Google allows you to filter search results by “reading level“, and shows what percent of that query results are “Basic”, “Intermediate” or “Advanced”.
This can be used to find out how frequently advanced articles gets written about a topic. Lets run a sanity test. I axiomatically claim that advanced articles about Knuth gets written more frequently than advanced articles Kardashian. (In percentage terms.). Lets ask Google, and graph that.
(Hover over the chart to see the breakdown.)
Lets run another quick test. I claim that academic websites about programming and mathematics would have more advanced content than general interest websites. I am using the “site:” operator.
This passes too. Wolframalpha > Quora > Pinterest as expected.
Next we want to find which programming language had advanced articles about it. Its easy to intuitively deduce that non mainstream and academic languages would have basic articles written less frequently than Python or Ruby.
Looks like Haskell is the most “Advanced” language. 
.
[Disclaimer after reading some comments]
I am just claiming that Advanced articles about Haskell get written more frequently than Python. Its just an interesting observation, not a rigorous scientific experiment.
For some more fun here is the comparison of web frameworks.
Again, non mainstream frameworks handily are more “Advanced” than Django and Rails.
One more time, lets graph SEO terms.
Common and beginner SEO terms have more basic articles written about them. Conversion Rate Optimization is an “Advanced” technique, who would have thought.
Here is the raw data.
Disclaimer: Its not a rigorous proof of anything. Its just something I found interesting.
{ 6 comments… read them below or add one }
What the fuck is this, are you insane? What have you tried and plotted. Seems like someone who never understood statistics or data mining tried to show off his nerdity. This is seriously pathetic. And at the end of all the hard work, to what conclusion have you reached?
More importantly what have you tried to put over here? This data to anyone who knows even a dime of statistics appears absolute bullshit.
Ask this question to yourself: have the methodology you tried a standard one? If no, do you have any background in any sort of publication work doing which you feel you might have developed skills which lead you to build such models? If yes: you certainly have gathered no skills doing them. If no: you should consider stopping doing these bullshit.
I am tired of seeing all such newsletters and blogs from the self proclaimed geeks, entrepreneurs and nerds. Do mankind a favor: stop blogging! You are doing more harm than good. The cloud is getting polluted.
Ahmed, I am not claiming this as a statistically rigorous experiment. I am claiming this as a joke. There is no need for profanities.
You know your method is deeply flawed, right? Reading levels are really lousy, especially on technical text. What it can show at best is the level of the intended audience of the articles. So I could conclude that articles written about PHP are meant for a wider audience than those written about Haskell. Or I could conclude that people writing about PHP know better how to express their subject in short sentences than people writing about Haskell. Or perhaps PHP appears in texts on topics entirely different from Haskell. And a more technical issue: even the name of the language can cause a shift in readability, since e.g. Lisp is just one syllable, whereas Haskell takes two.
Ok, sorry to kill your joke…
Theo, Definately I am not claiming this as a statistically rigorous experiment. I am claiming this as a joke. I need to add more smilies. I have added a disclaimer.
Keep on blogging man, it looks like developers need to chill a little.
Your analysis must be wrong. Everyone knows that the most advanced programming language was, is, and forever will be Lisp
.