The temporary research center established by Silph Co. in Glendale had rapidly become a second home to Ray. Each morning after class, he would head straight there, where researchers and engineers now treated him less like a guest and more like a junior colleague. With Professor Oak offering support remotely and Dr. Lysand leading the engineering side, the project to improve the Aptitude Measurement Tool was in full swing.
That afternoon, the lab hummed with concentrated energy. Multiple terminals were lit up with real-time data streams as the modified scanner ran test cycles on volunteer Pokémon from the local academy. The air smelled faintly of coolant and polished steel.
"Running Calibration Cycle 14," announced Dr. Lysand, motioning for a technician to begin.
The machine emitted a quiet hum and scanned a Growlithe standing on the analysis platform. Lines of light traced across its fur before converging into the central core of the scanner. Numbers and color indexes began scrolling across the main screen.
Aptitude: GreenEnergy Sync Rate: 81%Reaction Curve: 76%Kinetic Output: Stable
Ray studied the readings, tapping on the projected interface with furrowed brows.
"It's still giving us a flat reading," he said, not surprised. "This Growlithe is more responsive than the last one we tested, but it's showing the same classification. Same base color. Same output."
Dr. Lysand adjusted his glasses. "That's expected. The current sensor model still rounds everything into the base classification threshold."
Ray nodded. "That's the flaw. The old model was built to identify base color thresholds—Red, Orange, Yellow, Green, Blue, Purple, Aurora—but it ignores the performance bands within them. Two Pokémon of the same color can be vastly different in practical capability. It's a gap I noticed a long time ago."
The engineers glanced at each other, and Ray tapped his tablet, pulling up a comparative breakdown. Two Pokémon—both labeled Green—were displayed side by side. One was a Beedrill with high attack response but low endurance. The other, a Shroomish, had excellent energy regulation but slow reaction times.
"Same color aptitude," Ray continued, "but anyone who's ever trained knows these Pokémon wouldn't perform equally in the field. That's why we need to assign concrete numeric ranges within each aptitude."
He swiped again, this time bringing up a diagram that split each color into measured spectrums without visually labeling them as sublevels. This was part of the plan—keep internal metrics precise while allowing trainers and evaluators to maintain a simple interface.
"Each aptitude band—Green, Blue, etc.—should cover a numeric aptitude range. Let's say Green spans from 61 to 80. We assign a weighted score based on specific metrics—energy efficiency, reflex time, durability, synergy with moves. The final score places the Pokémon in the appropriate range."
Lysand leaned in. "And we use those ranges to differentiate similar Pokémon without changing the visual output for most users."
"Exactly," Ray said. "The front-end still displays 'Green,' but under the hood, the machine knows if it's closer to 61 or 80. That's enough to make better calls in breeding, battling, or field assignments."
The team murmured in agreement. One engineer, formerly skeptical, muttered, "I think this kid is ten years ahead of us."
As the others resumed calibration work, Ray opened a new file and entered raw data from Pidgeotto's last few sessions. Combat performance, maneuver recovery, emotional synchronization, stamina decay curves. He was still experimenting with the best formula for final scoring, but every trial brought him closer to a consistent model.
Later that day, he received a massive file package from Professor Oak. The message read:"As promised—baseline performance scans from my lab's research. Over 200 Pokémon, tested under controlled and natural conditions. You'll find edge cases and outliers especially useful. Let me know what you see."
Ray's fingers flew over the keyboard as he began comparing Silph's early scan data against Oak's library. He spent hours isolating traits that standard machines missed: how Pokémon responded to confusion status, variance in focus duration, microsecond differences in attack preparation.
One pattern became clear—many Pokémon previously considered "underwhelming" in a shared aptitude group were actually bottlenecked by misread internal efficiencies. With proper identification and training adjustment, they could easily outperform others in the same class.
By evening, Ray had rewritten half the interface logic used in the scoring model. Silph's team was already adapting his framework into a test-ready version.
"Tomorrow we'll begin sensor calibration against real-time sparring matches," Lysand said, rubbing his eyes. "We've already prepped ten academy Pokémon for structured testing."
Ray nodded. "I'll adjust the formulas tonight. I've got a few things to fix."
As the others filtered out for dinner, Ray stayed behind. In the quiet glow of the lab's central screen, he reviewed his notes again. The complexity of Pokémon potential couldn't be captured by colors alone. They were alive—dynamic, evolving. His system would reflect that.
Tomorrow, the machine would read more than just color.
It would read truth.