up into smaller numbers, this is true, and i agree, i kind of wish we had 12 fingers and used base 12 maths, but we didn't, but still imperial/standard measuements are not standardized in this, there 14 pounds to a stone for example 16 onces in a pound, then there are 2 types on ton, a short ton (2000lb) and a long ton (160st /2240lb) although the metric inventors just made it worse by calling 1000Kg a tonne at least i know that there's 1000Kg in a tonne without googling it.

[uncivlengr]

As a Canadian engineer that does a lot of woodworking and carpentry, I can generally summarize my opinion on the matter by saying I'd much rather design something in SI units, but build it in imperial units.

I do, however, raise issue with this point:

The second reason, a derivative of the first, is that customary units divide better than metric. Metric is base 10, which is a terrible base. Most customary units use either base 12 (feet, time) or some binary base (volumes, weights). I say this is a derivative of the first because I'm sure these bases were chosen due to their practicability. Division by two, three, four, etc. is more useful than division by five and ten, so units that allow division by three or multiple divisions by two have become dominant.

It might seem like it's more practical, but think about it - how often do you actually have something of exactly 12 units in magnitude to divide into equal portions? A case of wine/beer?

Going back to carpentry, I have a wall that's 11'10", or 3607 mm, neither system lends itself to easily dividing it into three equal portions. Sure, if you intentionally designed the wall to be 12' instead, it'd be easy, but so would intentionally designing it to be 3600 mm.

That said, it's a heck of a lot easier, when you do need to go to the calculator, to find the answer to 3607/3 than (11*12+10)/3.

The only reason I have for preferring to work in imperial units is familiarity - I have a good spatial grasp of inches over millimetres, and when I need to judge the size of something in mm for work, I'll actually think of it in inches, and do the conversion in my head.

[enk]

the metric inventors just made it worse by calling 1000Kg a tonne

Yeah, I think Mg (megagram) would have been cooler. It would also teach people to mind the effin case so they don't confuse it with milligram (mg).

But actually amid all the consistency, metric is still not "perfect": The lower case prefixes are <1 while the upper case ones are >1 except for kilo which was made lower case! (i.e., it's "kg", not "Kg")

When people want something to be 20 MM long, it tell them that molar is used for molar concentration, not length, and also that megamolar is not really a sensible unit for molar concentration anyway.

*shakes fist*

[Thirty-one]

In the process of looking for wedding bands, I came across a site that listed the width of the bands in mmmm. They didn't look tastier than normal gold in the pics, but I suppose that's why they listed the values.

Personally I grew up with metric for everything, and I'm quite happy about that. If I need to measure something that would lend itself to feet, I tend to
see roughly how many half meters would go into it instead. Or I'll break out some measuring tape.
With division I can see the /3 /4 issue, but I also hate conversions in Imperial, and I think that overall that makes calculations easier with metric, for my uses.
I'm also happy with the Celcius scale for temperatures. Everyone where I live know where body temperature lies in it anyway, and having the boiling point and freezing point as 100 and 0 feels pretty natural.

[AvatarIII]

the metric inventors just made it worse by calling 1000Kg a tonne

Yeah, I think Mg (megagram) would have been cooler. It would also teach people to mind the effin case so they don't confuse it with milligram (mg).

But actually amid all the consistency, metric is still not "perfect": The lower case prefixes are <1 while the upper case ones are >1 except for kilo which was made lower case! (i.e., it's "kg", not "Kg")

When people want something to be 20 MM long, it tell them that molar is used for molar concentration, not length, and also that megamolar is not really a sensible unit for molar concentration anyway.

*shakes fist*

it didn't stop them from making dm (1/10 of a meter) and Dm (10 meters) although they are both quite rare measurements to use, and i didn't actually realise kg was supposed to be lowercase,

[Derek]

Obviously customary units are not perfect, but I would rather have an inconsistent but convenient system than a consistent but inconvenient system. The great tragedy of metric is that if they had just used base 12 and not based everything on arbitrary properties of water it could have been a really great system. But instead its doomed to never adequately divide by 3.

Something else occurred to me today. As I mentioned already, the Frency tried to decimalize time at the same time they created the metric system. They also tried to create a metric angle to replaced the traditional 360 degree system*. Obviously neither of these caught on. Why? (Speculation begins here) These two units were already standardized across Europe! The units that metric replaced were the ones that were chaotic. Not because metric was better, but simply because it offered a standard. When faced with an already standardized system, metric failed despite its "advantages" of base 10 conversions.

Ok, that's my metric bashing for today. Thoughts?


*Of course, radians written in terms of Tau are obviously the only real system to use for measuring angles.

[uncivlengr]

Obviously customary units are not perfect, but I would rather have an inconsistent but convenient system than a consistent but inconvenient system. The great tragedy of metric is that if they had just used base 12 and not based everything on arbitrary properties of water it could have been a really great system. But instead its doomed to never adequately divide by 3.

That's an argument over duodecimal vs decimal systems, not imperial vs metric.

10 inches doesn't divide into three any easier than 10 centimeters.

[EvanED]

I don't agree with the Fahrenheit too cold/hot idea. Zero degrees Celcius makes more sense: it's the threshold where rain turns to snow (Edit: and water on the roads to ice, more importantly) which is pretty relevant for daily life.

Point taken; I hadn't really thought about that before.

Though I will say that setting a single temperature for that belies the fact that you can't really do that; there's a lot of other things that will affect whether you'll get icy roads: how sunny it is, what the ground temperature is, etc. So it's really more like "if it's in the lower 30s or below, you should watch out", which is like "if it's below about 2 C, watch out".

I still stick with the "wonderfully human" description.

Obviously customary units are not perfect, but I would rather have an inconsistent but convenient system than a consistent but inconvenient system. The great tragedy of metric is that if they had just used base 12

That gets some of the benefits but loses many others.

...and not based everything on arbitrary properties of water it could have been a really great system. But instead its doomed to never adequately divide by 3.

As uncivlengr pointed out, it's still arguably easier to divide metric units by 3 than arbitrary imperial ones after you start mixing units (because you "can't" give a precise length with the larger ones). See his 11'10"/3 vs 3607/3 example.

(Though I will say: easier than (11*12+10)/3 then converted back to feet/in is 11/3=3 2/3, 2/3*12=8, 8+10/3=3 1/3, so the answer is 3'11 1/3". That really isn't too bad. You could almost do that calculation as you say or write down the answer.)

[uncivlengr]

(Though I will say: easier than (11*12+10)/3 then converted back to feet/in is 11/3=3 2/3, 2/3*12=8, 8+10/3=3 1/3, so the answer is 3'11 1/3". That really isn't too bad. You could almost do that calculation as you say or write down the answer.)

I agree it's not too bad, but twice as bad as one calculation in your head in mm.

Also, you run into a point I didn't bother mentioning, which is that thirds of an inch aren't a standard unit.

Related, I'm working on the erection of a bridge with a span of 100 m, with lateral deck beams spaced at 4166.7 mm, because the designer decided to use 24 spaces along the length. If they had simply added one more beam, everything would have been neatly spaced at 4 m even. Most "everyday" things are easily managable in this way, regardless of units.

[EvanED]

(Though I will say: easier than (11*12+10)/3 then converted back to feet/in is 11/3=3 2/3, 2/3*12=8, 8+10/3=3 1/3, so the answer is 3'11 1/3". That really isn't too bad. You could almost do that calculation as you say or write down the answer.)

I agree it's not too bad, but twice as bad as one calculation in your head in mm.

Like I said, I do buy what I see as your overall argument, which I would characterize as follow: for the sorts of measurements you have to deal with in real life, the "it's easy to divide" argument is vastly overstated as a benefit for the imperial system, to the point where it winds up about the same overall. (I would say some divisions definitely are easier in the imperial system, and others are easier in metric.)

However, I continue to disagree on the details of that example; I would not characterize the imperial calculation as 3 different calculations and the metric one as just 1. I strongly suspect 3607/3 isn't something you'll do as a unit (I certainly wouldn't); in practice, I would count that as two: 36/3=12 in the hundred's place and 07/3=02.3 in the one's place.

Also, you run into a point I didn't bother mentioning, which is that thirds of an inch aren't a standard unit.

Sure, but that's natural: thirds of a mm aren't either. But I think that that's a bit unfair to criticize about.

Thirds of a foot are an even number of standard units, as are any length that's an even multiple of feet. There is no metric unit for which a third of one unit is an even number of another unit.

(Of course it's not like that's typical for the imperial system either. A third of a gallon isn't an even number of any smaller units, nor is a third of a pound, a third of a ton, or scanning through the wikipedia page, a third of anything that isn't based on the foot.)

[Xanthir]

(Of course it's not like that's typical for the imperial system either. A third of a gallon isn't an even number of any smaller units, nor is a third of a pound, a third of a ton, or scanning through the wikipedia page, a third of anything that isn't based on the foot.)

A third of a tablespoon is a teaspoon. (And a tablespoon is half a liquid ounce.)

[AvatarIII]

(Of course it's not like that's typical for the imperial system either. A third of a gallon isn't an even number of any smaller units, nor is a third of a pound, a third of a ton, or scanning through the wikipedia page, a third of anything that isn't based on the foot.)

A third of a tablespoon is a teaspoon. (And a tablespoon is half a liquid ounce.)

a teaspoon in the uk is generally considered to be 5ml, a desert spoon 10ml and a tablespoon 15ml. (according to a converter, it's actually ~6, ~12, and ~18, but when it comes to measuring spoons and recipes and stuff, it normally uses the metric version.

[ShootTheChicken]

Just throwing out an opinion:

I can't speak for any other Canadians, but apart from government documents everyone I know uses imperial for height and weight. I'm 6'1 and 175, I have NO idea where I am on the metric scale.

Everything else I vote metric. Especially distances. Typical highway speed in Canada is 100km/hr, so on long drives you know you have 250 kms to go and you know that'll take you 2.5 hours. SO convenient.

Although I assume everyone is biased to what they grew up with.

EDIT: Oh and baking! Baking is an imperial domain.

[mcvoid]

I go with metric. I don't buy the base-10 argument: it only applies when using MKS or another naming scheme: it's not hard to agree to a power-of-2 naming scheme for special use and it would not only have the same applicability as the power-of-10 prefixes, but also a known, consistent conversion to and from known bases. Also, the way you can combine the elementary units into more spoisticated units, like kilograms, meters, and seconds, into Newtons, is very nice for science. In that way, does not metric have the Unix nature? Besides, I like having a way to measure mass that doesn't depend on the gravity of the place that it is being measured.

[EvanED]

I go with metric. I don't buy the base-10 argument: it only applies when using MKS or another naming scheme: it's not hard to agree to a power-of-2 naming scheme for special use

If you are thinking about kilobytes and such (vs kibi) there's probably a religious war thread on that, so I won't bring up my pretty strong opinions on that here.

Besides, I like having a way to measure mass that doesn't depend on the gravity of the place that it is being measured.

You sort of have one, though it's not commonly used.

And really you don't need it anyway. "Pound" definitely isn't exclusively force -- there's pound-force and pound-mass, and there are situations in which "pound" means both. If you want to specify mass, just say "pound-mass".

[Derek]

If you need to distinguish between force and mass you've left the realm of everyday use and entered the realm of science, for which you should use SI (which is also not quite the same as metric, though it largely has the same advantages and disadvantages). Likewise for combining basic units into more complex units.

Everything else I vote metric. Especially distances. Typical highway speed in Canada is 100km/hr, so on long drives you know you have 250 kms to go and you know that'll take you 2.5 hours. SO convenient.

I can top that with customary. 100 km/h is about 60 mph, which is a typical highway speed in the US. If some place is 140 miles away, you know it will take about 140 minutes, or two hours and twenty minutes. I would say that's even more convenient because you don't have to convert base 10 fractions of an hour into minutes.

Of course, both of these are a just a coincidence of highway speeds, and quickly become useless as soon as you start pushing the speed limit to get there faster

[ShootTheChicken]

You bastard, that is more convenient.

I think I'm going back to my 'whatever you were raised with' answer. You can become fluent in any method if it's how you were raised. I would expect however, that someone alien to both systems would find metric to be easier.

[phlip]

It's all wrong anyway... as a rule of thumb, if the speed limit 100km/h, then you'll probably actually be travelling at an average of 80km/h, give or take a lot, once you take into account slowing down for various reasons... 100km/h is your maximum, not your mean. I find that subtracting around 20km/h from the speed limit gives you a reasonable guess for your actual average speed, and gives you a better estimate of your travel time than just taking the speed limit as-is. Of course, this doesn't apply in places where driving 20km/h over the speed limit at all times is customary... sure, the sign says 100km/hr, but you have to leave a 20% tip on top of that. Or something.

[enk]

Danish highways are 130 km/h in some places and 110 km/h in others. An avg. of 120 km/h gives 200 km <--> 100 minutes which is not too bad either

[cjmcjmcjmcjm]

It will be a cold day in Hades before I order a litre of beer.

N.B. What is it with the small pints here in the US?

[AvatarIII]

It will be a cold day in Hades before I order a litre of beer.

N.B. What is it with the small pints here in the US?

i've had a litre of beer in germany, and considering it's closer to 2 pints than it is to 1, i'm totally happy with it

according to google a US pint is only 473mls instead of 568ml (a UK/imperial Pint), that is weird...

[phlip]

N.B. What is it with the small pints here in the US?

The pint is defined as one eighth of a gallon. Other versions of the gallon were defined for different commodities, and there were equally many versions of the pint.

America adopted the British wine gallon (defined in 1707 as 231 cubic inches exactly (3 in × 7 in × 11 in)) as its basic liquid measure, from which the U.S. wet pint is derived, and the British corn gallon (⅛ of a standard “Winchester” bushel of corn, or 268.8 cubic inches) as its dry measure, from which the US dry pint is derived.

In 1824 the British parliament replaced all its variant gallons with a new imperial gallon based on ten pounds of distilled water at 62 °F (277.42 cubic inches), from which the UK pint is derived.

Seriously, fuck the imperial system.

[ShootTheChicken]

It's all wrong anyway... as a rule of thumb, if the speed limit 100km/h, then you'll probably actually be travelling at an average of 80km/h, give or take a lot, once you take into account slowing down for various reasons... 100km/h is your maximum, not your mean. I find that subtracting around 20km/h from the speed limit gives you a reasonable guess for your actual average speed, and gives you a better estimate of your travel time than just taking the speed limit as-is. Of course, this doesn't apply in places where driving 20km/h over the speed limit at all times is customary... sure, the sign says 100km/hr, but you have to leave a 20% tip on top of that. Or something.

It's Canada, so 100km/h actually means 'YOU'RE ONLY GOING 100 KM/H?' 120 tends to be the minimum that I set my cruise control at , but indeed it depends on traffic and such. The first time I went to the states I was amazed that people followed the speed limits. They're more... suggestions... where I live.

[Derek]

I would say that in the US ten (mph, of course) above the speed limit is pretty typical. Of course, some people (including myself) prefer to follow the speed limit.

[KestrelLowing]

I would say that in the US ten (mph, of course) above the speed limit is pretty typical. Of course, some people (including myself) prefer to follow the speed limit.

It really depends on where you are, and what time of day. I'm from Detroit, and if you're not going 5 mph over, there's a possibility you'll get rear ended. Also, in most places I've been, a cop will not pull you over for going 5 over. Most won't pull you over until you're doing 10mph over, which is the preferred speed for most Detroit drivers.

IME, it's more likely to see people following the speed limit in more rural areas. This of course has exceptions - if it's a highway going through a rural area. People in Wisconsin seem to follow the speed limit (their cops can be absolutely brutal on out of state plates!), people outside of Detroit in Michigan seem to do the standard 5 over with a fair amount going the speed limit, Indiana just sucks (sorry) and Illinois goes as fast as you possibly can with the traffic (or if you're in a clear area, 5-10 over). Obviously, these are just my biased observations.

Oh, and with respect to metric or not, I'd really like it if we could switch over to metric, but my brain has no idea what a kg is, or km. Meters make sense, and cm are wonderful, liters are good (the pop thing worked!) but I can't get the hang of Celsius or grams. Conversions are way easier and I'd prefer that.

One place where I might resist would be baking. I'd have to learn the chocolate chip recipe all over again. I've had it memorized since I was 8, and that would just be annoying. Besides, everything just seems to work with a tsp of baking soda!

[Xanthir]

Besides, everything just seems to work with a tsp of baking soda!

Replace it with 5mL, easy peasy.

[enk]

Besides, everything just seems to work with a tsp of baking soda!

Replace it with 5mL, easy peasy.

I dunno about other metric countries, but in Denmark we still use teaspoons and tablespoons for recipes.

[cjmcjmcjmcjm]

Regarding my previous post, when I was in the UK, it seemed they they used any and all kinds of unit that felt convenient at the time. I noticed that yards and miles were always used for distances (the instantaneous speed limit, IIRC, is in mph), but meters measured height. So, you'd have a car that is 3 yards long and 1.7 meters tall. They also used (kilo)grams for mass and stones for weight. The important thing is that they have a better pint than in the US.

[Deutschbag II]

I was raised in the States, so the Imperial system is second nature for me. But I truly believe that it has only stuck around this long because people don't like change. I occasionally try to make a mental switch to metric, but it's really difficult because I'm surrounded by imperial. I don't think that imperial is common here because it's more convenient or easy. It's just because people are resistant to something different, especially with a recent anti-European sentiment that I've been noticing lately.

But really, think about it. 5,280 feet in a mile doesn't seem to have a basis in any kind of sense. Sixteen ounces to a pound is just silly. And so forth. But since we're raised with them, we're used to them.

[solune]

my 0.20 euros:

On the inch/foot/yard issue, it's true that a base 3 is good for everyday use, but it doesn't cover all the range of every day uses: if you have to talk about 3 eights of an inch or even a 16th of an inch, you've got a problem (why did we switch to base 2 btw ?). There's also a hole between "our garden is 300 feet long" and "walk 1/4 mile to the post office"

This is not really a metric/US issue, but I 'd rather measure fuel consumption in l/100km than mpg : The problem I'm trying to solve is: "I've got 500 km do drive, how much will it cost me ?", not "I've got 10 gal in my tank, how far from my mother-in-law can I get ?"

As for the Celsius, it's not a proper metric unit. If it was, it would start from absolute 0, and it would take one joule to warm 1kg water for 1°, not 1 calorie. But in everyday use, the freezing point is an important reference point.

[EvanED]

This is not really a metric/US issue, but I 'd rather measure fuel consumption in l/100km than mpg : The problem I'm trying to solve is: "I've got 500 km do drive, how much will it cost me ?", not "I've got 10 gal in my tank, how far from my mother-in-law can I get ?"

Now, this is a whole other dispute. And I'll take the other side.

First, why does it matter what it'll cost you to go that 500 km? Are you going to make a decision about whether you're going to take the trip based on how much it costs? Are you going to fill up your tank with only enough gas to get to your destination? About the only realistic scenario I can think of is that you're trying to compare, say, the cost of driving vs the cost of flying. Then it's useful to know how much gas you'll use on a trip.

Second, you can put it as flippantly as you want , but the second one actually is useful: it helps you answer "can I make it to the next milestone on the gas that I have now." If you're driving out west in the desert and coming up on a town, and you see "Littleville: 2mi, Smalltown: 73mi" and you're not convinced there's something between them, it behooves you to figure out if you'll be able to make it to Smalltown on the gas you have or if you should fill up in Littleville. Even in more mundane situations, at least I play that game. Especially along routes I've done a couple times, I tend to have places where I prefer to stop. Maybe you cross a state line and the gas price drops a lot, or something like that. If you can make it across, you'll save a couple bucks.

It's not that the latter calculation is difficult with a volume-per-distance measure, but I do think it's less direct.

[Derek]

As for the Celsius, it's not a proper metric unit. If it was, it would start from absolute 0, and it would take one joule to warm 1kg water for 1°, not 1 calorie. But in everyday use, the freezing point is an important reference point.

That's SI, not Metric

Kelvin is SI, Celsius is metric, at least that's my understanding of it (likewise, joules are SI, calories are metric). At the very least, Celsius is metric in practice, if not theory, since it has (as far as I'm aware) 1:1 correspondence with metric usage. Furthermore, this topic is about everyday use, and no one seriously suggests using Kelvin for everyday temperatures. Therefore celsius is as much a part of this debate as meters.

[Odd_nonposter]

Out in the midwestern states, it would be silly to use metric for distances, since everything is divvied into square mile blocks. A 6x6 square of them made up a township, so many townships made up a county, etc. Not that that matters for our crooked state highways, which were based off of Indian trails (which is why I call them "drunken Indian roads"), creek paths, property lines, topography, and whatever little bumfuck town paid to have a big road run through it.

EDIT: Oh and baking! Baking is an imperial domain.

I'd beg to differ on that. I've used the US system for all of my recipes, and it works well if it's something you've created and made several times yourself. But when you try to make somebody else's recipe, a bunch of things can go haywire. Flour density is a big one. Depending on brand, sifting, and measurement technique, the mass of a cup of flour can vary by 20 grams or so. That can mean the difference between a fluffy cake and a brick. The same problem also exists with brown sugar: is it packed or poured freely? Measuring things by mass is far more consistent, and metric is a lot more precise at the scale of most recipes.

[MHD]

I say we abandon all this anthropocentrism and turn to natural units.

All hail c, G and \hbar!