jasmine mithani


Words That Matter

June 2015 - May 2016

problem, methods, solution, and results

DoSomething.org had launched a campaign in April 2015 called "1 in 3 of Us." The campaign focused on educating high school and college students about the signs of unhealthy relationships. The campaign received 30,000 less signups than anticipated. In all aspects, the campaign was considered a failure.

"1 in 3" struck me as incredibly important, and the space I wanted to explore while interning at DoSomething were the sex, relationships, and violence campaigns. I decided to take initiative and design a successful campaign about unhealthy relationships.


I talked to several members of the DoSomething staff that worked on the campaign, trying to pick apart what went wrong. I also examined other DoSomething campaigns in the same problem space, as well as other national campagins to raise awareness about unhealthy relationships. I gathered two major insights:

  1. Campaigns targeted too wide of an age range. High schoolers and college students experience relationships in very different ways.

  2. Education is focused on physical and sexual abuse, neglecting more subtle signs of an unhealthy relationship.

This information pushed me to aim this game at a younger demographic, in order to start education about healthy relationships early. High schoolers especially have little experience with deeper emotional relationships, and as such are at a high risk for unhealthy relationships. As someone who was not-quite-so-recently in high school, I saw this reflected both inside and outside the classroom. I wanted to make a game that empowered players to recognize the signs of unhealthy relationships and take action to stop these behaviors.

I spent my first week at DoSomething reviewing academic literature on games, social activism, and offline impact. With no prior video game work done at DoSomething, I created a repository of articles, studies, and interviews with game developers to inform our process. From both published research and interviews, I identified one driving factor in all successful games: the invocation of empathy. While many factors can help make a game memorable - such as interesting mechanics or innovative gameplay - without making the player feel something, the cause is lost. Designers can focus on certain values to elicit an emotional response from the player. From there, I researched signs of emotional abuse, many of which had to do with communication and language used to undermine a friend or partner.


In order to best mimic decisions in everyday life, I determined a dialogue-based point-and-click game would allow the content to shine over complicated mechanics. Many games in this style focused on bystander intervention, but I wanted to go in a different direction - at least initially. A part of my vision for the offline impact of the game involved the player not just stepping in to educate others, but also watching and correcting their own behavior.

Dialogue chart I created for the second scene, taking place at the dance. Blue is the player, with other colors representing supporting characters. Choices are indicated by double-headed arrows.

The basic layout of the game was established: exhibit abusive behavior, learn that it was abusive, learn how to make up for past mistakes, and then educate another character through bystander intervention. Of course, the game had to resonate with high schoolers especially, so I chose a quintessential secondary school experience: prom.


Before we could begin the development process, we needed to choose a platform. Throughout this process, my foremost concerns were reaching the most people, eliminating engagement blockers, and designing for virality. Our two options were either a web game or a mobile application.

DoSomething is a global compay, and engages 2.8 million members through SMS. I was worried that having to go to the app store to download a game would be a blocker for engagement, and would also eliminate a huge part of DoSomething's member base that was exclusively contacted through email. To make the game accessible to as many people as possible, decided to host on the web, and make the game responsive to mobile devices. After researching different game engines, we decided to use Phaser, a JavaScript/HTML5 framework, over others (like Unity Web Player) that at the time required a separate download.

We quickly iterated through several prototypes, and ended the summer internship with our first "official" prototype, which we had hosted on a GitHub Pages website. We sent it out to several game developer friends, tested it with our target audience, analyzed their feedback, and put together a plan for the next few months.

Early Phaser prototype testing content flow and UI elements.


The first prototype we worked on over the summer had three short scenes of dialogue choices, along with narrative copy to set up the game. I first drafted the game dialogue in a physical flowchart made from post-its notes, then implemented it in our prototype. As our project progressed past the summer, I did a complete rehaul of the simplistic game script. I created a prototype to test content flow using Twine, a text adventure game engine, and then reached out to DoSomething contacts at Planned Parenthood, the Center for Disease Control, and ROAR for Good for a review. I wrote everything from scratch, and while I had done extensive research about unhealthy and abusive relationships, I wanted a professional opinion about the scenarios we presented and the actions we recommended.

Testing another UI design in Phaser using stock photos.

For our final demo, my team hired a contract artist to illustrate one of the scenes from the game. The demo showed content progression, potential UI design, and a product that management could touch and feel. The demo functioned as an example of what DoSomething could produce in the digital game space.

Screenshot of the final demo, with art by Kimberly Lee.

Our first complete Phaser prototype from the summer is available to play on the web here. The Twine content flow is playable in the web browser here. The final demo is available for download on Github here.