"Musical Queneau" Test #2 RESULTS 🎨 13 Votes, 3D Graphs & Two Different Rankings

➡️ *Voice-over and subtitles available* 🎨 *Who Won?* Votes, 3D Graphs & Rankings. In this video I present how each of 13 participant distributed 10 points across three different musical styles, how the votes are represented in a 3D space, and how the rankings change depending on the method used to calculate them.
This is not only a question of “who won”.
It is also an exploration of how different ways of looking at the same data can reveal different meanings:13 votes, 3D graphs, two different rankings… and one surprising question:
can the same votes tell different stories?

In Test #2 of the "Musical Queneau" project
https://youtu.be/wANEHF1T9XU
the same original AI Song was presented in three different musical styles:
A = R&B & Soul
B = Guitar Ballad
C = Brass Reggae

*Here you find the voting RESULTS*
*TIMESTAMPS*
0:00 - Intro
0:25 - 13 Votes in 3D Space
1:48 - The x + y + z = 10 Plane
1:58 - All Votes & Final Rankings: Absolute vs Relative
2:17 - Euclidean Distance from the Group Average
2:28 - From the Closest to the Group Average to the Most Individualistic
2:46 - Euclidean Distance Ranking
2:55 - Absolute & Relative Rankings after Two Tests
3:27 - Test #3 Preview
3:52 - 3 New Styles
4:11 - Call to Action

Each participant had 10 points to distribute across the three versions.
This means that every vote can be represented as a point in 3D space:
X axis = Style A
Y axis = Style B
Z axis = Style C
So a vote such as:
A=4, B=2, C=4
becomes a point in space:
X=4, Y=2, Z=4

Since every vote satisfies the equation:
A + B + C = 10
all valid vote points lie on the same plane, a triangular surface inside the 3D space.

🎯 THE AVERAGE POINT (x=3,44 y=3,05 z=3,51)
After plotting all individual votes, I calculate the general average of the votes
This average point represents the “center of gravity” of the whole set of preferences
It shows where the group, as a whole, is positioned in the 3D space

🥇 *_ABSOLUTE_* PREFERENCES RANKING
The Absolute Ranking shows the total strength of each style.
ThIs first ranking is based on the *average* score received by each style
In this “absolute preferences” ranking, the result is very compact:
Style C=3.51 comes first
Style A=3.44 comes second
Style B=3.05 comes third

*Style C* wins the absolute ranking, but the three styles remain very close.

🔁 *_RELATIVE_* PREFERENCES RANKING
The second ranking uses a different criterion
Instead of averaging all points, I assign 1 preference point to the style preferred by each participant.
If a participant gives the highest score to more than one style, the preference point is split equally.
For example:
- if A and C are tied as preferred styles, each receives 0.50
- if A, B and C are all tied, each receives 0.33

Using this method, the 13 preference points are distributed as follows:
- Style A = 6/13 = 46,154%
- Style C = 5/13 = 38,461%
- Style B = 2/13 = 15,385%
Here the ranking changes because *Style A* jumps to the first position.

*This is the most interesting shift* :
Style A does not win in *_absolute_* preferences, but it wins when we count *_relative_* preferences

Same votes: different method, results, and meaning

This is very close to the *spirit* of Raymond Queneau’s *Exercises in Style* : the form of representation changes what we see.

📐 EUCLIDEAN DISTANCE OF EACH VOTE FROM THE AVERAGE
The final analysis measures the Euclidean distance between each participant’s vote and the general average point
This is not a ranking of quality
It only shows how close (or far) each individual vote is from the average voting pattern
Some people vite like the average and others, instead, in a personal (isolated) way
Who voted *closest* to the group average?
Who moved *farthest* away from it?

So we can read the results not only as a ranking of styles, but also as a map of individual positions around the collective center.

🧠 GLOBAL INTERPRETATION
The RESULTS are a small map of different ways of listening.
C wins by Absolute score
A wins by Relative preference
The 3D graph shows that listening choices are not linear: they form a space of possibilities
The melody stays the same
The meaning does not
This video combines AI music, audience voting, music style comparison, 3D data visualization and a creative reinterpretation of Queneau’s “Exercises in Style”.

🎨 TEST #3
Finally, as shown in the video, I compare the results of Test #1 and Test #2 and introduce the next link (Test #3) in the chain:
Test #1 ➡️ Test #2 ➡️ "Test #3"

Musical Queneau continues with *Test #3*
My original AI Song: *The Memories of Youth*
Style A: Jazz Ballad
Style B: Celtic Pop
Style C: Classic Rock
*Online from June 10, 2026*
https://youtu.be/VcUfF-uoTIg

Which style will win next?
Listen ➡️ Compare ➡️ Vote
Musical Queneau continues...

"Musical Queneau"
https://www.youtube.com/playlist?list=PLnD0ArA_BGHv79KTLSxTKwD2_slefiZKT

My Best AI Songs
https://www.youtube.com/playlist?list=PLnD0ArA_BGHuQ14Mp86yX5RCcTp2hUs4H

#MusicalQueneau #AIMusic #SunoAI #MusicExperiment

Видео "Musical Queneau" Test #2 RESULTS 🎨 13 Votes, 3D Graphs & Two Different Rankings канала Mario Nicola Misino