The Macbook "Brick" • Hacker News Board

Re: The Macbook "Brick"

#41

Post by qqq » Mon, Oct 06, 2008, 2:08 PM UTC

mattmaroon wrote: post #39

unalone wrote: post #38

Earlier quoted context omitted.

An incredibly weighted sample. That's like searching for commentary on any default Linux distro. Negativity rises to the top, because it's more urgent and more entertaining. Again: I'm running my MBP from college, and I'd say that probably half my floor runs Mac. I've been dubbed the techie, unfortunately, and I've been asked no questions regarding Mac bugs yet. The biggest problems people've been having is with usin…

Very true. But, if you assumed Macs failed 10% of the time, and that to be a high rate for notebooks overall, a sample of a ten of which none failed would happen ~35% of the time. A sample of which 3 out of 4 failed would be much more rare. Thus you can glean more information from a sample of 3/4 failures than a sample of 10 with no failures. I don't remember my combinations well enough to figure out the exact number…

You're saying, now, that it is statistically relevant.

Re: The Macbook "Brick"

#42

Post by unalone » Mon, Oct 06, 2008, 4:18 PM UTC

qqq wrote: post #41

mattmaroon wrote: post #39

Earlier quoted context omitted.

Very true. But, if you assumed Macs failed 10% of the time, and that to be a high rate for notebooks overall, a sample of a ten of which none failed would happen ~35% of the time. A sample of which 3 out of 4 failed would be much more rare. Thus you can glean more information from a sample of 3/4 failures than a sample of 10 with no failures. I don't remember my combinations well enough to figure out the exact number…

You're saying, now, that it is statistically relevant.

And it's not. Because from the data I've looked at, 10% is a vastly high number. You can't make numbers up and call them statistically relevant. And you can't disregard the fact that 3 out of 4 could just as easily be the statistical anomaly.

However, the other comment I've made still stands. Can we all drop it? Nobody cares, nobody will prove anything either way. I like the MacBook Pro, love it, think it's the best machine I've ever used. You, Matt, don't. Neither of our opinions matter to the other person, and this entire thread was started pointlessly and seemingly with intent to inflame. Let's drop it and go argue our operating system of choice somewhere that invites actual DISCUSSION.

Re: The Macbook "Brick"

#43

Post by mattmaroon » Mon, Oct 06, 2008, 11:54 PM UTC

qqq wrote: post #41

mattmaroon wrote: post #39

Earlier quoted context omitted.

Very true. But, if you assumed Macs failed 10% of the time, and that to be a high rate for notebooks overall, a sample of a ten of which none failed would happen ~35% of the time. A sample of which 3 out of 4 failed would be much more rare. Thus you can glean more information from a sample of 3/4 failures than a sample of 10 with no failures. I don't remember my combinations well enough to figure out the exact number…

You're saying, now, that it is statistically relevant.

No, I'm not saying that. I'm simply saying a small sample of failures (assuming failures are significantly less likely than non) gives you more information than a large sample of non.

Re: The Macbook "Brick"

#44

Post by qqq » Tue, Oct 07, 2008, 3:36 PM UTC

mattmaroon wrote: post #43

qqq wrote: post #41

Earlier quoted context omitted.

You're saying, now, that it is statistically relevant.

No, I'm not saying that. I'm simply saying a small sample of failures (assuming failures are significantly less likely than non) gives you more information than a large sample of non.

Statistics take into account any valid ways of getting more information. And they also take into account what we know about how to avoid making crap up. You need to either claim to have a statistically relevant amount of information, or to know nothing about overall failure rates of apple laptops.