Eighty-four percent of top-tier mobile strategy titles include a “hidden handicap” that reduces enemy efficiency by up to twelve percent after a player loses two consecutive matches. It is a quiet, algorithmic mercy, baked into the code by developers who have realized that a frustrated player is a player who eventually closes the app. In the industry, they call it “churn prevention.” In the living room, at , Iris calls it a lie.
She sat in the quiet of her apartment, the air still smelling faintly of the rain that had finally stopped an hour ago. On her lap, the tablet was a window into a dozen different battlefields, and she had conquered them all. She had led armies, solved intricate puzzles, and outsmarted “expert” AI across five different genres. But as she watched the “Victory!” banner scroll across her screen for the ninth time that night, she felt a strange, hollow gnawing in her chest. It wasn’t the satisfaction of mastery; it was the suspicion of a fix.
She had noticed it three games ago. She had made a blatant tactical error-the kind of mistake that should have ended the match instantly. She had left her flank exposed, a gaping wound in her formation. The AI had the resources to crush her. It had the logic to see the opening. But instead of striking, it had hesitated. It moved a unit to a meaningless corner of the map. It “fumbled” just enough to let her recover.
Iris didn’t want to be a queen. She wanted to know if she was a good player. But the world she lived in-this digital, curated landscape of engagement metrics and retention loops-wasn’t interested in her growth. It was interested in her time. And time is best harvested in a state of mild, constant flattery.
I understand her frustration more than I’d like to admit. Last Tuesday, I managed to lock my keys inside my sedan while it was still running in the driveway. It was a stupid, human error, born of rushing and a momentary lapse in spatial awareness. As I stood there, listening to the muffled hum of the engine through the glass, I realized something: the car didn’t care.
It didn’t “sense” my escalating panic and decide to pop the lock as a gesture of goodwill. It didn’t analyze my “user journey” and determine that a locked door would lead to a negative brand sentiment. The lock was a physical reality. It was an uncompromising “No.”
The Integrity of “No”
Physical systems don’t have a mercy setting. They offer the dignity of a real consequence.
There is a profound, almost spiritual respect in a system that refuses to budge. The physical world, for all its jagged edges and inconvenient physics, is honest. When you fail in the physical world, the failure is yours. It belongs to you. You can own it, study it, and eventually overcome it. But when a piece of software “lets” you win, it steals your agency. It turns your effort into a pantomime.
This shift toward “soft” difficulty isn’t accidental; it’s the result of a massive industrial pivot that occurred in the . In the era of the arcade, games were designed to be brutal. The “opponent” was a quarter-hungry predator. If a game of Gauntlet or Dragon’s Lair was too easy, the machine failed to generate revenue. The ceiling was high because the floor was expensive. But when the primary theatre of gaming shifted from the arcade to the home console, the economic model flipped.
Suddenly, the goal wasn’t to kill the player in three minutes; it was to ensure the player didn’t return the cartridge to the store the next day. Developers began experimenting with “Dynamic Difficulty Adjustment” (DDA). If the player died too many times on Level 4, the game would secretly lower the health of the boss or increase the frequency of power-up drops. This was the birth of the “padded” experience. We traded the dignity of a hard-earned loss for the cheap sugar high of a managed victory.
Arcade Era
Designed for high skill-ceiling to maximize quarter-per-minute revenue.
Home Console
Designed for retention to minimize returns and maximize playtime.
The economic flip that traded the dignity of failure for the comfort of engagement.
Searching for the Ceiling
The problem with this is that you can never truly measure yourself against a ghost. If the goalposts are on wheels, you never know if you actually kicked the ball far enough.
Iris started searching. She wasn’t looking for a “game” anymore. She was looking for a ceiling. She wanted an opponent that didn’t have a “mercy” setting. She eventually drifted away from the flashy, high-budget titles and toward the stark, minimalist world of recreational mathematics. That’s where she found the game of sim.
On the surface, it looked like child’s play. Six dots arranged in a circle. Two players taking turns drawing lines between them. The goal? Don’t be the one to complete a triangle of your own color. It’s a game of forced outcomes, a classic demonstration of Ramsey theory. Specifically, it’s a manifestation of the theorem , which proves that in any group of six people, there are either three who all know each other or three who are all strangers. Translated to the game board, it means that a draw is mathematically impossible. Someone must lose.
This is the kind of honesty Iris was craving. There is no “luck” in a game like this. There are no dice, no random spawns, and no “rubber-banding” AI. There is only the graph.
However, even in the world of simple logic games, the “retention trap” exists. Most versions of this game you find online are built by hobbyists or app-factories. They use basic heuristic AI that plays “well enough” to be interesting but often falls into predictable patterns. They are designed to be “fun,” which is often a coded way of saying “winnable by a casual user.” They are still part of the hall of mirrors.
Then she found Triad.
Triad is different because it treats the game not as a toy, but as a mathematical truth. It doesn’t just offer “Easy” or “Hard.” It offers a mathematically perfect solver. This is the “perfect opponent” that the rest of the industry is terrified of. If you play against the solver, and you make a single sub-optimal move, you have already lost.
The machine will not fumble. It will not “hesitate” to give you a chance to recover. It will follow the cold, unwavering logic of the graph until you are forced to complete that final, losing triangle.
Playing against a perfect solver is a disorienting experience. It feels less like a competition and more like walking into a wall. You realize, very quickly, that your “intuition” is actually just a collection of bad habits that other, softer AIs have allowed you to keep. You realize that you haven’t been playing the game; you’ve been playing the developer’s mercy.
When Iris played her first match against the Triad solver, she lost in eleven moves. She sat back, staring at the screen. For the first time all night, the gnawing in her chest was gone. It was replaced by a sharp, cold clarity. She hadn’t been cheated. She hadn’t been “handled.” She had simply met the edge of the world.
The Solver’s Respect
There is a common assumption in game design that meeting a perfect AI is discouraging. The theory goes that if a player realizes they can’t win, they will stop playing. And, for many, that’s true. But for the person who actually wants to master a craft-whether it’s chess, or coding, or a simple pencil-and-paper game of dots-the perfect opponent is the only one worth meeting. It is the only one that can tell you the truth about your own progress.
The solver doesn’t hate you. It doesn’t want to crush your spirit. In fact, it doesn’t want anything at all. It is simply a reflection of the game’s inherent structure. When you lose to it, you aren’t losing to a “boss”; you are losing to the math itself. And that is a respectful way to go.
We spend so much of our lives surrounded by systems that are trying to “engage” us. Our social media feeds are tuned to our outrages; our streaming services are tuned to our comforts; our games are tuned to our egos. We are being “managed” into a state of perpetual, shallow satisfaction. We have become so accustomed to the “mercy rule” that we’ve forgotten the thrill of a genuine challenge.
I think back to my car door. Standing on the driveway, feeling the cold metal of the handle, I was forced to problem-solve. I had to call a locksmith. I had to wait. I had to pay for a three-minute job involving a pressurized bag and a long metal rod. It was inconvenient, expensive, and entirely my fault. But when that door finally clicked open, the relief was real.
It wasn’t a “Victory!” banner generated by an algorithm to keep me “retained.” It was the resolution of a physical conflict.
The Triad implementation of Sim offers that same kind of resolution. By providing a perfect solver alongside the educational resources to understand why the solver is perfect, it moves the experience out of the realm of “distraction” and into the realm of “learning.” It gives you the floor and the ceiling, and then it steps back and lets you climb.
It is the only thing in the room that respects you enough to let you fail. Iris didn’t close the tab after her first loss. Or her second. Or her twentieth. She wasn’t “churning.” She was recalibrating. She was looking at the six dots and the intersecting lines and realizing that the game was much larger than she had thought. It wasn’t about winning; it was about the pursuit of a perfect line.
We don’t need more games that let us win. We have plenty of those. We have entire industries dedicated to the art of the participation trophy. What we need are more systems that are honest enough to show us exactly how far we have left to go. We need the “No” of the locked door. We need the cold logic of the graph.
We need to meet the opponent who refuses to move the goalposts, because only then can we trust the moment when we finally, truly, put the ball through.
Late that night, when Iris finally did manage to push a match into a deep, complex mid-game against the solver, she didn’t need a “Victory!” banner. The tension in her shoulders and the focus in her eyes were proof enough. She had found the ceiling. And now, for the first time, she knew exactly how high she could reach.