MMA meets 50 Shades of Grey

Quilt.AI Name Generator: MMA Event Names

Quilt.AI
Quilt.AI
4 min readFeb 15, 2019

--

One Championship Events

Entourage fans will remember Ronda Rousey getting out of a car and clobbering Turtle’s car hood or hearing him nearly choke before he blurted out a dinner date invitation while she clobbers him in the ring. We’ve been noticing a surge in women being interested in MMA thanks to her. But the names of the MMA events by themselves felt (to the women we spoke with) to be too masculine. Earlier today, Elon Musk shared and then shut down Open.AI’s fake text generator. Ours isn’t anywhere near that good but here is what we found….

But first…some context on the MMA. MMA was once on the verge of being banned because it was a violent “human cockfighting”. Today it is the world’s fastest growing sport, with UFC worth over $4billion dollars (from $2 million in 2001). What changed?

Widespread spectatorship changed the game. Especially among 18–34 year old fans in the USA,where MMA is more popular than even the NHL or NASCAR. Today, these events are the mainstay of MMA, with sales of UFC 229 being pegged at an average of $1000.

Stepping into the MMA event arena, is spectacle of masculinity. Bright neon spotlights. A caged ring. Men slick with sweat. The presenter’s deep, guttural call to fight. The cues are perfected. What preempts all of this is the nomenclature of the event itself:

Destiny of Champions. Roots of Honor. Warriors of Light.

There is a primordial challenge buried in these names. They conjure up images of power, grandeur, and war. We live in an age where ‘modernity’ is associated with efficient technology, clean design, glass buildings — essentially, high control and predictability.

MMA events speak to a deep need for chaos, with the display of physical strength and courage as both a means to provoke but manage it.

To see if our neural nets are able to learn and replicate the meanings embedded in MMA event names, we ran an experiment to train the machine on MMA event names. A database of event names was extracted from the One Championship, and broken into a logic to teach the machine.

The results?

Quilt.AI Machine Generated Results (Column F)

Pretty on point.

What if we trained the machine with words from 50 Shades of Grey?

MMA received another huge boost in popularity in 2013, when women were added to the UFC roster. Ronda Rousey’s successful defeat of Liz Carmouche in the first ever Women’s Fight in the UFC marked this change, leading to the declaration of the “golden age” for women’s MMA. MMA has been attributed to an equalization of female bodies, and seen a rise in females among its fanbase.

However, the names and meaning behind each event have not changed. We wondered — what if words and values from these other sites of interest could be injected into the MMA naming conventions? Might it be a step towards appealing more to women, and capturing a larger demographic that’s currently cold to the calls of MMA?

As part of a further experiment, interests of males and females who liked MMA are compared. There is a correlation between high masculinity textual cues among men as those interested in MMA also interested in media sites such as “Muscle Madness” and “Speed Society”. In contrast, female fans of MMA are more interested in engagement rings, Cosmopolitan magazine, and Fifty Shades of Grey.

Text from Fifty Shades of Grey was extracted and vectorized into Quilt.AI AI naming engine we had trained only on MMA names.

Previous event names created by humans
Quilt.AI Machine Generated Names

AI-generated MMA event names have more consistent results than human-generated names

We designed an A/B tests for two groups of communication — human-created names by One Championship, and AI-generated MMA event names that had taken into account words that appeal more to females interested in MMA

The A/B test was run over two days in Kansas, Missouri, to 50,000 women aged 18–35 years old, and who are interested in MMA.

AI-generated communications resulted in more consistent results, with a delta difference of 0.32 across campaigns, whereas human-created communications had a delta difference of 0.66. Human-created communications drew both the highest as well as lowest scores in terms of appeal, while the AI-generated communications remains consistently appealing.

This highlights the potential for AI-generated communications to be more predictable, and for its efficacy in generating new content tailored to specific fan demographics.

Implications

  • The ability to draw inspiration and interest areas from digital sources should be considered more seriously when brands move to capture new audiences or want to expand their market share.
  • Exploratory research that requires foray into new cultures, demographics, or spaces, would be well supported through AI collaboration.
  • Using AI in designing communications can make the process faster and deliver more consistent results.

Read more stories on Quilt.AI.

--

--

Quilt.AI
Quilt.AI

We are a culturally rooted, AI powered insights firm that converts millions of data signals into human understanding. Visit us: https://quilt.ai