By on April 9, 2012

Mazda makes fun cars. Too few car buyers care. Mazda has been losing buckets of money. What to do? Mazda is betting that a focus on fuel economy without going hybrid will reverse their fortunes without costing them a fortune. To deliver big mpg gains, and further enhance the driving experience as well, the folks in Hiroshima have creatively re-engineered conventional engines, transmissions, suspensions, and body structures, with an emphasis on light weight and improved efficiency. But talk is cheap. Do Mazda’s “SKYACTIV” innovations actually deliver?

The suspension and body of the 2012 Mazda3, the first car to get SKYACTIV, haven’t changed. The former remains the best aspect of the car while the latter (along with the interior, Mr. Whipple-worthy dash pad notwithstanding) remains the worst. Refrigerator-white paint and undersized wheels don’t help.

The bits that have the largest impact on fuel economy—the engine and transmission—have been replaced. The old 148-horsepower, 135-pound-feet 2.0-liter four-cylinder engine managed EPA ratings of 24 miles-per-gallon city, 33 highway with a four-speed automatic transmission and 25/33 with a five-speed manual. The new 155-horsepower, 148-pound-feet 2.0 achieves a much more competitive 28/39 with a new six-speed automatic and 27/38 with a new six-speed manual. Not quite the magic “40” achieved by others, but in the ballpark.

How did Mazda achieve these impressive gains? For starters, a much taller top gear with either transmission. Perhaps the oldest trick in the book, and one I’ve long wished for in my 2003 Protege5 (which struggles to top 30 mpg on the highway largely because its archaic 2.0 is spinning close to 4k).

Beyond this, the automatic transmission employs an energy-wasting torque converter (a fluid coupling) only for what these do best: smooth takeoffs. Once the car is moving the torque converter locks up and a computer-controlled multi-plate clutch manages shifts. Other automatic transmissions have either a torque converter (usually) or a multi-plate clutch (rare—only the Mercedes-Benz MCT used in some AMG models leaps to mind). Mazda’s innovation: both in the same transmission. The upsides are takeoffs as smooth as those of a conventional automatic and shifts with the direct feel and efficiency of a manual transmission. The downsides, of course, are cost, complexity, and curb weight, as you have two systems instead of one. Mazda has mitigated the weight gain by making the torque converter more compact than it would be if it had to handle all shifting. The quick shifts are an achievement in themselves. Single-clutch automated transmissions have been widely considered an evolutionary dead end, as they’re neither as smooth nor nearly as quick as a dual-clutch automated transmission. But, through extensive tweaking of the transmission’s electronic controls and mechanical bits, Mazda has gotten its box about as close to a dual-clutch system as a single-clutch system is likely to get.

As for the power part of the powertrain, the new direct-injected engine incorporates clever combustion chamber design, sophisticated fuel injectors, and lightweight parts to reduce internal friction and permit an ultra-high 12:1 compression ratio. This compression ratio is actually down from 14:1 in other applications and locations, because a “SKYACTIV” 4-2-1 exhaust header won’t fit in the Mazda3 and Americans prefer regular unleaded gas.

The pudding? Results in previous TTAC tests have been mixed. Brendan achieved 33 mpg. But Derek managed only 25 mpg in town in an automatic and 26 mpg with a healthy helping of highway miles in a manual.

What gives? Brendan drove his car as if it were a rental. Which it essentially was, only rent-free. Derek drove his pair in the Canadian winter with high rolling resistance Blizzaks. Also, his typical trip was only six miles. Engines are terribly inefficient when warming up from very cold temps. With the ambient temperature near freezing, the blue light that indicates a cold engine (in place of the absent temp gauge) stays on for about two miles. The Blizzaks likely knocked off another mpg, judging from Tire Rack tests. Then again, the trip computer likely added at least one mpg, as these devices are wont to do.

For my own tests, I had an automatic car briefly from a dealer and a manual for a week from Mazda. Driving the automatic after the engine had warmed up with a light foot, I observed a trip computer average of 34.6 in suburban driving. Pretty good. But returning on the highway with the engine spinning a leisurely 2,100 rpm at 70 mph (the posted limit), the car managed only 33.5.

I had more of an opportunity to experiment with the manual transmission car, which, like Derek’s, was shod with Blizzaks. Traversing the suburbs in my normal, semi-casual style, I observed about 27.5 on the trip computer—virtually identical to my old-tech Protege5 when handicapped by winter tires. Spot on the EPA number, yet somehow disappointing.

It turns out that the SKYACTIV Mazda3 is especially sensitive to driving style. Driving with a very light foot, shifting short of 2,000 rpm, and paying extreme attention to impending red lights (so as to get off the gas as soon as possible), I managed a high of 41.0 mpg on my standard drive from home to the kids’ school (with a trip computer reset after the engine had warmed up). Now this is more like it!

On the highway, though, the manual transmission car fared no better than the automatic. Speed is a big factor. First, air resistance rises at the cube of velocity. Second, engine speed increases in lock step with vehicle speed, and at a certain rpm efficiency begins to fall off dramatically. Driving at a steady 78 mpg, the trip computer reported 30.6 (again close to my nearly decade-old Protege5). Drop ten mph, and the number ticked up to 31.3. Drop down to 62, and it took a larger jump, to 33.6. Perhaps at the double nickel the promised 37 would materialize—but I just can’t drive 55! The implication is clear – the EPA highway figure is only going to happen at speeds much lower than most of us drive.

The new SKYACTIV powertrain also makes the Mazda3 more fun to drive compared to the former 2.0-liter engine. Though it delivers little in the way of visceral thrills, the SKYACTIV four revs very smoothly, and pulls well at high rpm. For me the automatic succeeds mostly by never calling attention to itself. Shifts are quick and smooth in the rush to get into top gear and when downshifting readily in response to a heavy right foot. For big grins, the manual remains the way to go. Throws have the tight, precise, well-oiled but direct feel of a rifle bolt. I can’t recall a better-feeling linkage in a front-wheel-drive car. Yet all is not perfect: the knob atop the shifter is far too tall, affecting both comfort and the perceived length of the shift throws. Not much of a problem—mods don’t come simpler than replacing a shift knob.

Bottom line: your satisfaction with the Mazda3 SKYACTIV will depend on your driving style and expectations. Expect good looks? Eye of the beholder. Fun to drive? It’ll deliver given sufficiently frequent curves and a willingness to live a grunt-free lifestyle (Texan torque junkies need not apply). Fuel efficiency? Highly dependent on driving style and conditions. Replicate the driving style and conditions of the EPA’s tests, and you’ll meet or exceed the EPA’s numbers. Intensively employ the pedals, drive short distances in cold weather, or exceed 55 on the highway, and those numbers aren’t happening.

Suburban Mazda of Farmington Hills, MI, provided the automatic-transmission car. They can be reached at 877-290-8110.

Mazda provided the manual-transmission car with insurance and a tank of gas.

Michael Karesh operates TrueDelta.com, an online provider of car reliability and real-world fuel economy information.

Get the latest TTAC e-Newsletter!

136 Comments on “Mazda3 SKYACTIV: The Truth Behind The EPA Fuel Economy Numbers...”


  • avatar
    Orangutan

    Did you double check the computer against actual fuel consumption?

    • 0 avatar

      No, I did not. In general I don’t put gas in the cars I test.

      Even if I did, I’d only have the overall average for the tank. With the trip computer I can measure specific trips driven a specific way.

      When I have had both sets of numbers the trip computer is typically high by about one mpg.

      • 0 avatar
        tuffjuff

        I don’t see a problem with this. Five years ago you’d be a fool to trust the onboard computer’s mileage estimation. Now, it’s pretty accurate. my 2012 Focus’ mileage calculation (per tank) has never been more than .3 MPG off from actual MPG acheived with a pen and paper. .3 MPG out of 33-35+ is hardly worth even thinking about, IMO.

      • 0 avatar
        Woochifer

        Based on pump calculations, the trip computer can be off by 2-3 MPG on my Skyactiv Mazda3, if the trip includes a lot of hills (which my daily commute does). With more driving on level terrain, my variance is closer to ~1 MPG.

    • 0 avatar
      DeadWeight

      A critical point in the quest for greater fuel economy has been ignored by most people most of the time: It’s not just what you drive, but HOW you drive, that is a very large determining factor as to how efficiently you’ll drive.

      Yes, development of components such as direct fuel injectors and such have allowed for leaps forward in fuel economy, but it’s only half the story.

      If I drive a new – and I’ll arbitrarily pick examples here – let’s say a 3.6 liter Doge Charger with the conscious goal of preserving fuel, I bet I can do nearly as well on mpg as I would driving one of the newest proclaimed fuel sippers without paying much attention to fuel economy, and the kicker is that I’ll probably make better time.

      Driving the Charger 3.6 in a way that minimizes hard braking at red lights and fast starts, and also keeping the car around 70mph on the highway, is going to get me within 15% (at most) of the driver of a Mazda3 or Chevy Cruze is isn’t much paying attention to driving style.

      And the Charger is going to produce as fast a 0-60 time at 35% throttle engaged as the Mazda3 or Cruze will at 50% open.

  • avatar
    Mark MacInnis

    “Mazda is betting that a focus on fuel economy without going hybrid will reverse their fortunes without costing them a fortune.”

    Mazda SHOULD be betting that getting rid of front fascias that resemble demented children’s cartoon characters will reverse their fortunes without costing them a fortune.

    Those Pokemon front-ends just plain creep me out….

    • 0 avatar
      morbo

      The car just wants to say hello…

      …while it eats your nightmares and rends your soul.

    • 0 avatar
      mtymsi

      +1 Mazda’s current front end styling keeps the 3 off my shopping list. You’d think they would have figured out by now that styling is costing them sales but apparently not.

      That and the fact that the last time I drove one several years ago the road noise was horrendous. Having test driven the 3 several years prior to my last test drive I was quite surprised at the road noise.

  • avatar
    zerofoo

    It’s been said before, but I’ll say it again:

    The “anime” design language of their cars is costing them a fortune.

    Is it really that hard to design cars that don’t look like cartoon characters?

    • 0 avatar
      Russycle

      I don’t get it either. I can’t imagine they’re finding focus groups that are telling them that they love that big dumb grin.

      I do applaud Mazda for realizing that even 4-bangers can use a nice tall cruising gear. Honestly, it’s not that hard to downshift when one encounters a hill on the highway.

      • 0 avatar
        rnc

        It’s not the focus groups its hearing what you want to hear. like when Ford re-did the taurus (ovaloid version). Focus groups kept referring to the car as european looking, ford took that and ran with it (don’t remember who said it, it’s been awhile, but those were the exact words, customers feel it looks european).

        Problem – When it comes to looking european, there is good and bad, good, say BMW or Audi if we are going for cars, bad PSA or renault. Personal example, older italian man who shops at same grocery store, always wears skin tight 501’s, air jordans and a yellow leather jacket, not something you would want to purchase (probably), but very european looking.

        Focus group – he looks european, in a creapy sort of way
        Ford People – he looks european, la la la la la la la la

        But atleast the ovaloid gave us the almost destruction of what was the most profitable automaker in the world at the time and allowed us to witness an almost unthinkable comeback (as documented here)

      • 0 avatar
        TonyJZX

        where I am the Mazda 3 consistently gets the top sales crown every month, month after month… even pre-restyle

        here’s the reason why: clever unique styling, quasi lux-sport, good equipment levels, perceived sporty handling and demeanour… its almost a poor man/clever man’s Golf

        people seem to want a bit of flair in their day and if that’s a hatch… good!

    • 0 avatar
      righteousball

      I’m a big fan of Mazda because when they were non-anime, they did a lot of design work I find subtle and exquisite and way above other Japanese marques at the time. The Xedos9 and the Millenia, the last american 929 (1st gen Sentia), the first MX-5 and the last RX-7 and more than a few JDM cars…

      But based on my experience, most car guys will say the cars I mentioned are not distinctive. Therein lies the rift and the reasoning behind their design language now – it’s at the public’s request.

      I believe Ford also felt the same, i.e. Detroit likes to have their design so strong in a single-minded way it becomes caricaturized, and call it bold design. They did this to Jaguar and Volvo and Saab and the rest of them. Thus Mazda got zoom-zoomified. (and then even worse, Sustainable-zoom-zoomified, but that’s another story)

      Japanese car makers have hammered on for years in Japanese media that cars in Europe need to be instantly recognizable from a rear view mirror. Hence the large badge and big mouth from everyone.

  • avatar
    DC Bruce

    Your results are interesting and perhaps worth comparing to the new 2 liter GDI Ford engine in the Focus (which, at 160 hp, is rated 5 hp more than the Mazda engine).

    In my drive of a rental Focus “SE” sedan riding on whatever the stock tires were from Phoenix to Tucson, the trip computer was reporting better than 37 mpg at 80 mph. That was on essentially level ground with outside temperatures in the high 80s and the a/c running. And, of course, with the DCT transmission.

    It seems to me at least that the Mazda 3 and the Focus SE hatch are natural competitors and provide points of comparison with each other.

    • 0 avatar

      I didn’t record any highway numbers for the Ford. In the suburbs it did about as well as the Mazda, maybe one or two mpg lower. But I tested the two at different times of the year, the Mazda had winter tires, etc.

      As noted in the review, the Mazda does seem to do at least as well as competitors in suburban driving. It’s on the highway that it seems to fall short.

      • 0 avatar
        golden2husky

        Michael, at what speed would you be running when your Protege was at 4000 rpm. That was always a beef I had in my Probe GT, another Mazda design. At 78 MPH, I am spinning 4000 rpm. A poster here (Terry?) once posted about changing the 5th gear for better cruising mileage….

      • 0 avatar

        At 78 mph (my typical speed on the highway) the P5 is spinning about 3,700 rpm.

      • 0 avatar
        redav

        The P5 drops mpg drastically above 65 mph. Mine will get nearly 35 mpg at a steady, sub-60 mph, but barely manages 30 at 70. I have found that tire pressure makes a big difference at those speeds. It’s worth noting that the P5’s engine is much less powerful (only 130 hp peak, and less torque at lower rpms) than these new engines, so they wouldn’t drive anywhere near as well with a taller top gear. It also explains the rapid drop off in mpg on the hwy.

        When I test drove the SkyActiv 3, the trip computer said 36. I drove it on some hwy and around a park with a lot of stops, but mostly casual driving appropriate for the setting. My unscientific observation is that the new engine is far, far better at saving gas when coasting/idling than the P5’s engine, and that’s where the bulk of the improved city mpg comes from.

        Also, I was very impressed with the automatic transmission, and IMO, they deserve a ton of praise for that.

  • avatar
    deliverator

    I don’t think those mileage numbers are unreasonable. I’m rooting for Mazda. I’ve never owned one but if I could help them by buying like 5 Mazdas, I would. It’s so rare to find a cool driver’s car, and rare to find a driving enthusiast too.

    Winter driving kills fuel economy. Right now in my 2008 Civic Si in relatively warm Alberta weather I’m getting an average of 7.9 – 8.1 l/100km, which is about 30 USMPG. This is with somewhat spirited driving and about 70/30 freeway/city driving. I’m enjoying it, a lot.

  • avatar
    Felix Hoenikker

    I,ve never had a car that did not achieve the EPA highway mileage when driving on level terrain at 65 mph. This extends three decades across auto and manual trannys, coupes, sedans and a minivan. So I don’t understand how Michael Karesh managed to get 41 mpg in semi city driving, but lonly 31.3 mpg at 68 mph. Something doesn’t add up.
    As a comparison, my 2004 Saturn Ion 5 speed manual 2.2 Ecotec delivers a very consistent 37 mpg on my daily commute of 65-70 mph driving.

    • 0 avatar

      Yes, this is exactly what I found puzzling. Observed in two different cars, so it doesn’t seem to be a fluke. Note that to get the 41.0 in suburban driving I hit few lights and for most of the distance was able to cruise at 35-50 in top gear. It’s not regularly achievable in the suburbs.

      • 0 avatar

        That seems totally reasonable to me. You’re essentially doing country-sort of driving in the ‘burbs. I have always found that I can get incredible mileage with that sort of driving–both in my old Accord and my new Civic. At 35-50mph I may even do a bit better than 41 in the Civic–probably 42.

        I get 36-plus in the Civic (’08, 5speed) in flat highway driving. I recently got 42, I think (as measured by fillerup) in mostly Vermont windy-twisty-hilly, probably average of a bit under 50, with similar, but flatter driving in Mass, and 35 miles of relatively flat highway.

        I think people have a misapprehension that mileage should be worse in hilly terrain, though. I think it should be better. It’s like hypermiling. When you’re going uphill, you don’t do worse than, I don’t know, maybe 15, or 10 if you’re really having to push hard at low speed to go uphill, yet, you can get 9999 according to teh scan gauge when you’re going downhill.

        I should add that my Civic revs fairly high even in fifth. I think it hits about 3,000rpm around 65, and due to the gear ratios, which are short compared to the Accord, I’m now upshifting typically around 2,500-3,000 rpm.

        I haven’t driven the skyactive 3, althogh I did drive a rental last summer with a slushbox, but I am doubting I’d find it much sportier than my Civic, based on the previous experience, although I’m a bit hesitant to compare a care with a slushbox to one with a shtick. (Jewish joke) I love the Civic.

    • 0 avatar
      Woochifer

      Something definitely doesn’t add up. As I noted in my post below, in the 3+ months and 5,300 miles that I’ve driven my Skyactiv Mazda3, I’ve never had a highway run (of at least 5 miles) show less than 35 MPG on the trip computer; and my daily commute will typically range from 36-40 MPG even with hilly terrain mixed in. My overall MPG of 34.7 is closer to what Brendan recorded in his earlier review.

      • 0 avatar
        ringomon

        @David
        Simple conservation of energy principles make your claim about hillier terrain getting metter MPG false. Any potential energy that you’ve created going up in elevation had to be produced by burning extra fuel. Otherwise where is the energy coming from?

        If you have a hill with an elevation of 2 miles, one up, one down, and you get 15 MPG going up the first mile, and infinite coasting the second mile down, your average for the two miles is still just 30mpg. (You used 1/15th of a gallon to go 2 miles= 1/30th gallon to go 1 miles.) But good luck getting more than half your usual MPG going up a hill that would let you to go the same speed coasting down it. It’s not possible.

      • 0 avatar
        Woochifer

        @ringomon
        With the Mazda3, the trip computer definitely overestimates the mileage on the downhill side. That’s why the variance between the trip computer’s MPG versus my computed pump average MPG is usually 2-3 MPG when I drive on hilly roads, and ~1 MPG when I drive mostly flat roads.

      • 0 avatar
        redav

        ringmon, you are not correct about conservation of energy–

        Theoretically, all the extra energy spent climbing the hill will be returned to you on the way down. Reduced mpg implies there are losses in that process, but it is entirely possible that an engine under high load has a higher fuel conversion efficiency (meaning a higher percentage of the fuel’s chemical energy is converted to work) while climbing, and if that is greater than any losses from going down, it actually could be more efficient on hills than on flat land.

  • avatar
    L'avventura

    Correction: The Skyactiv engines have a 13:1 compression ratio in North America, not 12:1, down from 14:1 elsewhere.

    Source:
    From Mazda’s US website(click on ‘engine”, they use Flash so I can’t directly link):
    http://www.mazdausa.com/MusaWeb/skyactiv.action

  • avatar
    stuntmonkey

    The combination torque converter/multiplate clutch sounds interesting, but I get a sense that future iterations would be simpler, like just having a wet clutch system with lockup. I think I heard it somewhere that by the time we got to 6-7 gears, wetplate clutches would be feasible to replace the traditional torque converter.

  • avatar
    psarhjinian

    Meanwhile, per Consumer Reports’ rankings (and they get theirs by measuring actual fuel flow) the non-GDI, ancient-four-speed-equipped Corolla** about equals this Mazda in their mileage rankings, as well as is most real-world comparisons. As does also-supposedly-outclassed Civic, by the way. The “amazing” Hyundai and Ford are similarly “meh”.

    Now, I like Mazda, but I really think that SKYACTIV is highly overhyped, as are the ratings of the Focus and Elantra. None of these cars are game-changers, not even close.

    I think we can safely say that a) EPA highway is, at best, useful only for the hairshirt version of spec-sheet racing, b) Honda and Toyota don’t get nearly enough credit for making cars that actually get very good mileage without resorting to high-tech trickery, and c) the trip computer is worthless as a metric.

    ** It’s still a miserable car, but you can’t deny the results.

    • 0 avatar
      Richard Chen

      FWIW, there’s about 200lbs between the Corolla and Mazda3 Skyactiv. Less tech, less material = more profit

    • 0 avatar
      gslippy

      Mazda’s use of a slick name for their redesigned drivetrain is a clever marketing trick. What we’re seeing here is that they’ve certainly improved their fuel economy numbers, but just enough to be an also-ran.

    • 0 avatar
      SV

      The Corolla and the Mazda3, at 32mpg combined, are pretty far ahead of the rest of the class. The Civic gets 30, the Elantra 29, the Focus 28, and after that it’s the mediocrity of the Cruze (26), Sentra (27) etc.

      I wonder if Consumer Reports has considered retesting the Corolla. Back in 2003 they got 26mpg from a Malibu V6, which was impossible, so they retested it and got a more reasonable 21. Considering the much lighter tiny-engined Yaris only matches the Corolla in their testing, I wonder if something is afoot.

      The point is, the Mazda3 getting 32mpg doesn’t make it an also-ran, it makes it tied for best-in-class.

      • 0 avatar
        Woochifer

        Considering that Consumer Reports is in a cold weather area, I would be aware of anything that they report during the winter months. One thing about their CR’s methodology though, it’s a lot more reliable and comparable than what how Michael and Derek have reported the Mazda3’s fuel economy, because CR installs a flow meter in the fuel line to record actual gas consumption. The actual mileage will still vary depending on road conditions, weather, driving habits, etc. But, at least the fuel economy represents accurate data, and not subject to trip computer variances and/or short fills recorded at the pump.

    • 0 avatar

      ** busted the laughmeter!!!

    • 0 avatar
      Elorac

      +100, Psar

      Fuelly data also backs up the CR results.

      Interestingly, all trip computer MPGs I’ve experience have been inaccurate, and always on the high side, with the sole exception of my former 2009 G-37, which was generally spot-on. Perhaps when fuel economy isn’t a selling point of the car they don’t bother to make the trip computer excessively “optimistic”…

      • 0 avatar
        burgersandbeer

        You are probably on to something with the G-37. The computer in my 540 is accurate, but considering the car had a gas guzzler tax when new, the truth doesn’t hurt as much.

  • avatar
    rem83

    Unfortunately, relying on an auto-journo road test, with OBC readouts and less than a tank of gas consumed, to predict real world fuel economy is only shade more accurate than relying on your local parts store counter jockey for model reliability information. I’d take a look at the Mazda 3’s page on fuelly, which unfortunately does not really sort drive-trains very well, but many users have identified their cars as Skyactiv with their vehicle name. It looks like there are a pretty significant number of users managing to average 34-37 mpg, which matches the EPA numbers substantially better than many of Hyundai / Kia’s “miracle” cars. http://www.fuelly.com/car/mazda/3/2012

    • 0 avatar

      What we don’t know from fuelly are driving style and driving conditions. As noted in the review, by varying these it’s possible to get anywhere from mid-20s to low-40s. Many people on fuelly are trying to get the best possible numbers, so their driving style will likely be less aggressive than the typical Mazda owner.

      • 0 avatar
        rem83

        Well, as a long time fuelly user, I certainly disagree with your premise that most fuelly users are hypermilers, but – even if that were the case – comparative data is still useful. You could even reasonably assume that most Civic and Corolla drivers would drive even more conservatively than Mazda 3 drivers, yet the ’12 Civic appears to post about 34-37 mpg real world, while the Corolla, with a much smaller sample size, is in the low 30s. The ’12 Accent is also in the low 30s, with the bell curve sharply peaking at 33 mpg and the Elantra at 31 mpg. So the Mazda 3 Skyactiv actually seems to be right on target with, or even somewhat better than, similarly rated cars.

        I’m surprised that someone who’s such a big proponent of TrueDelta would dismiss fuelly so quickly, as well. Couldn’t I argue that your reliability data is skewed as most TrueDelta users will maintain and pay attention to their cars better than the public at large? User reported data collection is what it is – people who don’t care about something aren’t going to report that data. Again, I’d certainly take a large sample size of people who might care a bit more about gas mileage than most vs. one person’s computer reported data.

      • 0 avatar
        Elorac

        Given how heavily the Ford and Hyundai ad campaigns hype MPG, one would expect the Fuelly users of those cars to skew toward hypermiling as well, which makes the actual results all the more surprising.

    • 0 avatar
      Feds

      I came here to discuss the fuelly results, because they are interesting, especially over all 2012 cars, as the distribution is definitely bimodal.

      Per Michael’s comment, Fuelly is a good reference point for that reason: the type of people who use fuelly (myself included) probably drive similarly, meaning vehicle-to-vehicle comparisons are valid.

      Corollas have a pretty even distribution around 30 MPG. Mazda3’s in 2012 have a grouping around 35, and then a flat down to 26-ish. 2011 Mazda3’s are clustered around 26-27.

      Assuming the 35mpg group are skyactive cars, the technology on its own is worth 8-9 mpg, or 30-35%, when compared to the 2011, non-skyactives.

      • 0 avatar
        Woochifer

        From Mazda’s sales numbers earlier this year, about 2/3 of the Mazda3s sold in the U.S. are the Skyactiv models. This means that 1/3 of the 2012 Mazda3 models still use the less efficient MZR drivetrains. 35 MPG sounds about right, since my overall calculated mileage is right at 34.7 MPG.

  • avatar
    chrishs2000

    You should be well aware that there are far more variables to highway fuel economy than “vehicle speed”. You failed to report the ambient temperature, humidity, wind direction, road type (yes, asphalt vs. concrete does make a difference) and condition, distance between the car infront of you and type of vehicle, etc etc etc. If you’re going to make a scientific-sounding post, at least do it correctly.

    As an example: My wife and I drove to Grand Rapids (~150 miles due west from us) and back in the same day. In my 2003 Accord V6 6MT with 205k miles, I managed 35mpg. The car is rated for 19/27 IIRC. Can I make conclusions on the vehicle’s fuel economy with just that information? Of course not – by the way, there was a 15-25mph westerly wind, so the split was roughly 29mpg/40mpg there and back.

    • 0 avatar
      redmondjp

      I have had similar experience with wild variations in fuel economy while travelling over the Cascade mountains – I will get 28mpg in my Civic going east, and then close to 40mpg on the return trip. I finally figured out that the difference is due to the prevailing winds on my weekend of travel.

      • 0 avatar

        I’m not claiming to have performed a scientific test here, but can report that there wasn’t much wind on the day in question. I drove the car to and from Ann Arbor, and observed just over 30 in both directions with a speed in the high 70s. I then checked the impact of vehicle speed heading in a third direction, and again observed a similar result. As noted in the review, what did make a difference (as it always will) was reducing speed.

      • 0 avatar
        chrishs2000

        Michael – Not trying to get on your case because I do appreciate and respect your work, but to report something and then claim that what you were reporting wasn’t intended to be accurate is a cop-out. Your entire point, even the title of the review, was that the car didn’t meet its EPA hwy rating (unless it was driven unrealistically slow).

        I have had this same discussion with several auto-journalists – I think a few Design of Experiment (or similar) classes should be mandatory for any auto-journo that wants to publish “their own” statistics. There is simply no way to accurately measure “real world” fuel economy based on one trip because there are so many variables that will bias one test. I have been watching my ScanGauge II like the anal retentive automotive engineer that I am for the last ~150k miles, and the variance is huge – for example at a constant speed of 70mph, I will get anywhere from 25 to 40 mpg depending on all of the variables that affect fuel economy. YMMV – that’s why your truedelta site is just so interesting!

  • avatar
    NormSV650

    Trip computers are usually optimistic!

    I usually can exceed EPA numbers by 20-40%. Less with V8’s, more with turbo 4-cylinder. Manual transmission and higher tire pressures along with the mentioned above driving style.

    I got 42+ mpg on a 150 miles from south of Columbus to south of of Cleveland in a 2000 Saab 9-5 5-speed. Two stops including 4 miles of one small town 50F averaging 57 mpg. Damn front quarterly head wind @ 12 mph prevented me from mid 40 mpg range. I’ll be testing an 02 9-5 with automatics tranmisssion soon for comparison.

  • avatar
    supersleuth

    If I were looking at this car I’d be a little queasy about how long that complicated automatic transmission would last and how expensive it would be to repair / replace. (Good thing I prefer manuals!)

  • avatar
    mjz

    Just looked at a new 3 today. Mazda is making a HUGE mistake by putting the combined mileage in big bold print on the window sticker. The one I saw said 33MPG in giant type. I thought that was the highway MPG. I had to search for the tiny print that said 28/40. I think they should have emphasized the 40 highway MPG’s instead.

    And the front grille, though better, is still incredibly ugly.

    • 0 avatar
      mikedt

      I thought the format of the mileage section was government mandated?

    • 0 avatar
      mcarr

      This is true. You have to look closely to see that the number is the combined number, not the hwy number. Dumb, dumb, dumb!! I’m rooting for Mazda, but dang, they have to stop shooting themselves in the foot.

    • 0 avatar

      I agree with you on the front grill. An abomination.

    • 0 avatar
      Woochifer

      As mikedt pointed out, I’m pretty sure this is a government mandate. The factory sticker on my Mazda3, which arrived on the dealer lot last year, has 28 and 39 MPG in big bold type, while the 32 combined MPG is less prominent. On all the new cars I’ve seen in recent weeks, the combined MPG is most prominent. Considering how much Mazda is pushing that 40 MPG number in their Skyactiv marketing, I doubt they would change the factory sticker unless it was required.

  • avatar
    Tinn-Can

    So what you are really saying is that I should wait until they roll out a new less goofy looking mazdaspeed3? Can do…

  • avatar
    ringomon

    I own a 2006 Mazda 3 (Hatch- natch) and a 2010 Mazda 3 (not a Hatch- new wife’s choice.)

    Anecdotally…
    I calculated my own fuel usage for a 1 year period (measured gallons purchased vs. odometer mileage) after first buying the ’10.
    My fleet average was a shade under 30 MPG. I had a highway commute to work that was about 30 miles round trip, with just a few miles of surface streets at both ends… but my wife drove mostly surface streets when she did drive.

    I am not exactly a speed-demon- but not exactly a slowpoke either. If the roads open I enjoy spirited driving- although I never do jackrabbit starts or accelerate into red lights or into tailgate position into the car in front of me (like the rest of the word around me seems to enjoy doing so much.)

    This is in Ohio so you have to factor in the Winter dip and Summers hot enough that I used AC.
    I had road trips of highway cruising (average about 5-10 mpg over posted speed limit) where I averaged 35-36 MPG in both cars often travelling through the mountains of KY, WV, and Tennessee.

    All in all I’ve been pretty happy with the real world fuel results of both cars- and would like to get my hands on a sky-active to see what I could get with her.

  • avatar
    Contrarian

    Don’t y’all forget the 10% ethanol penalty. I figure that costs me about 1.5 MPG highway in my “35MPG” car. If you tested it with 10% booze and the EPA tests are done with pure dino, there’s some of the discrepacny we all see in our cars vs EPA numbes.

    • 0 avatar
      supersleuth

      My Fit, on 10% booze, easily beats its EPA highway number both in the suburbs and on the highway. So I doubt that’s the explanation.

    • 0 avatar

      Good point. I don’t know what was in the car, but there’s a very good chance it was 10% ethanol.

      This still wouldn’t explain, though, why the car easily exceeded its “city” number but not its “highway” number.

      • 0 avatar
        supersleuth

        Poor aerodynamics would be the most likely culprit in that scenario. I wonder what the Cd is.

      • 0 avatar
        WheelMcCoy

        @supersleuth

        Did a google search a found the following:

        2010 Mazda3 sedan had, at the time, a class leading cd of .29.
        2012 SkyActiv Mazda3 sedan has a current class leading cd of .27.

        I think the hatchbacks add 0.01 to the cd.

        So I doubt it’s poor aerodynamics.

      • 0 avatar
        supersleuth

        Hmm. Well then, I’m pretty stumped by Michael’s observations, especially since I see further down the thread that CR got really good highway numbers.

    • 0 avatar
      supersleuth

      I looked up this question and the answer turns out to be NEITHER gasoline nor gasohol, but pure indolene, which they say is to eliminate variability in emissions (which, remember, is the primary purpose of the tests) due to fuel quality. What this means for real-world vs. EPA results I don’t know. They correct for the higher energy content of indolene, but I’m not clear on whether they use “pure” gasoline or E10 as the target for this correction.

  • avatar
    Contrarian

    No doubt, but simple chemistry and math tells us that there is a penalty for using 10% ethanol vs pure gas.

    • 0 avatar
      Lumbergh21

      I can vouch for that with real world experience. I take trips from California to Idaho several times a year. Oxygenated gas is not mandated in Idaho and you can still get non-ethanol gas for a $.20 premium per gallon. During one previous trip from Boise to Yellowstone, I used oxygenated gas for the first part of the trip then straight gasoline driving around the park at the first part of the trip back (two full tanks). My mileage was approximately 20% higher using straight gasoline than using gasahol, even though the gasohol was straight freeway driving versus the numerous short hops around the park with some 65 mph highway driving between the park and Idaho Falls, my last stop in both directions to and from Boise. Going back further, I always noticed improved fuel economy while driving in Oregon before they started requiring oxygenated fuels. For me, it’s a case of do I believe what the ethanol defenders tell me or the fuel economy records that I have meticulously maintained on my last three cars over the past 12 years? I guess I’ll believe “my lieing eyes.”

      • 0 avatar
        supersleuth

        How does 10% ethanol reduce mileage by 20%? “Chemistry and math” certainly don’t support that. Ethanol doesn’t have negative 100% BTU content.

      • 0 avatar
        Morea

        It is within the realm of possibility that adding ethanol prevents the engine from extracting as much energy from the non-alcohol components of the gasoline. This would make it **appear** as having negative energy density when what it is really doing is lowering the **effective** energy density of the hydrocarbon fraction.

      • 0 avatar
        Lumbergh21

        +1 Morea.
        I was surprised by that one trip to Yellowstone, but if anything, the road time spent using gasahol was the easier flat protion of the drive and all highway miles with a warm engine (not really an issue given the driving distance and the time of the year, late May). Maybe driving around Yellowstone at 25 to 45 mph is more fuel efficient in a Mazda6 than driving down a flat interstate at 75 to 80 mph. But, there’s no doubt in my mind that based on past experience with straight gasoline versus oxygenated gasoline that gasahol does have a mpg penalty associated withit well out of proportion with the ethanol content and its relative energy density.

  • avatar
    mic

    My ’10 6MT Versa hatch gets 30 mpg consistently in combined driving which is above EPA combined. That’s without hypermiling. But it’s geared so low that going above 70 mph makes the mileage drop fast. BTW, it turns 3100 at 70, I almost drooled when I read the Mazda turns 2100! The one thing no one is saying about the taller gear is engine wear. I’m turning 60k more every hour! This will definitely affect the life of the engine.

  • avatar
    Volt 230

    People who buy into Mazda Zoom Zoom BS, drive harder than those who go for Corolla’s more laid-back, relaxed driving style this could explain the better mileage figures for the Toyota’s better, more consistent numbers.

  • avatar

    On a recent trip from NYC to DE, round trip, I drove 455 miles, out of the 455 miles, 70 were city driving with many short trips on cold engine, HWY speed was 65 to 75 MPH.
    I ended up with 28.5 mpg (REAL), trip computer show 29.1 mpg, that’s about the EPA number HWY only.
    The car, 2011 Mazda 3 hatch, 2.5 liter, 5 speed AT.
    I live in NYC, my trip computer will show 23.5 mpg in normal driving all year unless I have a long trip on the HWY.
    To MHO, I can get to 40 mpg on the new 2012 SKYACTIV with no problem.

  • avatar
    CelticPete

    Problem is you can get like 30mpg in a Dodge Charger on the highway.. There is going to be backlash from the low speed EPA tests of small hatchbacks.

    Hatchbacks have more drag then sedans and the highway miles are tested at low speed.

    • 0 avatar
      mic

      That’s pretty optimistic, Fuelly says 2012 Chargers average 20 mpg.

      • 0 avatar
        rem83

        Yeah, but that’s average mixed mileage over several vehicles, which wouldn’t exactly represent highway mileage. I’m currently seeing 2 RWD V6s, 1 AWD V6 and 1 V8. The AWD V6 has a peak tank of 28.5 mpg while one of the RWD V6s has a peak tank of a hair over 31.3 mpg. I’d say that there certainly aren’t enough data points to get anything conclusive, but if you were going to take a cross country trip in a V6 Charger, you’d probably be seeing very high 20s or low 30s on the open road.

    • 0 avatar
      bunkie

      That doesn’t surprise me. On a recent road trip, I was seeing almost 32mpg average on the trip computer in my CTS 3.0 Wagon. This is a 2-ton car. Of course, around-town driving is closer to 20 than 30, but it proves that even large, heavy cars can get good highway mileage.

      And this, I think points to Mazda’s problem. If you look at a Cruze Eco, (which is heavier than the 3) you will see slightly better EPA numbers and a nicer interior experience by far. The Cruze feels like a big car, the Mazda, despite it’s better driving dynamics gives up too much in perceived experience which is more a factor of comfort and quietness than “sportiness”.

      • 0 avatar
        Woochifer

        But, the Cruze Eco is not the volume seller version, and actually pares about 200 lbs. off compared to other versions (supposedly by using thinner sheet metal, lighter suspension parts, a smaller gas tank, no rear center armrest or headrest, etc.). The Skyactiv version of the Mazda3 makes up 2/3 of that models total sales. In a journalist roundup last year that Mazda hosted for the Mazda3, along with several high mileage variants of competing models, the Mazda3’s recorded fuel economy (avg. 33.7 MPG) came in 2nd to the Civic HF (34.5 MPG). The Cruze Eco was 6th out of 6 (29.8 MPG).

        http://www.vehix.com/blog/reviews/first-drive-review-2012-mazda3-with-skyactiv-delivers-zoom-zoom-and-efficiency

        GM couldn’t have timed the Cruze’s introduction any better, and unlike Ford’s more cautious rollout of the Focus (which uses the same platform as the Mazda3), the Cruze went into full volume production from day one. The Cruze is in line with more traditional American tastes — bigger and softer. Just look at how that formula worked wonders for the VW Jetta’s sales, even as reviewers roundly panned that car. The Focus only recently began outselling the Cruze, but it remains to be seen if that trend holds up.

        Mazda doesn’t compromise the chassis dynamics, and that’s why it remains a niche player in the U.S. market, but also why it has a dedicated following in enthusiast circles. That said, Mazda3 sales so far in 2012 are up over 40% compared to a year ago, and the Skyactiv drivetrain is a big reason, because it directly addresses the Mazda3’s primary liability — fuel economy. Being able to tout 40 MPG puts Mazda into the conversation with buyers looking for a more efficient car.

  • avatar
    vvk

    The biggest change is that we finally are able to get the hatch with the smooth, refined 2.0 liter engine. It used to be that if one wanted the hatch, one had to settle for the rough, loud and incredibly thirsty 2.3/2.5 liter. The excellent 2.0l was exclusive to the impractical sedan.

    • 0 avatar
      ciddyguy

      For the Mazda3, yes, but the Protege5 had the 2.0 and 2.3L motor, depending if it’s the Mazda speed or not.

      Even the older 2.0L had vibrations at idle with the sport stick autobox (as it has a lock up torque converter) as I have that combo and so far, the best I’ve been able to obtain on a FULL tank of gas was 22mpg, combined and much of that being rush hour and/or city driving with some highway driving of at least 60mph to as fast as 70 on average with occasional short bursts to as high as 80 and that was on Saturday when I filled the tank with 12.663Gallons of gas (when the low gas light came on).

      I would not mind it if the car got 28, highway with ease or even 24 city with ease, but even with the current EPA numbers, 22/28 with 24mpg being the combined cycle and I run 10% Ethanol gas here in Washington State it’s likely I’m not getting the best this car can get at the moment. I don’t know if that’s year ’round or if winter use only, only time will tell as we get into the warmer months and I run Falken SS 535 summer tires as that’s what was on the car when I bought it back in January.

      But it seems to be comparable to what the EPA says for similar sized cars like the 1988 Honda Accord with the 2.0L 4 @120hp gets with the 5spd manual as the EPA shows the exact same numbers so on that instance, I think the Mazda’s mileage is OK, not great but OK.

      • 0 avatar
        redav

        Your numbers are problematically low. On predominantly hwy driving, I can average 30+, and it only gets down to 25 even in heavy city driving. The best I’ve ever accomplished on a single tank was 36, but that was 100% hwy (with a tailwind, I assume). Over 10+ yrs, my cumulative total is over 27. I have found that I am almost always between the old EPA city/hwy limits: 25/30.

        Also, I’ve never heard of a P5 Mazdaspeed. I didn’t think they introduced one of those until the Mazda3 was released.

      • 0 avatar
        ciddyguy

        Redav,

        You are right, the Mazdaspeed Protege was only for the sedan, not the hatch, though Mazda did offer the speed in the hatchback 3, beginning in 2006 for the ’07 MY.

        I’m still learning the car’s ways but the EPA does state the 2.0L 4 gets 22/28 under the current numbers so I’m not that far off I don’t think.

        Also, I use oxiginated or rather, ethanol blend of 10%, can’t avoid it I’m afraid, at least during the winter months and if they DO go to full gas for the summer months, they may make the change sometime between now and May when the summer vacation trips generally begin.

      • 0 avatar
        redav

        I remember the MP3 but not a speed version. They sold it in 2002–it had a P5-esque front end but was a sedan. (The sedan Protege that year didn’t get the new styling of the P5.)

        The change in EPA ratings was for driving style, so one would expect some difference, but something still seems amiss. I have the 10% ethanol mix here, too. It’s required by law because of the pollution in Houston. Also, the max speed limit in any of the counties around the city is 65. You should see the same cruising (45-60 mph) mpg as I do, which is definitely over 30. It does drop pretty fast when you get up to 70, though.

      • 0 avatar
        ciddyguy

        Redav,

        Yes, in 2001, Mazda brought out the MP3 variant of the Protege, which was a 1 year only model. Then in 2003, the Mazdaspeed Protege came out and was a 1 year only offering before the speed became a permanent model within the Mazda 3 lineup beginning in 2006.

        The Protege Speed was an update on the MP3 essentially and many of what it had I think became part of the Speed3 later on.

        As to my driving, I live in Seattle, which is hilly, especially Denny Way up into Capitol Hill and a good chunk of Downtown from the water front east to Capitol Hill. I’m also aware that driving habits and the terrain can have an effect and where you live and yes, much of the freeway through Puget Sound is 60-65mph, traffic permitting.

        My commute is roughly 10 miles each way, I live about 5 blocks from the freeway, but in the mornings, I-5 is slow going through Seattle but I only go a short distance before getting off onto the collective distributor that takes me onto I-90 eastbound to Bellevue. That’s roughly 7.4miles of moderately flat terrain, but I DO have to descent a short section of Denny Way over I-5 when getting to the freeway.

        Traffic in the mornings flows along at a decent clip of mostly 60mph, but sometimes I go faster than that if I can. I get off at Richards Road and work is right there, I only have to go through 2 additional lights after the exit’s intersection with Richards Road.

        The commute home is often slow going, at best maybe 30mph many days for most if not almost the entire stretch back into Seattle via I-90 and when a ball game is going on, it’s more stop and go.

        I get off at Rainier Avenue in Seattle but do have to contend with some hills back into my neighborhood through surface streets. Errands on the weekends when I do them are mostly city and some highway, though not often fast up or down I-5 or utilizing Aurora Avenue and occasionally through downtown if coming up from South Seattle home.

        As to my driving style, it’s spirited. I use the manual mode in the Mazda as I’ve driven a stick shift for years, 1st gear often gets shifted at 3Krpm, but 2nd gear doesn’t usually get shifted until 4Krpm for spirited driving as the motor really doesn’t wake up until you hit 2Krpm and above. The autobox shifts way early, keeping the revs to barely 2500rpm and uses top gear too often. Around town, I don’t use 4th gear and most often my rpms are roughly 2500rpm, if not a touch below that but don’t let it drive for long periods above 3K tops if I can help it (usually can though).

        Nothing like a spirited shift into the next gear to make driving fun and thus that’s part of why my mileage is where it is on this car and I’ve not had a chance to do a full tank of strictly highway yet as my tankfuls have been in mixed driving since I bought it.

        So I’m dying to see how it does when I take it on a longer trip, hopefully not before too long.

      • 0 avatar
        burgersandbeer

        The MP3 and the Mazdaspeed Protege both had suspensions tuned by Racing Beat. The Mazdaspeed version added a turbocharger, limited slip differential, wider tires and bigger brakes. Suspension tuning was revised.

        I’m pretty sure later Mazdaspeed3 shares no mechanical parts with its predecessors. Different engine, turbocharger, differential, brakes, etc.

    • 0 avatar
      ciddyguy

      Burgersandbeer,

      The Mazda3 is considered a successor to the Familia/GLC/323/Protege line that stretches back to the mid 60’s in all its forms though is NOT mechanically related to them at all.

      The 3 uses the global C1 platform that underpinned the European Focus and Volvo S40 and came in 2.0L and 2.3L motors the 4spd autobox remained though by 2006, an optional 5spd autobox was available, but only with the 2.3L motor. All cars got a 5spd manual standard.

      The 3 got the MZR variant of the 2.0L motor, the Protege got the FS-DEC variant of the 2.0 (ES and P5 only) for 2001-03 years only. the Protege Speed got a turbo variant of the 2.0L motor, known as the FS-DET instead.

      In 2009, the 2.3L 4 was replaced with the now current 2.5L 4 in the 3. The Speed3 got the 2.3L MZR-DISI motor, fitted with a turbo and received a limited slip diff and ABS/TCS/DSC standard along with a 6spd manual only. The Speed3 got extra structural bracing and received Volvo’s larger discs (12.6″ Front, 11.0″ Rear), the Speed 3 also got 17″ wheels instead of the standard 15″ units on the regular 3. Don’t know what the current Speed3 gets, if it’s still the 2.3 turbo motor.

      So while it still has some of the same types of features to the Protege Speed, they are not interchangeable.

  • avatar
    Woochifer

    FWIW, I just went over 5,300 miles on my Skyactiv Mazda3, and my calculated overall fuel economy (from the pump reading, not the trip computer) is at 34.7 MPG, with ~75% highway driving. In my experience, the trip computer is usually about 2-3 MPG over the pump calculation. My fuel economy per tank has ranged from 30.4 to 39.1 MPG.

    But, I’ve also found that the fuel economy drops quite a bit, and trip computer is far less accurate as well, if the drive includes a lot of hilly terrain. On a 40+ mile stretch with flat terrain, the trip computer showed over 42 MPG, while my daily commute that includes two large hills, usually shows around 36-40 MPG (depending on speed, traffic conditions, and whether I use the cruise control).

    Unless the highway trip was short (under 5 miles), I’ve never had a trip computer reading under 35 MPG on an all-highway run, even during the break-in period when I was purposely doing a lot of accelerating and decelerating. And this also includes runs where my average speed was close to 80 MPH. This is based on 3+ months of daily highway driving.

    Local driving does take a big hit on the fuel economy, even if the city driving accounts for a relatively low percentage of the miles driven.

  • avatar
    darex

    Not impressed. My Veloster hovers at 40mpg on the highway and exceeds 45mpg on the highway routinely. Not once: Every highway trip!

  • avatar
    johnhowington

    in other words, SKYACTIV is really just marketing hype.

  • avatar
    PaulVincent

    Using my 2012 Focus Titanium to take me to various hunting spots (mostly highway mileage) and having the Blizzaks winter tires, I was able to get a maximum of 48 mpg while averaging 50 mph. This was on a 184 mile round trip of which only eight miles were not Interstate. The vehicle had anywhere from 2700 miles to 6200 miles on the odometer as I made this trip many times over the course of eight weeks.

  • avatar

    I am definitely disappointed in the skyactive. I would love to see a pure ICE car this size doing at least as well as the hybrids.

    • 0 avatar
      Lumbergh21

      That is literally impossible if both cars have the same power rating. For starters, the hybrid regenerative braking technology and engine start/stop always assure less wasted energy; therefore, better mileage.

      • 0 avatar
        Woochifer

        It will be interesting to see how the new Skyactiv Mazda6 does when it comes out early next year. It will supposedly incorporate engine start/stop as well as debut their i-ELOOP regenerative capacitor feature (the regenerative braking feeds a set of capacitors rather than batteries). And the rumors on the Skyactiv diesel coming to America continue to circulate, but I’ll believe it when I see it.

    • 0 avatar
      Adamatari

      No current ICE car does as well as the hybrids in the US. Not even on the highway. The Prius is still king. Having driven one on a long highway trip and still managed 50 mpg (in Florida and Georgia, so no real hills), I can say they really did what they set out to do. Maybe if manufacturers started to actually try to improve their aerodynamics they could get there. They’d still need an Antkinson cycle engine, though.

  • avatar
    A09

    My wife has the Mazda3 Grand Touring Sedan, Skyactiv with automatic. She averages 38mpg on the highway. In winter months it averages 36mpg. Best observed average 41mpg. She drives like there is a balloon between her foot and accelerator pedal.

  • avatar
    robc123

    It has nothing to do with their cars- their cars are great (in Canada the price points are very high compared to the USA but the 3 was Canada’s best selling car)

    The problem is how the company is being managed.
    1. they own a baseball team
    2. they own hospitals and dormitories
    3. they own $5 billion worth of real estate
    4. their overhead is too high- as high as honda and suzuki.
    5. They make 80% of their cars in japan and export 70%
    6. kill the SKYACTIV- too complicated, go conventional and make it simple and durable that gets great MPG.

    But still they managed to have a record quarter selling more cars than they ever have abit at a loss.

    They need new executives focused on:

    1. building cars abroad- now that ford doesn’t hold their hand anymore.
    2. selling excess non-car businesses and using cash for starting factories in other countries and R&D.
    3. Focus on China, their future market.
    4. Reduce overhead lower by replacing old guard
    4. redo the Miata , make it cheaper, lighter, faster with better MPG, a turbo and better colors- go manual windows, big deal.
    (that’s just my opinion at the end, but really when a miata in Canada is just 10k way from a boxster with a 6cyl. with better mpg, we have a problem)

  • avatar
    DannyZRC

    reporting fuel economy over the course of a single tank of gas (or less), using only the on-board economy readout, especially without a giant disclaimer to the effect of “these numbers are totally useless and should be ignored” is quite frankly completely sub-par.

    This isn’t the truth about anything.

    • 0 avatar
      Woochifer

      +1
      I will say though that at least Michael is up front about how the fuel economy was recorded, and how he arrived at his readings. But, you’re right in that driving less than a full tank with only the trip computer results, makes for an unreliable fuel economy report at best.

      If I was to go strictly by the trip computer, the results on my Mazda3 would be closer to 37 MPG, rather than my calculated (at the pump) average of 34.7 MPG. In my experience, the Mazda3 trip computer results can vary a lot, depending largely on the terrain (it way overprojects the fuel economy on downhill cruising).

      What’s sad is that this is the second article in a row on this site about the Mazda3’s fuel economy that uses a less-than-reliable methodology. At the very least, there should have been a calculation based on actual pump totals (yes, this is also fraught with issues, but it’s certainly better than relying just on the trip computer and it provides an additional data point). The previous review from Derek was just a hatchet job, with minimal explanation about how his driving conditions and habits (with winter tires, subfreezing weather, and mostly short local trips while admittedly driving aggressively) might have influenced the subpar fuel economy numbers that he recorded.

      • 0 avatar
        Brendan McAleer

        FWIW, I measured my economy figures by hand. I didn’t really delve into the fuel economy figures too much, as I only ran through a little over a half-tank in the week. You’d need an average of three-four tanks methinks, and a more varied range of driving than I did. I simply report that I was fairly impressed with the mileage, considering that I drove it not in a hyper-mile fashion.

      • 0 avatar

        Two mpg does not strike me as a huge amount of variation.

        Also, every report suggests that the trip computer is consistently a little optimistic (vs. unreliably all over the map, as some of these comments suggest). So if the trip computer average (not instantaneous) readings are consistently lower than expected, a different measurement technique isn’t going to yield a more positive result. A slightly more negative one seems much more likely.

    • 0 avatar
      DannyZRC

      Mr Karesh, this site as well as headline for your article read “The Truth”.

      This isn’t. end of line.

  • avatar
    Herm

    Wayne Gerdes tested the Mazda3 Touring with the Skyactiv engine.. unfortunately in blizzard conditions and with the Blizzak tires also, here are the steady state results he obtained at cruise speeds, calibrated:

    http://www.cleanmpg.com/forums/showthread.php?t=42433

    50 mph 50.0 mpgUS
    55 mphS 47.4 mpgUS
    60 mph 44.1 mpgUS
    65 mph 38.4 mpgUS

    This is not hypermiling, just driving at steady speeds on the hwy. I’m sure it will do much better in the summer with LRR tires.

    • 0 avatar

      While these numbers are higher than the trip computers in both cars led me to expect, they are consistent with my observations in one regard: a significant decrease going from 60 to 65 mph. More than most other cars I’ve tested the SKYACTIV engine seems to drop rapidly in efficiency once out of its sweet spot at a low rpm. At 60 mph the engine is spinning about 1,800 rpm. If you want excellent fuel economy out of the SKYACTIV engine, keep it under that number.

  • avatar
    Terry

    Unless you actually do fill/miles divisions, you are pissing in the wind. Relying on a trip computer is anything BUT “The Truth About Cars”
    0 avatar
    gslippy
    April 9th, 2012 at 2:05 pm

    Mazda’s use of a slick name for their redesigned drivetrain is a clever marketing trick. What we’re seeing here is that they’ve certainly improved their fuel economy numbers, but just enough to be an also-ran.
    GSlippy–give it up–you are an admitted airmchair auto reviewer, and a broken record at that. Credibility is not your strong suit.
    Maybe this site should stick to what it does best–fawn over dinosaur Panther cars, check the interiors for hard plastics, and lament how “Honda has lost its way.”

    • 0 avatar

      I’ve responded to this a few times already. In a nutshell: trip computers never seem to read slow, and usually CONSISTENTLY read a little high. If the reported numbers are lower than expected, it’s unlikely that a more precise measurement would help.

      The trip computer is good for testing how much mpg varies by driving conditions and style. You’re not going to burn through a whole tank at 60, then another whole tank at 65, etc. The best method would actually combine both data sources. Alas, the budget doesn’t permit spending $50+ on gas for each review.

  • avatar
    kevlar

    if mazda was serious about taking a different approach in NA they would have announced the mazda 6 2.2L diesel, with manual and auto tranny, at the NY auto show.

  • avatar
    icemilkcoffee

    Didn’t Consumer Report do an instrumented test of all the ’40mpg’ cars? As I recall the Mazda3 did OK on their test, and their freeway mpg matched or beat the EPA number. And CR did it with an actually gauge installed into the fuel line- a far more accurate measurement than either the onboard display or the gas station pump.

    • 0 avatar
      KixStart

      Not all of the 40mpg cars, but quite a few:
      http://news.consumerreports.org/cars/2012/02/reality-check-will-your-car-actually-achieve-the-advertised-40-mpg.html

      • 0 avatar

        I don’t doubt CR’s road test procedures–I think much more highly of these than their reliability survey methods, which are much older and which they put a lot less effort into improving and fine-tuning.

        I am surprised they achieved 43 MPG in a 65 mph highway drive in the Mazda3. This is about 10 mpg higher than the trip computers in the two cars I drove were reporting. Perhaps the Mazda3 is fitted with the most pessimistic trip computer ever?

        Another possibility: cold temps tremendously affect the Mazda. It was about 18 degrees out when I tested the manual transmission car on the highway. Perhaps Mazda needs to work on their cold weather calibration.

      • 0 avatar
        Woochifer

        @Michael Karesh “I am surprised they achieved 43 MPG in a 65 mph highway drive in the Mazda3.”

        I don’t know why this would be surprising. It’s consistent with what other sources have reported, and if you look at the CR table, they got high fuel economy results from most of the other cars that they tested.

        Just from the trip computer, I recorded 42.0 MPG last week on a 40-mile highway trip with the cruise control set to 70 MPH much of the time (mostly level terrain and mild temps).

  • avatar
    robc123

    Who wants to drive like a weeny, at less than 2k RPM’s when accelerating to get just 34 MPG?

    Maybe what we need is a 70 MPG car so that when people drive normal they get 35 MPG.

  • avatar
    Gardiner Westbound

    Been reading automotive reports for a long time. This is one of the most thorough and well thought out items I have seen. Thank you Michael.

    In my view overemphasizing fuel economy is imprudent. Cars of similar size and performance consume roughly comparable quantities of fuel. Gasoline is a small portion of vehicle ownership expense relative to depreciation, interest, taxes and insurance. A couple of miles per gallon is not significant even with gasoline at today’s prices.

  • avatar
    ciddyguy

    As I said above to another poster, I got 22mpg on my last tankful in my 2003 Mazda Protege5 over the weekend with the sport stick autobox, doing mostly slow rush hour and city traffic with some highway. That’s the best I’ve gotten so far in roughly 3 months of ownership. I briefly thought I’d gotten 30mpg, but I think I misread the trip odometer wrong as I’d put some 8. something gallons (a little over half a tank) of gas to top it off and thought I got 250 miles, at that point, it would’ve been 30mpg average. Instead 150 miles for the same amount of gas added yielded 17mpg. So I think I did something a little bit more than 150 miles as my previous mileage has been just over 20mpg average on the past several fill ups.

    This has been due to colder weather, running Falken SS 535 summer tires and me learning the car and its driving dynamics. This last tankful has been with the weather considerably warmer than it’s been as the temps are now in the 50’s for the most part though over the weekend, we hit the low to mid 60’s, 65 on Sunday and nice and sunny too for a lovely Easter for a change.

    I did my first longer highway trip in the car so it’ll be interesting to see what I get when I fill it up again as this tankful will include the highway (I-5) trip to Olympia from Tacoma and the trip from there all the way back to Seattle (an hour and a half roughly trip, one way) dong on average, 65-70 most of the way in heavy-ish traffic as many were driving home from the holidays.

    I can’t wait to take it out on the highway for a serious road trip to see what it gets with strictly to mostly highway speeds.

    So far it seems to be better than my 1992 Ford Ranger truck with the 4.0L V6.

  • avatar
    daveainchina

    This has me asking this question. What are the real world comparisons in mileage of vehicles.

    What I’d really like to see is some sort of standardized paths where all the vehicles get tested. And it would need city, mixed, and highway paths.

    With the same driver for consistencies sake. And I’d love to see which vehicles are doing well in real world type testing instead of epa numbers etc.

    At the moment I don’t know of anyone doing this, they just say ohh we got xxx mpg when we were out testing. But no standard course/path etc. So how can we compare when we don’t know the differences in terrain, elevation traffic etc.

    This would be real information to judge vehicles I think.

    • 0 avatar

      The 2008+ EPA ratings aren’t far off, in my experience–if your driving is similar to the test. Consumer Reports also performs a standard test.

      For my own testing, I measure fuel economy on the same routes with each car. But I have to rely on the trip computers, and cannot control the timing of red lights, so this method is not very precise. I do usually drive the same route multiple times, seeing if the numbers are similar each time.

  • avatar
    shaker

    CR’s numbers were C18/H40/Overall 28 for the 2010 Sedan with the 2.0/5 Auto. For the 2012 – H22/H43/Overall 32 for the SkyActiv 2.0/6Auto.
    The numbers seem pretty consistent with the improvements for the city readings achieved by the sophisticated tranny, and the HWY ratings (the higher compression ratio, and taller top gear).
    YMMV, after that.

  • avatar
    HalfMast

    Jury is still out on SkyActiv…. The Mazda3 is not a “full” SkyActiv car. It’s really more of a SkyActiv mod (or maybe a “preview”) because they didn’t want to do the full car redesign yet. So they dropped the engine and transmission in (to be fair, the biggest parts of the program) and then tried to get weight out where they could. But the CX-5 is the first car to be fully designed under the SkyActiv program. I’d be interested to see TTAC’s review of that one.

    I’m definitely rooting for Mazda on this one… I know the aesthetics aren’t hitting it out of the park, but I still think that they are closer to the pulse of “car people” than any other brand.

  • avatar
    Hildy Johnson

    “Speed is a big factor. First, air resistance rises at the cube of velocity.”

    IIRC the air resistance usually rises with the square of velocity. Furthermore, this applies to the force as a function of velocity; the energy expenditure would only increase linearly with velocity.

    • 0 avatar
      Rada

      Power is a product of force and velocity, so if force is proportional to the sqare of velocity, power will increase as the cube of velocity: P = F * v = k * v^2 * v = k * v^3

      • 0 avatar
        Hildy Johnson

        That is correct – power rises with the cube of velocity. However, energy expenditure per distance only rises with the square (but not linearly as I said above).

  • avatar
    L1011

    Good article about the SkyActiv gasoline version. I look forward to a review of the SkyActiv-D diesel, if it ever materializes here in the US. I’m really, really hoping Mazda brings that vehicle to the US in form of the CX-5 AWD. It’s very likely that will be my next new car if they do.

  • avatar
    a242sj

    The Mazda3 Skyactive is advertised in Canada as consuming 4.9L/100km.
    I’ve driven mine for around 17K km and it has never come even close to 6L/100km. Average is 6.8L/100km, and I drive for 100km/day mostly on highway, 110km/h avg. I am a bit dissapointed not about the car but about the fact that the numbers do no reflect reality, and I think consumers should not be misleaded by unrealistic/lab-obtained numbers. I also reckon that it may not be possible to give a single number for all the different driving conditions, but this fact should be taken into account by manufactures and the agencies.

    • 0 avatar
      ciddyguy

      a242sj,

      While I’ve not driven the Skyactive equipped Mazda so I can’t verify how close to the EPA ratings it actually gets in real life but like all cars, your mileage is largely dependent on how you drive and what kind of driving you do.

      Most people do an average of city and highway, some skewed more towards city, others highway – and even driving in rush hour can have a negative effect if it’s generally pretty bad (like here in Puget Sound). That 4.9/100L mileage looks to be the highway estimate.

      To be honest, I’d say, go by the combined average, not just the highway miles for precisely this reason, you are more likely to get the combined, rather than the highway mileage in real life due. The only time the highway mileage should be heavily considered is when you are on a long road trip that requires the use of the interstate where you can lock the speed in with your cruise control and leave it there, only stopping for fuel periodically.

      A case in point, I drive a 2003 Mazda Protege 5 with the 4spd sport stick automatic (not my choice, but I did buy the car used) and I have so far gotten no more than 25mpg (US) and that even if I mostly did highway driving and that’s doing between 65-70mph, but since a lot of where I drive isn’t all interstate and the speed does vary some due to some of my highway driving are on slower state highways and secondary US routes when not on the interstate, and even there, it’s often in rush hour, that’s the best I can hope for as once I get over 65-68mph, the revs begin to creep up over 3Krpm and this iteration of the Mazdas tended to be very sensitive to your driving habits and the mileage is more greatly affected at higher speeds than in many other cars. The EPA mileage ratings are 22 city, 28Highway with a combined average being 25 or so. Under the old numbers, it was 24-32, combined 26 or so.

      I once had a 1988 Honda Accord LX-I 4 door which had the 2.0L inline 4 (fuel injected) with all of 120hp, but with a 5spd manual and the EPA mileage is 22-28 under the newer numbers, the same as the Mazda and it has the same sized motor (but with dual cams) and at 130HP with the 4spd autobox.

      Also, the terrain of the area you drive your car also can have an effect as a hilly terrain can affect your mileage negatively to some extent as you have to climb and descent those hills. Even where “flat”, it’s not really truly flat, usually, there will be some undulations even there. About the only exceptions will be in the midwest where it IS flat (or Florida for that matter).

      Again, observe your driving habits, the combination of city/highway and whether much of that is during rush hour or not and adjust your expectations accordingly by relying on the combined average – and this goes for just about every kind of car you drive.

    • 0 avatar
      burgersandbeer

      I think all the talk about cars not meeting Government estimates are pointless. There will always be well-intentioned, standardized tests designed to help consumers compare fuel economy. So long as standardized tests are the norm, manufacturers will always try to cheat the system to post better numbers.

      This is not a secret. Anyone with fuel economy as a top priority should also read surveys of real world mileage. To TTAC’s credit, they seem to be one of few sources of reviews that actually try to achieve realistic fuel economy numbers during their test. Read reviews like that as well.

      Many C-segment cars are also available at the rental counter. After narrowing down your selection via dealer test drives, renting your top pick is a good idea if that’s an option. That would give you the best mileage data you could hope for.

      • 0 avatar
        ciddyguy

        Burgersandbeer,

        Good points there and I would agree with what you say here as well.

      • 0 avatar
        gfcpolo

        So first tank in Manual hatchback, 70mph highway with 3 passengers registered 39 mpg. daily commute of 12 miles, combined, but mostly country roads at 45-55, computer says 37-38 mpg, obviously I drive easy but is anyone else getting this mileage?

  • avatar
    herb49

    well,i think you are all wrong.i just purchased a brand new mazda 3 with skyactive this morning,and,even on the test drive,it showed 41.7mpg,and i was going over 60.also,on my way home,cruising about 65mph,the trip went to 47.8 for a long stretch of road.i was very impressed,to say the least.you say what you want about the ‘smiley face’,because i think it adds character to the lines.i also got a great deal,as mine stickered for almost 23,000,and i paid,before down payment,19,200 !!!


Back to TopLeave a Reply

You must be logged in to post a comment.

Subscribe without commenting

Recent Comments

New Car Research

Get a Free Dealer Quote

Staff

  • Contributing Writers

  • Jack Baruth, United States
  • Brendan McAleer, Canada
  • Marcelo De Vasconcellos, Brazil
  • Vojta Dobes, Czech Republic
  • Matthias Gasnier, Australia
  • W. Christian 'Mental' Ward, Abu Dhabi
  • Mark Stevenson, Canada
  • Cameron Aubernon, United States
  • J Emerson, United States