MPG seems like a useless metric

"Jake - Has Bad Luck So You Don't Have To" (murdersofa) 05/27/2014 at 15:38 • Filed to: None

1

27

Fuel consumption is often rated in "Miles per Gallon", with lower scores being worse than higher ones, because you can travel farther on a tank of gas. MPG is given in highway (uninterrupted driving at 60mph or so, pretty much the best you'll ever get) and city (stopping and starting, going around 35mph or so).

Problem is, I really don't like it.

What about idle fuel consumption? What if I was going through a city with no stop lights?

I really think Gallons/Hour @ Speed would be a better metric, with a car rated at

G/H @ 0MPH (idle)

G/H @ 30MPH (city/neighborhood driving)

G/H @ 60MPH (generally where a car gets its best mileage)

Just adding one more set of numbers and changing the measuring stick seems like it would give way more useful information to a consumer than the seemingly-arbitrary (disclaimer: I haven't done any research on how city/highway is determined because I'm lazy) city/highway numbers.

Thoughts? Or am I just an idiot?

yamahog > Jake - Has Bad Luck So You Don't Have To 05/27/2014 at 15:42

7

Personally, I believe in Smiles per Gallon, because looking at the actual gas mileage I'm getting in the XJ will make me cry.

Arch Duke Maxyenko, Shit Talk Extraordinaire > Jake - Has Bad Luck So You Don't Have To 05/27/2014 at 15:43

4

Just adding one more set of numbers and changing the measuring stick would scare and frighten the average consumer. Which is why MPGe was pulled out of the EPA's ass to sell people electric cars.

Jake - Has Bad Luck So You Don't Have To > Arch Duke Maxyenko, Shit Talk Extraordinaire 05/27/2014 at 15:44

1

Ugh, don't even get me started on MPGe.

RamblinRover Luxury-Yacht > Jake - Has Bad Luck So You Don't Have To 05/27/2014 at 15:45

1

Those figures are going to be significantly less helpful than EPA cycle numbers because different vehicles are going to have very different fuel usage accelerating/decelerating, and like it or not, that's a very important thing to know. Under your proposed constant-speed metric, it'll show heavy aerodynamic cars very favorably compared to the real world, and light, non-aerodynamic cars very poorly indeed above 30 compared to what you'll actually get.

Take a diesel Benz and a Fiat 500: the former's numbers by this metric will be an outright lie. Full stop. Only somewhat accurate for a non-congested highway, and accurate-ish standing still, but nobody ought to plan around idle usage anyway. The Fiat will show far better numbers around town or even on the highway, with only hints that will occur from the 0MPH usage rate.

Saracen > Jake - Has Bad Luck So You Don't Have To 05/27/2014 at 15:49

0

You're right, it is arbitrary, and it never accurately reflects real world driving. Pretty much any such metric is going to have its advantages and disadvantages because it is entirely situational.

Using mileage ratings for static speeds as you suggest doesn't really work because it doesn't account for how much fuel the vehicle uses while accelerating.

BiTurbo228 - Dr Frankenstein of Spitfires > Jake - Has Bad Luck So You Don't Have To 05/27/2014 at 15:51

2

The difficulty with having mpg at a constant speed is that it doesn't really reflect what you'd expect to get in the real world. A good deal of fuel is spent in accelerating up to a speed, which is where the bulk of fuel spent in city driving is used.

There's a standard cycle of acceleration, deceleration and cruising that makes up each fuel consumption statistic, which is tailored to represent real-world driving as closely as possible.

How close it does this is open to debate.

The issue I have is that MPG is not a linear scale. 10-15mpg is a bigger gap than 25-30mpg. That's something that could be changed.

Garrett Davis > Jake - Has Bad Luck So You Don't Have To 05/27/2014 at 15:51

0

I don't see what that fixes. You're just changing the metric, not the results. Speed over time isn't really more useful because, like measuring in miles, this doesn't take into account the environment. 60mph on a 3% grade is much different than 60mph on flat ground. Another issue with speed is that you're almost never going to be going one of those prescribed speeds on average, unless you take a freeway trip on cruise control with no traffic. Even then, taking a trip up through the mountains on one trip is going to give you much different results than the last tank where you drove across Kansas.

Also, its a very inconstant metric between vehicles. Aerodynamics play a huge role in mileage at a certain speed, and also something like a hybrid will consume 0 fuel at lower speeds, then more than average at higher speeds, skewing the whole metric. You could have two vehicles with very similar MPG that get wildly different G/H @ X speed.

EPA listed MPG isn't really useful for real world numbers, what it is useful for is to compare it to other vehicles to see which is more efficient.

Most importantly, MPG is very easy to visualize. If my car gets X MPG on average, that means I can go X amount of miles before filling up. As opposed to, "well, if I go 60mph, I can travel for X amount of hours before having to fill up." That only works if you can reliably predict what your average speed for a trip will be, and I challenge anyone to plan their roadtrip that way.

Nibbles > Jake - Has Bad Luck So You Don't Have To 05/27/2014 at 15:51

2

I prefer rods to the hogshead.

RamblinRover Luxury-Yacht > Nibbles 05/27/2014 at 15:55

1

Kilocubits per decadram, bitch.

Shankems > Jake - Has Bad Luck So You Don't Have To 05/27/2014 at 15:55

0

How about an OBDII plugin that records your engine/driving habits and then applies those habits against a list of all other known vehicles. Then you could see whether you're going to net something or if you're a leadfoot and it doesn't matter!

Jake - Has Bad Luck So You Don't Have To > Nibbles 05/27/2014 at 15:56

3

Teaspoons of fuel per Valve video game release cycle

llamaguy > Jake - Has Bad Luck So You Don't Have To 05/27/2014 at 15:56

0

I believe the EPA uses a set driving cycle to better emulate actual driving and not specific fuel consumption which would vary too much based on environmental conditions.

Nibbles > Jake - Has Bad Luck So You Don't Have To 05/27/2014 at 15:59

1

Hectares per cubic liter

Or H/L³

crowmolly > Nibbles 05/27/2014 at 16:02

1

PUT IT IN H!

Arch Duke Maxyenko, Shit Talk Extraordinaire > RamblinRover Luxury-Yacht 05/27/2014 at 16:07

0

Parsecs to Dram

Klaus Schmoll > Jake - Has Bad Luck So You Don't Have To 05/27/2014 at 16:09

1

How about gallons/100 miles or L/100km like the (civilized) world uses it. Sorry, couldn't help it!

See mpgs create a curve that is very steep in the beginning and then gets flatter and flatter. Going from 10 to 20 mpg will save you big bucks, form 20 to 30 it's noticeable but not really worth getting a more expensive car for, 30 to 40 is barely noticeable at the pumps, and 40 to 50 (in the future) won't actually do anything relevant in real life.

With L/100km or G/100miles for get a dead straight line. Much easier to relate to for the sheeple, erm the customer.

miadaman? yes please > Jake - Has Bad Luck So You Don't Have To 05/27/2014 at 16:13

0

Useless stat but it's beneficial to the auto industry that the number can be as ambiguously defined as possible. Anyone serious about tracking their MPG will just do some simple math at the pump.

Most likely the current standard is just a set of baseline driving habits and conditions to run each car through that best simulates "average driving" - which is purely speculatory at best.

deekster_caddy > Jake - Has Bad Luck So You Don't Have To 05/27/2014 at 16:21

0

The EPA has a specific driving pattern that all models are supposed to go through to get their 'average' numbers for both City and Highway. City includes lower speeds and stop and go. Highway includes a variety of terrain and speed. Unfortunately the manufacturers are generally left to do the testing for themselves, so the numbers get fudged a lot.

When Chevy initially ran the Volt through the test they got "230 MPG"!

So the EPA said 'oh, hybrids really screw with this test' so they started looking for a different test parameter for hybrids and EREVs.

(for what it's worth, I'm averaging "103.3 MPG" lifetime in my Volt).

davedave1111 > BiTurbo228 - Dr Frankenstein of Spitfires 05/27/2014 at 17:24

0

"The issue I have is that MPG is not a linear scale. 10-15mpg is a bigger gap than 25-30mpg. "

Er, that is a bigger gap. 50% v 20%.

BiTurbo228 - Dr Frankenstein of Spitfires > davedave1111 05/27/2014 at 17:40

0

Yeah, that's the difficulty. Most people don't see it that way, they just see it as 'x car is 5mpg better than y car', but the money saved decreases the further you go up the scale.

It might be needlessly complicated to change that, or perhaps not really possible, but it's something that's bugged me.

davedave1111 > BiTurbo228 - Dr Frankenstein of Spitfires 05/27/2014 at 17:49

0

I think most people aren't much good at talking about that kind of thing, but do actually understand. With modern cars it doesn't make so much difference anyway, because the denominator is higher.

Mosqvich > Jake - Has Bad Luck So You Don't Have To 05/27/2014 at 17:52

0

I have a 2012 Ford Fiesta. It consistently gets 28.6 MPG all around. The EPA rates it at 29 City/33 Combined/39 Highway. So, clearly I get 28.6 combined. On the highway we get around 34. There are a few factors impacting my fuel economy:

1. We live out west with 75 MPH speed limits. We drive the speed limit, more or less.

2. We live in Colorado at 7,000 feet altitude. This impacts our combined, but we've traveled to Missouri, Michigan, California, and points in between, so it shouldn't have that big of an impact on our highway economy.

Your idea isn't bad (it's good), but I think the highway number should be based on at least 65 MPH or perhaps even 70 MPH to reflect a greater cross-section of the nation.

wiffleballtony > Jake - Has Bad Luck So You Don't Have To 05/28/2014 at 00:34

0

Might as well do a metric of efficiency at 85 mph, no one does 60 in my area.

jdrgoat - Ponticrack? > Jake - Has Bad Luck So You Don't Have To 05/28/2014 at 02:57

0

Those metrics, as others have stated, are rather useless for most circumstances.

Also, as already mentioned, L/100km is a much better metric that we could use as G/100mi or some such variant. I have been keeping up a spreadsheet for the last few months that would be useful to compare all these metrics that we are making up. Well, except for yours where you require a set speed. Because that doesn't really tell anyone anything in the real world. Really.

Also, after seconds worth of my time, I was able to find the EPA traces that they use for determine the mpg figures you see on window stickers. Your descriptions of how they come to the numbers are so far off from reality that I want to bring some light in here. Also, the traces that I found appear to be from before they were revised in 2007(?).

City:

Highway:

(At this point, I apologize if Kinja screwed up some formatting.)

Also, I'm not sure why you think the EPA highway rating represents the best you'll ever achieve. I have bested that rating many times myself, and I'm not really trying. I have never taken the time to learn how to hypermile, nor do I want to be "that guy" on the highway.

The EPA tests use a dyno to load the car, and the operator must maintain the vehicles speed within a certain amount of the point on the trace. It even listed shift points when I drove it in a lab.

There's so much more on this that I wish I could continue with, but I'm getting sleepy.

thebigbossyboss > Klaus Schmoll 05/30/2014 at 08:41

0

I don't understand L per 100 km at all, and I live in a metric country. Someone told me my small car getting 10L to 100 km was terrible mileage...but I dunno seems pretty good to me I guess. How do you convert from one to the other?

Klaus Schmoll > thebigbossyboss 05/30/2014 at 08:47

0

It's not like kph to mph, where you simply take it times 1.62. You have to translate the gallons to liter and so on and so forth, much more math involved.

thebigbossyboss > Klaus Schmoll 05/30/2014 at 08:55

0

Apparently its 28 mpg. What da hell my neighbour talking about 28 mpg is pretty good!