The Face Game

The Face Game

The problem

A research project asking a question that turns out to matter for both behavioural science and AI alignment: how do profile pictures influence human and AI decisions? Specifically, do certain faces consistently get classified as "team players," "trustworthy," or "competent" — and how does that classification differ between humans and machine-learning models?

To answer it rigorously, the research team needed an interactive web platform that could:

• Run controlled experiments at scale.
• Collect structured behavioural data from human participants.
• Compare human classifications against AI model output for the same stimuli.
• Be reliable enough to underpin academic publications.

The approach

A two-year engagement to design, build, and iterate on the Face Game — a serious-game web application that doubles as a research instrument.

Key design choices:

• Game framing, not a survey. Participants stay engaged because the experience is structured as a game with feedback and progression — even though the underlying mechanic is structured experimental data collection.
• Lab-grade rigour. Trial sequencing, randomization, and data integrity were non-negotiable. The platform is built to academic-research standards.
• Live and operating. The product runs at the-face-game.net and has accumulated participant data over the engagement.

The impact

Two years of sustained work, a live research platform, and the working relationship to show for it. The engagement spanned the period before SwiftX scaled into a team — direct work between Khaled Yacout and the principal investigator.

I worked with Khaled Yacout for two years; we collaborated on developing an online serious game for research, and my positive experience with him speaks to the kind of service and value SwiftX provides. He was professional, responsive, and really took the time to understand what I needed, making the whole process smooth and straightforward. He consistently delivered high-quality work at a very reasonable rate, and I always felt like the project was in good hands.

— Edmond Awad, Principal Investigator