Methinks you don't know what "quality" means.

Your reasoning doesn't hold. The sample size has to be so small as to make the size of random chance distribution bigger than the size of the total possible distribution (i.e. 50% +/-50%, 70% +/- 70%) before you can't draw any conclusion whatsoever. If this weren't true, then even huge samples of millions of data points wouldn't provide enough information to draw any conclusion whatsoever.

There are two general areas in a statistical measurement: the area which shows a likely correlation, and the area which shows a correlation is unlikely. Small samples expand the area of a likely correlation, which is, interestingly, the area that doesn't tell you anything: there may be a difference of 1%, and a sample which provides a margin of error of 1.5% doesn't tell you if these things are the same or different. Large samples do largely suggest that the two things are the same--that the difference is *insignificant*.

In this case, the area that represents "insignificant" is larger than the area that represents "significant". What you learn is that the effect is almost certainly smaller than a certain size, same as when you have incredibly large samples.

Strunk and White: The Elements of Style.

I was carrying the argument ad absurdium, because life isn't so fixed that 3-ways are special. I have, in fact, had 4 girls proposition me all at once; I ran the hell away from that. This experience tells me that such things happen to normal people, at random, not necessarily due to them being super awesome in some way. I've also been told by a number of girls that they've had the same response from a bunch of guys when trying to set up a 3-way: a lot of guys are very much unenthused with the prospect of actually having sex with two girls at once, while it's relatively easy for them to find a girl that wants to go in with them.

The world is a strange place.

In practice, it is absolutely doable even for those of us who barely ever get laid. It is, however, awkward.

People often discount how social women are. You might have trouble getting some girl to know to hook up with you; but that same girl, if she gets it in her head to have a 3-way or something, might bring it up for no apparent reason if she's comfortable with you. Suddenly, you're like, "What, wait, what? 3 of your friends? And you? Just me? Wait, no, go back... I don't think I have the stamina for that... I mean that's a lot of multi-tasking."

It happens. It happens more often when you're unattached (your girlfriend is less likely to be hanging out with you plus a friend and move from cheeky girl-talk to raunchy situations since you're around; two girls you just hang out with sometimes might decide to go for it if it crosses their minds first). It's also an amazingly efficient way to break a man's boundaries and send him skittering to the nearest corner.

It's a potential opportunity risk analysis problem: the panel was added to take advantage of the opportunity of greatly increasing your ROI by having a long-term operable probe.

Applying the small costs additional to add tracking or mobility so the probe can recover from landing somewhere unfavorable (there is an entire dark side of the comet) potentially has huge returns in additional scientific data gathered: even absolutely minor, unique data is data you can't get without funneling billions of dollars and years of space travel into a comet excavation mission, so any additional data gathered is billions of dollars cheaper than it otherwise would be.

If the probe turns out like the mars rovers--able to operate for months or years on solar power, possibly occasionally powering down for a few days or weeks to let the batteries recharge for laborious drilling and analysis--then your small additional investment pays enormous dividends. If the probe is simply able to gather more data than predicted, but the data is good and interesting, your small investment is still paying off. If the probe simply fails, the small additional investment costs you little compared to the vast base cost of the mission itself.

This type of analysis tells you when to capitalize on risk opportunities. If you can turn a 10% likelihood of gathering 1% additional ROI into a 95% likelihood by adding 0.001% to the cost, you do that. You should look at all risks and combine as much mitigation as you can to minimize the discrete cost of each and maximize the total opportunity exploitation and total risk mitigation, and then add contingencies (alternate plans) which take effect at possible cost (maybe you can do X and Y, and it's expensive to be able to do BOTH, but you can implement the capability to do EITHER and sacrifice the other for cheap), but only when the threat or opportunity presents itself in earnest.

Well, apparently they planned to launch with the gravity assists and eddies needed, and to intercept the comet's trajectory. That's macro-scale planning: "I'm going to go grocery shopping at Big K-Mart".

They apparently didn't plan back-ups for parts failure (thruster on top), landing anomalies (bounces around), losing track of the probe, or landing in shadow (couldn't make it mobile enough to move to a lighter place). They also didn't consider the cost in delivering a payload, versus the cost in delivering a slightly larger payload--two probes, which is less than two probes and two launchers and two fuel loads. That's micro-scale planning: "I need to carry a second credit card in my wallet in case I run low on cash or the first has a mag strip failure".

It's certainly easier to plan a mission that can get there and do its work if all goes well compared to if things go wrong. Realizing that something can be done and drawing up the plans to point-and-shoot isn't "well thought-out"; risks and contingencies are what make a plan well thought-out.

No she's cute. She was a coworker, so not attractive to me (I have a psychological block in place; my emotions attach to concepts and ideals, not to individual people, and I dislike the concept of polluting a professional relationship), but she had objectively all the surface qualities in behavior, intelligence, and bodily appearance one would find attractive.

It's simply an interesting culture point.

The OTA updates they released also wiped the OS. They didn't OTA it because it was too big for their network. The prior updates were 30-ish MB, and this was 130-ish. Motorola spent some 4 months telling us they had an update ready, but didn't know how to roll out something larger than the amount of RAM in the phone.

A ring is a thing you can inspect as well: if it's shiny inside and dirty outside, she's been working it off her finger. This makes no sense in America, as people are prone to bone married women.

Ask yourself this question: why do they behave that way? Even a health nut would eat meat if it were the only healthy food available--they would know that the carbohydrate-based gruel and the apples they can get in prison are not an adequate fat and protein source, and will make them ill, and so they will need meat because nobody is going to give them a bowl of nuts and dried soy beans. What, then, would drive a person to such obsession, if not fear of health issues?

Similarly, animal rights and the issue of supporting the meat industry go out the window in situations where your change in consumption effect no change in society. For example, a wedding planner may cater the wedding such that the 40 pounds of meat and 25 pounds of vegetables are vegan-friendly: the vegetables may involve beans, salads, and some better-spiced vegetable dishes, but they'll still be 25 pounds of vegetables, and you'll still see 40 pounds of meat at the reception. Obviously, the prison system isn't going to reduce its ordering of meat to deal with the vegan population. In both situations, what ties you to adhere to a hollow principle? The "animal rights" and "support of animal torture" considerations are founded on sand, and the sand has shifted away, and yet you still attempt to stand firm on air.

In these situations, when a person experiences any distress at the consideration of consuming the proffered meat, that distress stems not from a violation of the world around them, but from a violation of themselves. There is no other explanation: it damages their image of who they are, which weakens their sense of social position. You offer only peacock feathers and dismiss the idea that the bird may have flesh beneath them; I offer that a peacock is a bird of flesh and fat, and the feathers are only the dressing within which it enshrouds itself.

Vegans are, ostensibly, an issue of morality. Marriage of 12-year-olds to 30-something men are also an issue of morality. There are many other social behaviors which can be said the same of, and they have all changed over time in similar ways, and have been studied to show similar drivers.

You're fixated on this "Vegans are a mystery box, and the way their minds work is deeply personal and does not follow any laws of science" ideal. Problem is their minds do follow the laws of science, and can be approached analytically.

That's not how statistics works.

With 10 people, a difference of 50% falls within random chance. A difference outside 50% is significant.

With 100 people, a difference of 10% falls within random chance.

Statistically, if you fall within the boundaries of random chance, you can't show that there is a difference between these groups. You can show that there is not a *significant* difference; the measure of significance depends on your sample size, as above (although you'd actually use correct figures, not 50% and 10% arbitrarily; there's math to compute alpha values, but it involves calculus).

Bennett claims there isn't a significant difference: that his small sample size indicates the probable difference between groups is bounded to a range smaller than posited. The range is large, but it's meaningful.

I know a woman who refuses to ever wear a wedding ring. She doesn't like the European position of women as property, and the band as a shackle.

Obvious is not scientific. It's obvious the sun revolves around the earth.