How MPG is mis-leading

[yesthattom]

Watch this before you buy a new car:

It would be amazing if they changed the car regulations to list gallons per 1000 miles, it would really drive the point home that efficiency is what it’s all about.

Comments

[rainbear.livejournal.com]

Yeah, it's one of the nifty things about the Prius (just like the Toyota hybrid they showed). It takes 5 minute averages on a graph, which shows the past 30 minutes of gas usage. The "gauge" to the right shows the current usage, which will range between 5 and 100mpg. The thing that made me think was "mpg is an average. It is not an absolute." Any car going uphill is going to get craptacular mileage. Any car going downhill will get better mileage. In the case of the Prius, it's smart enough to turn the engine off while "gliding" for a while, or while stopped, which improves mileage simply by the fact that the engine isn't running at all.

I think it would be cool if all cars came with that kind of 'instant gas gauge'.

[docstrange.livejournal.com]

Well... Even Cadillacs in the 1990s had immediate consumption gauges. More expensive cars have tended to have such nifty gadgets, and I've seen those gauges in Audis and some other cars in addition to the expensive GMs.

Of course, in the Caddies, the reason for the gauge was that you had to keep an eye on it to make sure the car didn't suddenly consume the last 1/4 tank on one decision to pass that truck while going uphill.

I agree on immediate-consumption gauges being really useful - they show drivers the blow-by-blow impact of their driving habits and decisions.

[catness.livejournal.com]

I guess I'm not sure why this is so revelatory. In what way are people thinking about mileage that's different from what this video explains?

Maybe I'm further afield than most of the world, but I watch mileage all the time, keep track of every tank, and if my averages start to fall off I do maintenance on my car(s) to bring miles per gallon back up to maximum vehicle expectations. Of *course* you get crappier mileage if you're making your engine work harder - that's just common sense. Hills, extreme GVWR, high RPMs, lead feet... all contribute to poor efficiency. How is this news?

I travel a lot for music and racing and it's easy to see even how different drivers spend more and less fuel depending on driving styles. I have also noticed on my motorcycle how fuel efficiency is different for terrain and style: if I ride by myself on my Honda CB-1, I get about 152 miles before I run out of gas (when it's tuned and jetted correctly). If I go on a touring trip fully loaded with luggage, and riding with friends who generally have bikes with double the engine size of mine, I tend to run out of gas at around 126 miles. When I race the CB-1 at my home track (New Hampshire Motor Speedway), I'm fuelless at about 81 miles. The first time I raced at Homestead Race Track in Florida, I had to push my bike back to the pits twice before I realized that 64 miles is my limit there.

I've never actually needed bells & whistles on a graph to figure this stuff out, and I knew it before I did self-service tours or rode/raced motorcycles. If this isn't how mileage is considered by the average person, especially since new cars usually have a different estimated number on them for highway vs. city driving, I honestly don't know how it is.

[bored2sleep.livejournal.com]

What the video points out is that people instinctively compare the numbers linearly, even if their usage is non-linear. In this particular case, MPG would be linear to usage only in some contrived case such as if you were a traveling salesman and your company only reimbursed 2 gallons of gasoline per day. In that case, since your travel is limited by volume of fuel instead of distance, MPG gives you a linear improvement. Most people actually travel based on distance (e.g. if you drive 1000 miles per month to and from work regardless of how much gas that takes) which means that most people would be better served if they were told fuel efficiency in gallons per 1000 miles instead of miles per gallon.

The issue isn't that driving hard burns more gas (though instant fuel efficiency gauges could help people see that as well) but rather that people make poor choices when deciding between cars based on fuel efficiency. They won't replace their SUV with a Prius in any case, but if they could tow their 2000 lb boat with a 13 MPG 4WD Explorer or a 19 MPG Forrester XT, they might not think the difference matters unless you told it as 77 GPK vs 53 GPK. Or tell them that they'd save $1166 per year at 12,000 miles per year. Instead they replace their Honda Civic with a Toyota Prius, and the difference really doesn't matter that much even though the numbers are bigger.

[lonesomepolecat.livejournal.com]

This is why most countries use L/100km and we should too.

We'd gain a whole lot of advantage by changing to this way of expressing efficiency. Apart from other things, we could easily compare what the our EPA gives as efficiency compared to other countries efficiency standards.

Some basics to get you started (using current EPA estimates):
- Prius: 5.1L/100km
- Acura TSX: 9.4L/100km
- Audi TT: 9.8L/100km
- Subaru Legacy, Kia Sportage, BMW Z4: 10.7L/100km
- Chevrolet Colorado: 11.7L/100km
- Toyota Tacoma: 13L/100km
- Shelby Mustang: 13.8L/100km
- F150: 15.7L/100km
- Ford Expedition, Dodge RAM: 16.8L/100km

[agent_dani]

This seems to come down to how people are (un-)able to see the relation of various ways of viewing measurements. Personally, I prefer MPG as I think that way and can easily calculate out what I need, plus I find that much easier for figuring out how far a tank will go (really useful for those of us on machines without fuel gages.)

Overall, I see this as similar to the current disagreement among some meteorologists about how the moisture content air should be expressed - separate camps argue that percent relative humidity and dew point are better and the other a meaningless measurement. In truth, neither measurement is really meaningful on its own.

[http://users.livejournal.com/cgull_/]

I'll vote for L/100km-- all that goodness and the metric system too!

[shandrew.livejournal.com]

I don't think it's misleading; people are just bad at math. Thanks to New Scientist for pointing that out...

At this point, switching to gallons per mile would probably just confuse people ("wait, now smaller numbers are better?"). The EPA already publishes and every car sticker has estimated cost of fuel per year, which is the number that most people care about. A minor advantage is that also incorporates the extra cost of different fuel types. Maybe they just need to make the font for that bigger, like on refrigerator stickers.