Leo understood the pipeline cold.
He’d spent years at a studio watching games move from raw assets to playable builds: import the mechanics, wire up the interactions, sequence the levels, layer the audio, test the prototype, iterate on every friction point until the whole thing felt seamless.
The System would handle the engine work and code compilation, but the assembly, the sequencing, the creative decisions about what went where and why?
That was all him.
He cracked his knuckles and started building.
Over the next two hours, he directed the System through a rapid-fire assembly process.
Plant modules first, tested individually to confirm their behavior matched what he remembered.
Peashooter firing rate, Sunflower generation timing, Wall-nut durability.
He flagged three inconsistencies and had the System adjust them before moving on.
Then the zombie AI, level by level, making sure the difficulty progression felt natural rather than random.
Then the UI layout, which he scrapped and rebuilt twice before settling on a configuration that kept the action readable even during the heaviest zombie waves.
By the time he had a working prototype, it was playable but raw.
Now he had to make it shine.
The base version was solid.
Faithful to the original, mechanically sound, visually clean.
But “solid” wasn’t going to cut it.
He needed to walk into the PlayForge meeting with something that looked, played, and felt like the product of a professional studio.
The System’s optimization engine was the only tool that could get him there in three hours.
He opened the optimization interface.
[Ding!
Multiple optimization pathways detected.
For this opportunity, only one direction may be selected.] [Optimization tiers available.
Single-Item Optimization: 200 exchange points (targets one specific element).
Primary Comprehensive Optimization: 2,000 exchange points (full-spectrum polish for single-player titles).] [The host currently possesses one Single-Item Optimization token.
To upgrade to Primary Comprehensive, an additional 1,800 exchange points are required.] Leo’s jaw tightened.
When the hidden task had rewarded him with an “optimization opportunity,” there’d been zero mention of tiers.
No fine print.
No disclosure.
The System had handed him what he’d assumed was a full-service upgrade and then quietly revealed, at the exact moment he needed it most, that it was the budget option.
The real thing cost extra.
He couldn’t exactly pull the System out of his skull and give it a talking-to.
So he swallowed the irritation.
“Fine.
1,800 points.
Do it.” He still had over 10,000 exchange points in reserve.
Spending 1,800 stung, but it wasn’t catastrophic.
And if this optimization delivered, the return would be enormous.
Before he committed, though, a thought nagged at him.
If the System was introducing tiered pricing now, there were almost certainly higher tiers waiting.
He asked.
[In recognition of the host’s intention to establish a game development company, the System hereby launches its full Optimization Service menu.] [Tier pricing: Single-Item, 200 points.
Primary Comprehensive, 2,000 points (single-player indie).
Intermediate Comprehensive, 20,000 points (PC/console/mobile).
Advanced Comprehensive, 200,000 points (large-scale AAA productions).] Two hundred thousand points.
Ten players per point.
Two million paying customers before he could afford a single AAA optimization pass.
Leo sat back in his chair and stared at the ceiling for a long moment.
The System wasn’t just charging for services.
It was building an economy designed to scale with his ambitions, and the exchange rate was punishing.
But the scale was there.
Nearly ten billion people on this planet.
Even a fraction of that market, spread across enough titles, and the numbers stopped being scary and started being inevitable.
He just had to build toward it methodically.
For now: Primary Comprehensive.
One game.
Three hours until the meeting.
He confirmed the purchase.
1,800 points drained from his balance.
The optimization interface lit up.
The System’s Comprehensive Optimization wasn’t a magic button.
It was closer to an advanced design suite that responded to direction.
His direction.
The System could enhance visuals, restructure difficulty curves, refine audio, and streamline UX, but it needed Leo to tell it what to prioritize and why.
He started with the issue that had bothered him most about the original Plants vs.
Zombies: the sun-collection mechanic.
In the classic version, sunflowers generated suns that floated across the screen, and you had to click each one individually.
Early in a level, this was fine.
The pace was relaxed, clicking suns felt almost meditative.
But in the mid-to-late game, when zombie waves were pouring in and every second counted, those floating suns became a nightmare.
They’d spawn at the worst possible moments, forcing a choice between collecting resources and defending the lawn.
Click the suns, lose your position.
Ignore them, starve your economy.
Leo had argued about this exact friction point with colleagues at the studio in his previous life.
Their consensus: the sun-clicking mechanic was the single biggest barrier to onboarding non-gamers.
People who wanted to enjoy the strategy but couldn’t handle the dexterity demands at higher difficulties.
“System.
Sun collection needs a quality-of-life pass.
Two-second manual window, then auto-collect.
And add a purchasable sun collector as an in-game item so players can spend suns to save time.” [Sun collection UX redesigned.
Auto-collect delay set to 2.0 seconds.
Note: System has added a faint chime differentiating manually collected suns from auto-collected suns.
No gameplay impact.
Aesthetic only.] Leo blinked.
He hadn’t asked for that.
He tested it: a bright little ding for manual clicks, a softer hum for auto-grabs.
A tiny detail, completely unnecessary, and it made the whole thing feel ten times more polished.
He decided not to question it.
But when he actually played through a mid-game wave with the new system running, a different problem surfaced immediately.
With auto-collect active, the economy had gone slack.
Suns just appeared in your bank.
Zero tension, zero decision-making.
The entire resource-management layer that gave the mid-game its teeth had dissolved.
He sat back and chewed his thumbnail.
The original’s sun-clicking was annoying, sure.
But it was annoying on purpose.
It forced micro-decisions under pressure.
Remove that friction entirely and the strategy drained out of the game like water through a screen.
“Split the difference,” he told the System.
“Auto-collect stays, but add a fifteen percent generation penalty on auto-collected suns versus manual clicks.
Reward the players who engage without punishing the ones who can’t keep up.” The System adjusted.
Leo tested again.
This time the balance held.
Manual clicking was still optimal, still rewarding, but auto-collect kept casual players in the game instead of drowning them.
Good.
He moved on.
“Character art next.
Sharpen the plant designs.
Each species needs a distinct silhouette and expression, identifiable at a glance even at thumbnail size.
Same for zombie variants.
And push the resolution ceiling up.” [Character redesign complete.
47 distinct plant silhouettes generated.
Zombie variant library expanded to 23 base models.
Resolution ceiling removed.] “Audio.
Keep the core musical themes but deepen the production.
More layers, better mixing.
Sound effects for planting, sun collection, and zombie kills all need satisfying feedback.
Crisp, punchy, distinct.” [Audio remaster applied.] “Tutorial.
Remove text walls.
Let the mechanics teach themselves through level design.
Sunflower placement first, then peashooters, then the first wave.
By level three, the player should understand every core mechanic without reading a single tooltip.” [Tutorial restructured.
Text prompts reduced from 14 to 2.
Mechanic introduction now sequenced through level geometry.
Estimated new-player onboarding time: 90 seconds to first zombie kill.] Ninety seconds to first zombie kill.
He hadn’t asked the System to measure that.
It had calculated an onboarding metric on its own, unprompted, like a tool that was paying slightly more attention to the end user than it probably should.
A flash of light swept across the interface.
Optimization complete.
Leo opened the finished build.
The title screen loaded, and the music hit him before anything else.
Same melody he remembered, but warmer, fuller, with a layered arrangement that made the version in his memory feel like a rough demo.
He sat there for a few seconds just listening before he noticed he was nodding along.
He clicked into the first level.
The first sunflower he planted landed with a small, satisfying thunk of commitment, like setting a chess piece down hard.
He placed a peashooter beside it and watched the first pea arc across the screen.
The zombie it hit staggered back half a step.
That stagger was new.
The base version hadn’t had it.
The System had added a hit-reaction animation he’d never requested, and it worked perfectly: just enough feedback to make every shot feel like it connected.
He looked at the Peashooter more closely.
The little thing looked determined.
Jaw set, eyes narrowed, the face of a plant that took its job personally.
The Sunflower beside it radiated cheerfulness.
The Wall-nut, when he placed one on level four, wore an expression of stoic resignation, as if it had accepted its role as a punching bag and made peace with the universe.
They weren’t just game pieces anymore.
They were characters.
Leo played through the first ten levels and watched the difficulty curve the System had built.
Every five levels formed a distinct tier.
The first five were gentle, almost meditative, establishing core mechanics through play rather than instruction.
Then level six introduced a new zombie type, and the difficulty stepped up hard.
Levels six through ten required real strategic thinking, and clearing level ten delivered a rush of earned satisfaction he hadn’t felt from a game in a long time.
Then level eleven hit, and the difficulty spiked again.
It was a brilliant structure.
Five levels of building confidence, then a wall that forced you to rethink everything, then five more levels of earned momentum before the next wall.
Struggle, adapt, break through, coast, struggle again.
The kind of cycle that hijacked your reward circuitry and made “just one more level” feel physically impossible to resist.
His only concern was the severity of the walls themselves.
At the tier-five jump points, the increase was steep enough that a player who’d been cruising comfortably might feel like they’d been dropped into a different game entirely.
Unlike Getting Over It or Life & Death, which were merciless from the first second, Plants vs.
Zombies lured you in with warmth and satisfaction before slamming the door.
The emotional whiplash would hit harder precisely because the game had earned your trust first.
Leo could already picture the forum posts.
The System, it seemed, couldn’t help itself.
Even when optimizing a genuinely wholesome game, it found a way to weave in suffering.
He closed the laptop and checked the time.
Two hours until the meeting.
Enough time to shower, change, and make the trip downtown.
He grabbed his jacket, locked the door, and caught himself halfway down the block humming the main menu theme.
He’d been doing it for three blocks without noticing.