Not "fair" at all, and not at all appropriate. Still want the average, the mean, the expectation because of the law of large numbers and the best L^2 approximation property of the mean. Besides, often in practice, e.g., for even positive integer n independent, identically distributed samples from a distribution absolutely continuous with respect to Lebesgue measure, the median is not even well defined.
The median is good for politically correct news stories about data from political economics.
No, as is clear on the screen in front of you, you wrote
"I think it's fair to assume"
No it's not "fair to assume" that.
For what was stated and assumed, my example works.
Sure, my example looks contrived and pathological, but so it commonly is with simple examples.
Your claim that height, weight, etc. are normally distributed is also wildly unjustified.
You have swallowed a common bit of old quasi-mystical nonsense from 100 years ago that real data is commonly normally distributed. That claim is nonsense.
Read what else I wrote about Gaussian distributions in this thread.
Grow up. Learn. In practice, Gaussian is rare and otherwise fantasy land.
I gave a simple, mathematical example of how all but one person could be better than the average in response to the claim that this could not happen. You gave a way it could happen from how people might think intuitively, and I gave a mathematical way.
The people who wanted to argue with my math example kept claiming, based on quasi-mystical nonsense, that mean and median have to be the same due to various unstated, unjustified assumptions, that 'most people' mean median when they say average, etc.
So, sorry 'bout that, but here you and I agree: Most people can be better than the average. You gave one justification, and I gave a mathematical one. The people who want to argue with me want to go way out in fantasy land and claim that mean and median have to be equal in practice, which is total BS.
My example is dirt simple. It should be taken as funny. Alas, some people apparently got their Statistics 101 feathers and Gaussian religion in an uproar and wanted to argue.
A claim in this tread is that there is nothing worth learning in college. Hmm .... Looks like this thread is mostly high school dropouts self-taught in programming. Thankfully for US national security, our DoD is far ahead of such nonsense.
You guys are hopeless.
Also, this Web site doesn't work with Firefox 3.0.1.8 on XP 3.0 and, instead, keeps giving "don't match" or some such. The programming here sucks.
So, guys, on this thread, let's agree: The greatest 'business advantage' possible ever is to have some high gloss plastic case on some handheld consumer product from Apple aimed at teenage girls who want to gossip. Ah, of such stuff will be the future of technology! Gossip! That's the key to business! Who would have thought! SUCH progress!
Now, on to you being a jackass.
First, don't run Firefox 3.0.1.8 on XP SP 3. XP is very old and has several issues which make it a poor choice in 2010. This coming from someone who currently works at Microsoft. Oh and in case you didn't realize it, Mozilla ended support for Firefox 3.0.19 in March. Second, since you have found the time to complain about a technical issue on an unsupported browser on a 9 year old operating system, you are clearly a moron. Oh, and to give me tips on how bad at programming we are? That's hilarious. Now, I'm not claiming I'm fantastic, or even one of the top programmers on HN, but I'm certainly more competent than a twit that can't figure out the Firefox update manager.
No. That distributions are commonly Gaussian is some quasi-mystical nonsense left over from 'psychometrics' about 100 years ago.
The most common way to get a Gaussian distribution is from the central limit theorem, and it's much worse than clumsy to argue its assumptions here.
Gaussian needs MUCH more than your "a whole bunch of average drivers, and a few specially good and bad ones".
Your "well-behaved" is asking far too much for practice or reality. Gaussian isn't "well-behaved"; from Melon in 'Back to School', it's "fantasy land". E.g., there's no reason to believe that can have your "controlled skewness".
On your "scale", sure, if know the distribution and if it is absolutely continuous with respect to Lebesgue measure, then can pick a scale that maintains order and yields Gaussian. The educational testing people love to do this. It's quasi-religion. And we collect taxes to pay them for that?
Yes, for stock market data can argue that changes in price each 15 seconds are independent and identically distributed with finite mean and variance so that the central limit theorem applies so that the change in price from noon today to noon next week will be Gaussian. Of course there are traders driving Ferraris laughing at that.
Do yourself a favor: For real data, assume mean and variance exist and are finite, but without a really careful appeal the the central limit theorem dumpster Gaussian. For multivariate data, don't even think about Gaussian.
Uh, some diffusion processes in physics might give you 3D Gaussian quite accurately.
There's no reason to assume that commonly mean means median or that there are enough properties to make them equal.
Besides a claim about "most people" without some solid data is poor support for anything.
No, not "right".
For any finite (> 1) set of people on nearly any measure, it is easy for all but one to be "better than average".
E.g., for any positive integer n > 1, everyone gets a 1 except the one person with a 0 so that the average is (n - 1)/n < 1.
Done.