Statistic of used keys

Maybe you like to share a statistic of your database? It's pure scientific curiosity, no data abuse!

Can you please publish the percentage distribution of the used key of all uploaded tracks? There should be seen 13 keys! :)
Yes, 13 keys. You know, why.
Please omit all drum tracks.

And the same statistic but for tracks declared as "saxophone" only!

I know this will not be really true because tracks are declared as "saxophone" instead of "keys sounding sax". But nevertheless there could be a significant difference in both distributations.

Happy curiosity and thanks for your kind help!
:)

---

These statistics are already easily accessible to you, just head for the track search engine and check the "Tracks found" display for any criteria you woud like to investigate.
My personal "scientific curiosity" about this dataset is not that big to be honest, but if you feel it reveals answers to relevant questions to you, feel free to enjoy yourself :)

---

I'll try!
I have found 42 tracks with sax ("loose search") in the key of Bb major. The results presented for minor are exactly the same tracks in major but the headline changes to minor.

I have found 16 drums only ("exact search") tracks in the key of C and 5 tracks in the key of D. It's funny. :)

I am having a bit of trouble to understand how to work with "loose" and "exact". But of course sometimes I upload four instruments in a single track and this in first or second or third position in a branch...yeah, it gets to difficult without help. But the answer is: Sax players don't care about "easy" scales...
:)

Ps.
By the way.
The press needs informations to write something. Example: "The keys of music hits". (The key of Eb: Catch the wind)
:) :)

---

Well, I am now a bit confused. How many keys are stored in the database? I have been thinking of 12 plus "missing key".
If we are talking about 24 keys, I'll have to aggregate C-major tracks and A-minor tracks etc. etc.
The best solution would be a matrix with
Instruments in the rows
and
Declared track key in the columns.
These should be the number of "#"s or "b"s or "no" or "key n.a.".
With row-totals and column-totals.
This would give the overall distributions of any instrument. Even of percussion in Eb-minor :)

Because this would be the first statistic of this kind in the world, the press could report about it. Or some students.
:)

---

to perform this in a way you could get a useable result you need solid data well documented ... which is most probably not the case on the 100K tracks online ! i am guessing less than 5% of the published tracks have a Key declared (if not less) ...
i think trying to get stats in these conditions is kinda pointless ...

---

That's right!!
You need a robot to analyse the key and update/overwrite the declarations.

100k of songs will be work for one hour I guess.
:) :) :)

---

a robot ??? who needs a robot when you'll have so much fun analyzing all this material by yourself ????? have fun ! ;)

---

Talking about keys.

A keyword for companies is "Adding values". To make a living for all employees and the capital and the state, too.

You need to ADD values. My fun is no value at all.

Adding values.

I know what I am talking about because I left the IT-business 14 years ago. Why could I leave? "If someone asks you to go a mile with him, go two".
:) :)
Or stay unknown.

---