Arturo Lloyd-Escribano: Systems & Narrative Designer
Building play experiences that read the room and the data. From coaching an undefeated esports roster to turning live-service telemetry into design moves, my career revolves around one question: What story is the system telling, and how can I make it unforgettable for the player?
About Me
What do an esports coach, a game designer, and a data scientist have in common? They all seek to understand complex systems and use data to predict and influence behaviour. My career has been a journey through these different worlds, and the common thread has always been my passion for finding the story behind the numbers.
As a competitive E-Sports Head Coach, I didn't just motivate players; I analysed performance data and opponent tendencies to develop winning strategies, leading my team to an undefeated season. As an English Teacher, I moved beyond traditional methods, using student performance data and gamification to design personalised learning paths that boosted engagement and comprehension.
This fascination with systems and human behaviour led me to a Master's in Game Design, where I learned to build the interactive worlds that captivate millions. It was here that my path fully crystallised. I realised that whether it's a player in a game, a student in a classroom, or a customer on a platform, their actions tell a story. You just need the right tools to read it.
Ghunter: Subversive Narrative Design
My design for "Ghunter" focused on creating a subversive experience guided by three core pillars, ensuring every story beat, character motivation, and environmental detail contributed to a world that appears simple but hides unsettling truths.
Aesthetics: Cute-Gore
Blending innocence with brutality to create a dynamic, memorable world that keeps players engaged and slightly off-balance.
Motivation: Discovery
Rewarding curiosity with moments of surprise and novelty to foster long-term exploration and understanding of the world's mysteries.
Action: Hunt N Cook
Justifying and giving meaning to the core gameplay mechanic, ensuring gameplay and narrative are always coherent.
Dual-Layered Plot: A Subversive Story
The story is intentionally structured with two distinct layers, allowing players to choose their level of engagement and rewarding those who dig deeper.
The Surface Plot (The "Cute" Layer)
A heroic journey to become the best GnomeChef and save the ailing king. This path offers straightforward objectives, clear progression, and a classic "happily ever after" ending for traditional players.
The Subversive Plot (The "Gore" Layer)
For curious players, a darker, optional narrative unfolds. Hidden clues, esoteric side quests, and environmental storytelling reveal unsettling truths about the kingdom, leading to a darker, thought-provoking ending.
Branching Dialogue & Player Discovery
Key narrative branches and dialogue options are locked behind gameplay achievements, reinforcing the "Discovery" pillar and making truth-uncovering an earned experience.
Solve Statue Puzzles
Each statue presents a riddle requiring a specific dish, demonstrating mastery of "Hunt N Cook" mechanics.
Gain "Fragment of Lore"
Successfully solving a puzzle rewards players with a lore fragment, unlocking deeper narrative content.
Unlock New Dialogue
Possessing fragments opens hidden dialogue trees with NPCs, allowing players to question the narrative and delve into the subversive plot.
Uncover the Truth
This gameplay-gated storytelling ensures deeper understanding is tied directly to player skill and exploration.
Worldbuilding: Justifying the Absurd
The world of "Ghunter" is designed to be inherently fantastical to support the core gameplay. Its foundation, the "Cosmic Cauldron," immediately establishes a setting where absurd elements feel natural.
Key Characters & Narrative Function
- Guncle (Mentor & Mystery): The player's cynical uncle, whose scattered notes serve as tutorials and breadcrumbs to the subversive plot.
- King Grancis (Antagonist): On the surface, a merry but insecure king. Subversively, he's a manipulative tyrant addicted to "divine dishes" for euphoric visions.
- Goddess Caldaria (Eldritch Deity): Publicly, the goddess of feast. Lore reveals her as a horrific entity sustained by the suffering of creatures during "Hunt N Cook."
Ghunter's Core Mechanic: "Caza-Cocina" (Cook N' Kill)
The central pillar of "Ghunter" is its unique core mechanic, "Caza-Cocina" or "Hunt-Cook." This system redefines crafting within action-adventure by seamlessly integrating it into real-time combat. In "Ghunter," fighting is simultaneously an act of creation.
The "Cook N' Kill" Process
Engage & Season
Locate an enemy. Use ranged weapons (Grill, Boil, Fry) to damage them, simultaneously filling a 'Cooking Bar' to season the creature alive.
Add Flavours
Optionally, use crafted 'Skewers' to add sour, spicy, or sweet flavours, applying buffs/debuffs and further defining the final dish.
Execute Finisher
Once the Cooking Bar is full, the enemy is vulnerable. Deliver the final blow with a skill-based melee finishing move.
Create the Dish
The resulting dish is determined by the Cooking Style (ranged weapon), Finishing Move (melee weapon), Enemy Type, and any added Flavours.
Player's Toolkit: Ranged Weapons
Grill (Flame Shotgun)
Close-range, cone-shaped fire. Ideal for grilling multiple enemies, requiring close combat.
Boil (Water Beam)
Concentrated water beam, strengthening with sustained aim. Rewards precision and focus.
Fry (Oil Grenade Launcher)
Projectile launcher firing oil grenades, exploding on impact. Allows strategic, indirect area attacks.
Player's Toolkit: Melee Weapons
Cleaver (Slice)
Wide, slicing attack. QTE is a timed sequence of button presses, testing rhythm.
Tenderizer (Smash)
Heavy, smashing attack. QTE is a rapid button-mashing event, testing speed.
Skewer (Pierce)
Long-range, thrusting attack. QTE is a precision timing event, testing reflexes.
The "Hunt-Cook" mechanic results in a deep, combinatorial system for creating dishes. With multiple enemy types, three cooking styles, three finishing moves, and five flavours, the number of potential unique recipes is vast, encouraging constant experimentation and mastery of combat techniques.
Ghunter: Interconnected Systems Design
Ghunter's design is built upon a triad of deeply interconnected systems: Aesthetic (Cute-Gore), Motivational (Discovery), and Action (Hunt N Cook). This ensures every element influences and justifies another, creating a cohesive and dynamic player experience.
1
Aesthetic: Cute-Gore
Governs visual and emotional tone, creating thematic tension and influencing player perception.
2
Motivational: Discovery
Provides intrinsic goals, rewarding curiosity with new information, mechanics, and narrative threads.
3
Action: Hunt N Cook
The core gameplay loop, designed to support and integrate with both the Aesthetic and Motivational systems.
System Dynamics: A Continuous Loop
These systems don't exist in isolation; they constantly feed into one another, creating a rich and engaging gameplay cycle.
Motivation Justifies Action
The desire to Discover new things drives engagement with Hunt N Cook gameplay.
Aesthetic Influences Action
The Cute-Gore juxtaposition changes the impact of Hunt N Cook actions.
Action Fulfills Motivation
Creating dishes through Hunt N Cook provides moments of Discovery, fostering further curiosity.
Emotional Systems: Guiding Player Journeys
Beyond mechanics, systems were designed to evoke specific emotional responses, mapped out via detailed flowcharts.
The "Hunt N Cook" Emotional Loop
Craving (Flow State)
Maintaining engagement between boredom and frustration.
Triumph
Achieved upon successfully overcoming challenges.
Indifference
The outcome of failure, prompting re-engagement or adaptation.
The "Cute-Gore" Emotional Loop
Pity/Guilt
Initial reaction to the cute presentation and hunting creatures.
Sadism
Overcoming guilt to achieve objectives, a subversive transition.
Catharsis
The ultimate resolution, unique to Ghunter's identity.
By intricately designing these emotional pathways, Ghunter's systems work together to create a memorable and psychologically engaging experience that transcends simple mechanics.
Ghunter: User Experience Design
My design philosophy for "Ghunter" centered on creating a responsive, clear, and satisfying gameplay loop. Every element, from character movement to attack impact, was intentionally designed to feel intuitive and reinforce the core "Hunt N Cook" fantasy.
Game Feel & Player Movement
The "feel" of controlling Ghunter was a primary focus, ensuring predictable, responsive, and tactical movement. An adaptive camera system further enhances player awareness in different contexts.
Predictable Movement
Ghunter moves at a constant walking speed with no acceleration, ensuring reliable controls for mastery.
Tactical Dash
A dual-purpose mechanic for dodging and initiating Finishers, with a deliberate recovery time adding risk and reward.
Adaptive Camera
Automatically adjusts between isometric exploration and zoomed-out tactical combat views for optimal awareness.
Visual Clarity & Impact
Clear communication is essential. I designed "Ghunter's" visual language to be immediately understandable, keeping players immersed while providing crucial feedback.
Diegetic HUD
Minimal on-screen information; "Cooking Bar" appears over enemies, and weapon VFX indicate cooking style.
Attack Previsualization
Charged melee attacks show their area of effect on the ground, enabling precise aiming and strategic play.
Impactful Feedback
Slapstick-inspired animations, clear VFX (e.g., sword trails), and distinct SFX (e.g., "ding" for cooked enemies) provide satisfying action.
The Finisher QTE System
The Finisher Quick-Time Event (QTE) is the peak of the core gameplay loop, designed as a moment of skill-based satisfaction rather than mere chance.
Cleaver (Sequence)
A timed sequence of button presses, testing the player's rhythm.
Hammer (Smasher)
A rapid button-mashing event, challenging the player's speed.
Skewer (Precision)
A single, precisely timed press within a moving indicator, demanding reflex.
This system transforms the final blow into a rewarding mini-game, with "Premium" enemies featuring even more complex challenges.
Resource Management & Meaningful Scarcity
The game's economy and resource systems were meticulously designed to create meaningful strategic choices and compelling tension.
Finite Resources
Ammunition and ingredients are limited, requiring exploration and conscious decision-making about collection.
Risk of Failure
Over-aggressive ranged attacks "burn" dishes, yielding no reward, forcing a balance between preparation and defeat.
Cost of Death
Player death results in loss of extra dishes and a percentage of resources, making wilderness ventures significant.
Esports Experience: Head Coach at XAL Esports
My years as the Head Coach for XAL Esports, a competitive League of Legends team, provided me with a unique, high-pressure environment to study game mechanics and player psychology at the highest level. This experience was not just about leading a team; it was a continuous exercise in competitive systems analysis, strategic design, and team management.
Case Study: Building a Cohesive Team Identity
My primary goal upon joining XAL was to transform a roster of talented individual players into a cohesive competitive unit. This involved establishing a clear team identity, creating robust strategic frameworks, and fostering an environment of communication and growth to prepare the team for the NLC Division 2 qualifiers.
My Coaching Philosophy & Methodology
My approach was rooted in the understanding that a game like League of Legends is a hyper-complex system, sharing traits with both the methodical strategy of Chess and the improvisational harmony of Jazz. To manage this complexity, I developed and implemented several core coaching frameworks:
Establishing Team Identity
Collaboratively defined a core team identity emphasizing proactive, aggressive, mid-game compositions with strong team-fighting and crowd control capabilities.
The "Metronome" System
Introduced minion waves as an in-game "metronome" to synchronize player actions, providing a consistent 30-second rhythm for coordinated plays.
Structured Game Planning
Created a "Game Plan Roadmap" breaking the game into four distinct phases with clear objectives, enabling proactive macro-level decisions.
Player Development & VOD Review
Conducted detailed VOD reviews to analyze decision-making, improve communication, and ensure players understood their role within strategic frameworks.
Applied Systems Thinking: Lessons from Esports
My tenure as Head Coach at XAL Esports honed critical skills directly transferable to game design, particularly in understanding complex adaptive systems and human behaviour within them.
Systems & Meta Analysis
Deconstructed every game patch to understand its systemic impact on the meta. This involved identifying balance changes, theorising new optimal champion compositions, and designing counter-strategies before they became mainstream.
Strategic Design & Documentation
Designed and documented the team's strategic frameworks, including our core game plan, objective prioritisation (using the PVPE - Priority, Vision, Push, Execute model), and specific in-game plays.
Player Management & Psychology
Managed the team's internal dynamics, fostering clear communication and a growth-oriented mindset. This included helping players develop their individual skills, manage performance anxiety, and work together as a cohesive unit under high-pressure tournament conditions.
Relevance to Game Design
This experience provided me with a deep, practical understanding of what makes game systems engaging, balanced, and exploitable. The process of analysing a meta, predicting player behaviour, designing winning strategies, and managing a team of highly skilled individuals is directly applicable to designing compelling encounters, creating balanced character abilities, and understanding the complex interplay of systems in a large-scale RPG.
Data-Science Projects
My approach to data science is heavily informed by quantitative analysis. Understanding player behaviour and market trends through data is crucial for creating successful, engaging experiences. My experience in python and
Historical Sales Analysis
Explored videogame sales data (1985-2020) to confirm market saturation impacts sales. Quality and strategic release windows are key.
Predicting Steam Sales (ML)
Developed an ML model (CatBoost) to predict game KPIs before launch, highlighting success drivers like engagement and specific tags. Addressed data bias.
Carbon Footprint (LLM)
Led data science for an AI/LLM-powered app to predict product carbon footprints. Showcases complex data pipeline design and cross-functional leadership.
These projects demonstrate my ability to apply data science, machine learning, and cross-functional leadership to inform and de-risk game development.
Case Study 1: Exploratory Data Analysis of Historical Videogame Sales
Objective & Hypothesis
The core objective was to determine if market saturation within a game genre negatively impacted individual title sales, using a comprehensive 1985-2020 dataset.
Specifically, we hypothesised that higher publication frequency in a genre would correlate with lower total sales for new releases.
Methodology: A Multi-Stage Approach
Initial EDA
Conducted extensive exploratory data analysis on historical sales data to uncover high-level trends across sales, platforms, and genres.
Data Enrichment
Web scraped Metacritic for user scores and precise release dates, enabling a granular analysis of market saturation over time.
Correlation Analysis
Analysed the relationship between genre publication frequency within specific time windows and median sales per title.
Key Findings & Implications
Market Saturation Confirmed
A high volume of genre releases within a 2-month window generally correlated with lower median sales for individual games.
Quality as a Buffer
Higher-rated games (Metacritic 75+) proved significantly more resilient to market saturation, highlighting the importance of critical acclaim.
Seasonal Competition
Q4 (Oct-Nov) showed clear peaks in game publications, indicating the highest competitive density during this period.
This market analysis provides critical, evidence-based insights for strategic pre-production and marketing decisions, helping identify underserved niches and optimal release windows.
Case Study 2: Predicting Steam Game Sales with Machine Learning
Objective: Data-Driven Insights for Game Developers
In a market with thousands of annual releases, predicting success is crucial. This project aimed to develop a machine learning pipeline that forecasts key game performance indicators (KPIs) before launch, providing developers with actionable, data-driven insights. Target variables included: Copies Sold, Wishlists, and Bayesian Score.
Methodology & Data Processing
The project began by fusing several extensive datasets, followed by rigorous cleaning and filtering to ensure data quality.
1
Data Fusion & Filtering
Combined 98,000-game Kaggle dataset, web-scraped Gamalytic data, and HowLongToBeat playtime metrics. Filtered for games with 50+ reviews, resulting in a clean dataset of 7,029 games.
2
Engagement Ratio Calculation
Engineered an engagement_ratio feature to quantify how long players engage with a game relative to its completion time, reflecting player stickiness.
3
Bayesian Score Development
Developed a Bayesian_score by fusing positive review percentages with total review counts, providing a more robust measure of player reception.
4
Extensive One-Hot Encoding
Processed over 500 categorical features, including game tags, genres, and publisher class (Indie, AA, AAA), to prepare data for modelling.
Modelling & Key Findings
Model Selection: CatBoost
After testing multiple models, CatBoost was chosen for its exceptional performance with categorical features, leading to robust predictive capabilities.
High Accuracy Achieved
The final models demonstrated high accuracy, with an R² score of 0.83 for Copies Sold and 0.75 for Wishlists, indicating strong predictive power.
Key Success Drivers
Feature importance analysis revealed critical success factors: player engagement (e.g., engagement_ratio, time_to_beat), community features (e.g., Steam Cloud, Steam Workshop), and specific tags like "Cute" and "Story Rich".
Critical Finding: Data Selection Bias
A significant discovery was the presence of selection bias: the 7,029 games in the filtered dataset represented only the top 7% of the Steam market. This meant the model was trained predominantly on highly successful titles. For example, the 25th percentile of sales in our data was 201.5x higher than the market's overall 25th percentile, underscoring the model's focus on identifying patterns within high-performing games.
Case Study 3: Carbon Footprint Prediction
Project Objective
Designed a full-stack web application to empower brands to measure the environmental impact of their products. The core aimed to use AI and LLM agents for Life Cycle Assessment (LCA), predicting carbon footprints.
Methodology & My Role as Data Science Lead
I was responsible for designing and implementing the entire data pipeline within this multi-disciplinary team.
Web Scraping
Utilised a Firecrawl agent to efficiently scrape product information from user-provided URLs.
LLM Agent for Data Analysis
Developed a LangChain agent powered by Gemini to analyse unstructured data from diverse sources (e.g., PDFs, Excel files), extracting key product specifications.
LCA Prediction
The extracted data was compared against an openLCA database to accurately predict the final carbon footprint (CO2eq).
Deployment
The entire service was successfully deployed via FastAPI on AWS, ensuring scalability and accessibility.
Outcome & Relevance to Game Design
Project Outcome
A functional MVP was delivered, enabling users to input a product via URL or document upload and receive a detailed carbon footprint analysis. My role was central to creating the intelligent data processing core.
Transferable Skills for Games
This project demonstrates my ability to lead complex data architecture, manage unstructured data, and collaborate cross-functionally. These skills directly apply to analysing player feedback, processing telemetry, and making data-informed decisions in game development.