By February, Sonic had gone through another round of revisions, with noticeable changes made to the gameplay.
Now, following Tetsu Kobayashi's guidance, Atlus began its internal testing phase.
Kazuki Takahashi picked up the controller, looking a little lost. "I'm not an employee. I can still take part in internal testing?"
Yuji Naka replied, "But you've been sitting here for almost a month already. Hurry up, sign the NDA, and go test the game."
Takahashi could only sigh. "You know, I really—"
Yuji Naka rolled up his sleeves.
Takahashi immediately shut up, walked over, and picked up the controller.
This test build wasn't long, only half of a single stage, quickly moving from parkour gameplay straight into a boss battle.
Below the TV, sheets of paper were taped up, explaining the controls—how to move, how to attack, and so on.
Takahashi tried moving Sonic.
Fast.
After developing Jörmungandr and Off-Road Motorcycle, Atlus had already formed its own unique approach to handling high-speed gameplay. It was starting to feel like a defining trait of Atlus titles.
The rough placeholder character on screen already hinted at Sonic's animated form.
Two-head-tall proportions, a big head, spiky hedgehog quills, a blue body, running like the wind.
In just a few minutes, Takahashi understood the game.
Run, collect rings, jump around. Occasionally there were annoying enemies, but most of the time, just running up and punching once was enough to deal with them.
At first, it didn't seem like much—just parkour. But as the stage progressed, Takahashi gradually became immersed.
The controls felt incredibly smooth and satisfying. There were almost no unnecessary obstacles.
The player only needed to focus on controlling Sonic's movement. Everything else was on the level designers.
Guided by this deliberate design, Takahashi quickly reached the end of the demo and entered the boss fight.
The boss was an ultimate stitched monstrosity, combining several classic horror elements. It was the creature Ning Ning Jiezaki had once imagined, now painstakingly brought to life by the team.
The stitched monster swung a cleaver, while Takahashi dodged clumsily.
Even here, Sonic's mobility remained unchanged, but without any long-range attacks, fighting the boss became a serious challenge.
Takahashi died several times. As he observed more carefully, a realization slowly formed.
It wasn't that attacks were ineffective—he was hitting the wrong spots.
He started attacking the monster's joints, and this time it worked. After three hits, a joint was destroyed. The corresponding body part broke apart, and the boss switched to a new attack pattern.
"Tsk—" Takahashi sucked in a breath.
What a bold boss fight.
He hadn't expected the boss to have multiple combat modes.
The testing lasted the entire morning and only ended at noon.
Tetsu Kobayashi, Satoru Iwata, Yuji Naka, and the others made some brief notes.
Sitting across from Tetsu Kobayashi, Takahashi said, "I think it's amazing. The final battle is really impactful. When that huge thing approaches, the pressure is intense. I died a few times at first, but once I figured out the joint mechanics, it felt completely different."
He continued, "What impressed me most is that destroying different joints has different effects. Even the order you attack them changes how the fight plays out."
Tetsu Kobayashi nodded.
That had been one of the original design goals of the stitched monster.
How do you show a monster's intelligence?
Reading player inputs was the simplest—but also the most annoying—method.
A computer reading inputs is something players can never truly react to. Even back in the Famicom era, machines already outpaced human reaction speed.
For example, there was a famously unfair FC game, often translated as Kung Fu or Karate. One gate in the game opened and closed by reading player input. The moment a player stepped under it, it could slam shut and crush them—nearly impossible to react to. Even expert speedrunners could still die there unless using TAS.
Designing bosses with multiple abilities that respond differently to player actions was another approach—one that avoided that intense frustration.
Takahashi spoke again. "Also… it feels kind of laggy."
"Lag is normal. That can be ignored," Tetsu Kobayashi replied.
After all, the game was intended for the MS, not the SG, so testing it on the SG naturally caused slowdown.
Takahashi added, "But I think there are some real problems too. Lava floors you can't step on are fine, but in some places, the jump platforms are completely off-screen. You can only guess where they are. That's really annoying."
Tetsu Kobayashi suddenly looked up.
"Who designed that?"
Yuji Naka scratched his head.
He did.
"I thought it would increase the difficulty."
"That's not difficulty," Tetsu Kobayashi said. "That's pure annoyance. What, when you were a kid walking to school, you'd fall into pits just because you couldn't see them?"
Yuji Naka was stunned.
What was that? Yuji Naka's childhood lore?
"Raising difficulty is fine. We can do it through movement restrictions—debuff items that reduce Sonic's speed or jump height. But placing a platform where the player can't see it, especially on the main route, is just frustrating."
Tetsu Kobayashi didn't completely shut the idea down.
That kind of design could work in optional side paths meant for collecting items.
But with the limited storage capacity of games at the time, collection mechanics couldn't be a major focus, so there was little reason to implement it now.
Yuji Naka quickly wrote it down.
Top directive from the boss: no cheap, unfair traps in the game.
After testing ended, Tetsu Kobayashi thought for a moment and asked Iwata and the others, "Based on current progress, how long until development is finished?"
Iwata answered immediately. "Revising completed stages, developing new ones, polishing everything, and testing for bugs—if all goes smoothly, we should be finished by April."
Tetsu Kobayashi nodded quietly.
Originally, the SMS was set to release in October 1985 as Sega's answer to Nintendo in North America. With all the changes now, its release had been pushed to June.
That meant development would still make the launch window.
"But that means I'll have to find someone to manufacture the cartridges myself," Tetsu Kobayashi muttered, pressing his temples.
Even though, like Namco, he now held the right to independently produce cartridges—essentially a hidden profit boost—it came with problems.
Not only did it require tying up a huge amount of capital in cartridge factories, he also needed to find reliable manufacturers.
Small factories lacked resources, and low production efficiency would directly hurt sales.
Tetsu Kobayashi thought it over.
Actually…
He really did know an old friend who could help with cartridge manufacturing.
Please Support me by becoming my patreon member and get 15+ chapters.
[email protected]/Ajal69
change @ with a
Thank You to Those who joined my Patreon