@gav @genmaicha I wasnt defending it, just describing the reality of it. But I'll play devils advocate now :HazeSmug: :


>In statistics, the coefficient of determination, denoted R2 or r2 and pronounced "R squared", is the proportion of the variation in the dependent variable that is predictable from the independent variable(s).
The first sentence is not a definition but something to guide one's (read a mathematician's) intuition

> statistic used in the context of statistical models...

"statistic" is jargon with a very specific meaning and while it sounds funny in this sentence, it is still worth mentioning since it helps with understanding what role in statistics this quantity has, namely, exactly what is spelled out in the next half of this sentence.

> they exclusively define the ourcome and minutia but never define the actual operation

there is literally a "definition" section which spells out how to calculate it for a dataset with the steps for the mean, residual sum of squares, total sum of squares, and finally the formula. Thats a pretty clear algorithm in my books. Arguably what's missing are some examples to further spell this out.

For the variance stuff, the first sentence is pretty much an english language version of the formula definition of the variance.

Theres also some impedance mismatches between what you and the authors want here. The first is related to:

> the primary reason to go to a math wiki is what am i looking at, what is it derived from, what does it prove, how do i reproduce it

Like I mentioned in my previous post this is orthogonal to what the authors care about. Their target audience is not a math noob but a grad student (or higher) and I wouldn't be surprised if the majority of the actual readers of such pages were undergrad/grad students. They dont want to write a textbook but a refresher for someone who is aware of at least some of underlying theory.

The second mismatch is related to terminology and definitions. Most of the simple definitions at the end of your poast are not general enough for the mathematician. Take the variance case:

> a variance is the mean of differences of a set of numbers from their mean
>
> you immediately understand generically what it derives from how to calculate it and what sort of things it could imply

This only holds in the specific case when you are working with discrete probability spaces where the expectation value happens to be the same as taking a mean of some numbers. But the variance is defined for a larger category of things. If there is a continuous random variable being studied then its expectation value (and hence variance) will likely involve integrals (there are definitions of certain averages that use integrals so its not too out of left field). That is, the Wikipedia pages are trying to capture some of this abstraction and generality since the target audience is a mathematician and not a layman. Spelling it out more, to the the mathematician, your last point doesnt match the definition you gave for variance: it is not generic enough and as such does not capture the full scope of implications that would follow from the more general definition