Uncategorized

Mathematicians Discover Secret Formula For Writing A Hit Song

Scoreahit-logoThough math is usually used to tell you if a song was a hit based on sales, scientists from the University of Bristol believe they have "The Hit Equation." A group focused on artificial intelligence developed machine learning algorithms that mined the characteristics of top hits to develop a formula for key musical features that determine whether a song will be a hit.

The lads at the Intelligent Systems Lab, armed with the power of math, identified 23 features that were considered important from tempo and loudness to harmonic simplicity vs. non-harmonicity.

Hit-words

Features Included in the Hit Equation

The results are 60% accurate in predicting a top five hit or below 30 non-hit. They noted that their Hit Equation must be adjusted somewhat as tastes shift but it's apparently a fairly gradual process. This particular group of scientists aren't discussing commercial applications however their public site, Score a Hit!, suggests entertainment may be a key outcome.

Mark Brown discusses Score a Hit! and related services and points to Hit Song Science and their site uPlaya. They offer not only hit analysis but related marketing services designed to leverage that analysis.

I don't know how Hit Song Science does its magic, but they claim to analyze "fundamental characteristics of all music" including brightness, tempo and changes over time, to identify hit potential in relationship to changing trends. They also identify markets where each song would be most likely to succeed.

Though I can't recommend the service, the idea of analyzing songs to identify markets and then providing marketing services is an interesting application.

MORE: Why Justin Bieber Is Such A Big Hit, a guest post by FutureHitDNA author Jay Frank.

Hypebot contributor Clyde Smith maintains his freelance writing hub at Flux Research and blogs at All World Dance and This Business of Blogging. To suggest topics for Hypebot, contact: clyde(at)fluxresearch(dot)com.

Share on:

10 Comments

  1. Business is an iffy topic for many genuine artists, as most music/art minded folks only dive into the business part as a means to continue their artistic work. Though I am very open minded when it comes to contrast in art, “Score a Hit” topples the walls of pretense that pop is still music/art.

  2. 60%?
    Wonder how this stacks up to Jay Frank’s system(s)
    He’s basing a new business model around the concept…
    In the end, it’s pretty apparent that if one wants to play it safe, and try to score big there’s an equation you can follow, the KLF wrote about it 20 years ago in their manual.
    That hasn’t changed in decades, what has changed is the entire way humans interact with the world around them…let alone the music they listen to. If there’s a way to identify and quantify that variable of the equation, and how it effects things, well heck.
    That’s something to talk about.

  3. This service should be merged with Apple’s Garage band for iOS with Smart Instruments. This way you can immediately know whether you should continue writing the next bar.

  4. This software does not account for the fact that “hits” are driven a LOT by promotion. Almost 80% by promotion. A lot of trash (songs with those “rented rappers”) has scored well on the charts thanks to incessant promotion at the cost of all other music. Hence, the idea that these guys can predict hits is very questionable. They might be able to assess quality based on harmonic structures etc. but predicting hits is a “pipe dream”.

  5. Ever heard the term, “turntable hit”? Talking motion pictures were once a pipe dream, so I think the statistical technique(s) may eventually develop merit, but you’d have to control for a helluva lot of record label, independent radio promo and retail skullduggery. All of this whiz-bang tech is based on faulty human charts, despite the advent of SoundScan and any number of fast-fourier analyses involved.
    Musical tastes are also known to migrate over time. Ah, but what wonders of audio shall our offspring behold…er, behear, should we ever devise an accurate chart?

Comments are closed.