Why Apple can accurately estimate battery life and others can’t… yet

As the new Macbook Air gets to reviewers and users, it’s become clear that Apple has made a real breakthrough here. No, it’s not the slimness. Nor is it the responsiveness. It’s the battery life. Not that it lasts particularly long, but that their battery life estimates are proving highly accurate.

A month ago, Xavier touched on the topic of the accuracy of PC battery estimates, pointing out that smartphone estimates are close, while PC estimates are invariably off the mark. While we can disagree on the many reasons behind this, no one disagreed with the claim.

So then what to make of the accuracy of the Macbook Air’s battery estimates? Fluke? Apple magic? Neither. Let’s go back to what I said about the battery estimate problem in response to Xavier’s post.

The real problem is everyone highballs their numbers, so anyone who tries to give an honest estimate gets screwed. All else being equal, if consumers are given a choice between an “honest” 5-hour rating and an “up to” 8-hour rating, they’ll go with the 8-hour.

It would seem that Apple is disproving this statement, if it weren’t for one key phrase: “All else being equal…” You can get a Windows PC from Dell, HP, Sony, or any one of a hundred other vendors. They distinguish themselves from each other with their own unique details, but generally “all else being equal” is how they operate. Apple, on the other hand, offers products that are already distinct due to the software and other features. People have other reasons to choose a Macbook Air. Thus, advertising lower yet honest battery ratings don’t hurt them in the market the way it could hurt others.

While I agree with outcries for accurate battery estimates, such as a recent one from Harry McCracken at Technologizer, the conditions that allowed Apple to nail their advertised estimates simply don’t exist for everyone else. For many, their only option for accuracy will be to lowball their estimates, which I believe will hurt them in competition against those who will continue to highball.

So that’s it then? Under current conditions, PC makers can’t be honest without getting screwed, so they won’t be honest? Well, that’s one option. Another option is to change the conditions. Instead of competing on this, PC makers could instead cooperate on a new battery life standard (and by “new” I mean not MobileMark) that tests under real world conditions. Nobody takes the plunge first, but rather everyone goes at once.

Short term, the drop in advertised battery life (from “up to” to “real life”) may hurt them against Apple, but at least they won’t be hurting each other. Long term, if they don’t match Apple on this, they risk letting honest battery life estimates become one more of those other reasons to choose a Mac. Better to take a small hit now than get smacked hard with that later.

7 Comments

  1. RJ

    11/09/2010 at 5:46 pm

    The Ipad is thus far the only apple product that has lived up to its battery claim. (although in the past month that battery seems to last shorter)

    Every single Ipod I’ve ever had (7 thus far), the battery life was at best 50-75% of the claim.

    And my 7-Hour battery 2009 Macbook Pro with Wi-fi on and Brightness at around 80-90% never lasts more then 3-4 hours.

    Reply

  2. GoodThings2Life

    11/09/2010 at 5:59 pm

    I wonder… and perhaps one of the GBMers with a Mac can better answer this, but when you run Windows on your Mac, does your battery life diminish noticeably? I’ve always been perfectly fine with blaming Microsoft and Windows on battery life IF it makes sense and is accurate to do so. But if I can run Windows and Linux and Mac OS and whatever else and still get the same life, that argument becomes invalid.

    I know, for example, that my Dell Latitude E6400 gets about 5 hours of life… it did so on XP when I got it, and it does so now on Windows 7. I’ve never run anything else on it, and I know my usage habits, and it’s consistent. That’s enough for me… I’m rarely away from my desk that long, and I just take my adapter with me if I am. But who knows… if I ran Linux on it, maybe it would be better? Or if I actually broke down and bought a Mac, boot-camped it to run Windows and then maybe I’d get better than 5 hours with my same habits.

    The argument for me is rarely “can I get better?” It’s usually “why can’t I get as good as they claim?” But I’m a technical user, and I think rationally, so I can make the distinction between those questions. Your average consumer thinks, “Ugh, I bought a Dell, and it gets HORRIBLE battery life,” and they don’t add the requisite clause– because it only gets 4-5 hours instead of 6-19 like they promised.

    Reply

    • Anonymous

      11/09/2010 at 6:44 pm

      The W7 vs. OS X battery test has been done by a lot of people with OS X invariably winning out, but it’s always on a Macbook of some sort, which are, of course, optimized for OS X (and vice versa). Not sure what effect that has on the tests.

      To be clear though, this post is about accurate battery life estimates, not longer battery life. Accurate estimates should be possible regardless of OS.

      Reply

    • Xavier Lanier

      11/09/2010 at 6:44 pm

      Yes, running Windows via Bootcamp reduces battery life on MacBooks.

      Reply

  3. pyrotechnomimus

    11/10/2010 at 3:40 am

    As a note, and I can’t find the exact link from Apple, but there are tons of people on forums who mention the same thing… the reason why most macs get lowered battery time in Windows 7 instead of in MacOSX is because of lack of driver support in Boot Camp. There have been plenty of examples, the most notable of which is whenever they update BootCamp and offer better battery times, or beta drivers that support better battery times.

    As a person who has tried a hackintosh, I can also indicate that MacOSX with fully supported drivers, third party (or built into OSX) I get lowered battery life than Windows 7 or Vista. In fact, I get the same amount of Battery out of Linux as I do MacOSX, and part of that reason is how frequently Linux builds (and likely Unix based builds) access the hard drive. In addition there are manufacturer proprietary drivers on a lot of Windows systems (such as Sony or Lenovo) that support turning off or turning down various functions of your notebook while you are not using them. Again, this seems more of a driver support issue than an OS issue, although under most popular versions of Linux currently available I get about 2/3rds battery time compared to Windows Vista/7. XP actually gets me less as well, and a big feature in Vista (continued in 7) were specifically targetting how often your hard drive had to have data pulled from it.

    The two biggest sources of energy drawing from my laptops (over time) are the screen and hard drive, so the limiting of access to the hard drive definitely improves battery life. LED screens help too, but processor management is also seemingly better under windows than most unix derivatives (at least at base). My processor, under windows, can drop to as low as 10-20% of it’s cap, whereas under Linux I could only get it to stage down to about 1/3rd of it’s overall cap. Under MacOSX I could not find the control at all.

    As far as accuracy of battery times. If the Macbook Air actually gets 5 hours of internet surfing, I’d suggest sending it to a Flash heavy website. My friend’s mac displays unnaturally high levels of processing power when trying to do flash, whereas my lenovo, with a weaker processor (as mine is a ULV) rarely used a fourth of it’s processing power for the same websites. Another benefit of Apple’s more restrictive policies.

    I haven’t studied the Air yet, but if they’ve done well, good for them. When I’ve bought Sony they too do a good job on their site. They call it Standard Usage, and it’s listed usually as a range. My TZ, for instance, claimed 5-7 hours of battery time. I usually get 6 1/2 hours wifi surfing. Which is fantastic, it met my expectations. The lowest I ever got it, IIRC, was 4 hours and 40 minutes. More people with ranges or Standard Usage notations would be good. But, then again, Sony’s been doing this for years and yet Apple again gets the note for this on Gottabemobile.com. And I think Sony’s ranged battery times might just be the mix most companies need to go so they don’t get beaten down by the competition while still being able to offer their consumer report esque numbers.

    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *