第2章 The Pattern

Daniel returned to the civic center at seven the next morning. He had told himself he was coming back to review the Vance case, to prepare for the first appearance, to do the job he had been assigned. But the truth was simpler and uglier: he needed to know if the machine was right. He logged into Axiom Court Analytics and pulled up his dashboard. The Elias Vance case was still there, the red prediction unchanged. But something had shifted. A new tab had appeared on the left side of the interface, one he did not remember seeing the night before. It was labeled *Historical Analysis*. He clicked it. The page loaded slowly, as if the system were processing a large dataset behind the scenes. When it appeared, Daniel felt the floor tilt beneath him. The page displayed a list of cases, going back twelve years, from 2014 to 2026. Every case Daniel had personally worked on was there. Every defendant he had represented, every charge, every outcome. And next to each case was a prediction, generated sometime before the trial, with a confidence score and an accuracy rating. His stomach tightened. He scrolled through the list. A drunk driving case from 2015 — prediction: guilty, accuracy: correct. A drug possession case from 2016 — prediction: guilty, accuracy: correct. A domestic violence case from 2018 — prediction: dismissal, accuracy: correct. The system had been predicting his cases for years, and every single one had come true. But it was the cases he had lost that made his hands shake. Twelve cases where the prediction had been acquittal or dismissal and the outcome had been conviction. Twelve clients he had failed, twelve people who had gone to prison because the system had known something he had not. The machine had seen it all coming. It had told him, in cold statistical language, that he was going to lose, and he had lost anyway. Daniel closed his eyes. He thought about Marcus Wells, the armed robbery case from the night before. He thought about Elias Vance, sitting in a county jail right now, waiting for a lawyer who did not know how to fight a prediction. He thought about every client he had ever lost and every verdict he had accepted as inevitable. He opened his eyes and typed a new query into the search bar: *Elias Vance*. The system returned the case file, but this time it included something new — a timeline. The prediction was not just a single outcome. It was a series of projected events, mapped out day by day, from the first appearance through sentencing. Each event had a probability score. Each score was above ninety-five percent. Daniel read the timeline carefully. *Day 1: Plea negotiation fails. Day 3: Key witness recants. Day 7: Juror substitution — alternate replaces primary. Day 14: Closing argument shifts narrative. Day 21: Verdict: guilty.* He printed the timeline and folded it into his breast pocket. Then he printed every other case the system had on record, every prediction, every outcome, every failure. He stacked them on his desk and stared at them until his coffee went cold and the rest of the office arrived and began their own quiet wars with their own quiet defeats. At lunch he walked to the courthouse and stood outside Courtroom 3B, watching the lawyers file in and out. He had always thought of the courthouse as a building where truth was supposed to be found. Now he wondered if truth was just another variable in an equation he did not understand. A woman in a dark suit walked past him, heading into the courtroom. She was maybe thirty-five, with sharp features and sharper eyes, carrying a leather portfolio like a weapon. Daniel recognized her from a dozen trials. Claire Whitfield, assistant district attorney, conviction rate of ninety-two percent, known for never losing a murder case. She was the one assigned to prosecute Elias Vance. He followed her into the courtroom. She was already at the front of the room, arranging her files with precise, economical movements. She did not look up when he entered. She did not need to. She already knew what he was there for. "Mr. Mercer," she said, though she could not have heard him. His lips had formed the shape of his name, and she read it with the practiced ease of someone who had spent years watching lawyers instead of listening to them. "I was wondering when you would show up." "I've been assigned to the Vance case," Daniel signed, pulling out his phone to type the message. Claire glanced at his phone screen and nodded. "Yes. I know." She continued arranging her files. "You have read the Axiom report, I assume." Daniel froze. "How do you know about that?" Claire set down her pen and looked at him directly. Her expression was not unkind. It was something worse: resigned. "Because I have been using Axiom since it launched three years ago. Because every prediction it has made in this courtroom has come true. Because Richard Hale — the chief prosecutor — built the damn thing." Daniel felt the room shrink around him. "You are telling me this system is rigged." "I am telling you it is accurate," Claire signed. "There is a difference. The system does not rig outcomes. It predicts them. And it has never been wrong." "Then why are you still here?" Daniel signed. "Why keep prosecuting if you already know how it ends?" Claire's expression softened, just slightly. "Because knowing the ending does not make it any easier to live with. And because some of those predictions are wrong about the right things. People still deserve a chance, Mr. Mercer. Even when the machine says they do not." She picked up her portfolio and walked toward the door. At the threshold she paused and looked back at him. "The next hearing is Thursday. I suggest you come prepared. Not to argue the facts. To argue the system." Then she was gone, and Daniel was alone in an empty courtroom with a stack of paper predictions and the terrible certainty that he had just stepped into a war he did not understand. The room smelled faintly of floor wax and old wood, the same smell it had always carried, but now it felt different — as if the walls themselves were watching him, waiting to see what he would do next. He gathered the papers from his desk and slipped them into his briefcase, knowing that the case files contained more than legal documents. They contained possibilities — small, fragile, almost invisible chances that no algorithm could quantify and no machine could dismiss. ---