G.A.T.I.N.H.O.S. is a tower defense game created as an academic project at the Federal University of Bahia (UFBA), with the goal of combining entertainment, environmental awareness and fundraising for the cats living on the campus of the School of Fine Arts, animals with no institutional support from the university, kept alive only by students and one professor.
The Problem
From the users
Casual tower defense players abandon games because of repetitiveness, each match is identical to the last, with no real variation.
From the cause
The cats living on campus are in a vulnerable situation, with no institutional support from the university. The cause exists, but has no visibility or scalable fundraising mechanism.
From the product
We needed to build a digital product from scratch that generated real engagement with this audience and converted that engagement into visibility for the cause and donations. On top of that, the project had to align with Fair and Solidarity Trade principles, with at least one axis focused on sustainability and the environment.
Insights from the initial research
The genre was already defined, but before making any mechanics decisions, we ran a survey with 88 respondents to understand where the genre fails and what players actually value. The 3 insights that guided the product priorities:
Repetitiveness is the main reason people quit in this genre
Players want character progression systems
60% play on mobile
Solution (MVP)
We defined an MVP where each feature was mapped to a product need.
Random card system
Random card system: each match is different, fights repetitiveness without increasing development scope; came directly from the research insight
Resolution designed for mobile
Campus cats as the main characters
Wireframe
The wireframes were built in Figma without pixel art intentionally. Working in free resolution before entering the pixel grid allowed us to explore layout and information hierarchy faster, without being constrained by pixel art scale limitations.
Gameplay Loop
The core challenge of the loop was to create a cycle that was infinite without being repetitive. The solution was a random card system triggered between rounds: at the end of each round, the player can buy three randomly drawn cards that upgrade the cats with different abilities and attributes.
What the beta revealed
We launched the beta to the 88 participants of the initial research. To measure the experience, we used two approaches: a structured feedback form and an analytics tool to collect player metrics, tracking time in game, button clicks, wins, losses and skipped tutorials.
What the beta revealed
We launched the beta to the 87 participants of the initial research. To measure the experience, we used two approaches: a structured feedback form and an analytics tool to collect player metrics, tracking time in game, button clicks, wins, losses and skipped tutorials.
The game was too easy
59.1% of players found the level easy or very easy
Full game rebalancing.
Creation of a cat specialization system to increase strategic depth.
The tutorial was being ignored
52% of players skipped the tutorial, creating a barrier between the player and the game.
Tutorial integrated into the game flow.
A talking cat that interacts with the game and the player during rounds, without creating a barrier.
Nobody was coming back to play
The "play again" button was clicked only 25 times in 294 sessions.
Implementation of a ranking system with score based on completion time.
Creating competition between players and a concrete reason to try again.
Players were using cats randomly
The game did not show what each cat did before the player bought it, causing frustration when buying the wrong cat.
Creation of a detail page for each cat.
With a short story and context, making it easier for the player to understand if a cat fits their playstyle or not.
Final results
The numbers compared to the beta show the direct impact of each decision made from the data.
Antes
Depois
average time per player
average time per player
"Play again" clicks
"Play again" clicks
considered the levels "Balanced"
considered the levels "Balanced"
of game sessions were completed
of game sessions were completed
302
Players
R$ 55
raised for the cats
Learnings
Research was key to the final outcome of this project. The first round, with 88 players, defined where to focus before any screen was designed. The second, with 122 players, revealed what needed to change before launch. Without them, we would have shipped a repetitive game, with an ignored tutorial and no reason to play a second time, exactly the problems the data forced us to fix.