Chapter 44: The Crack
[Westfield Defense Systems, Seattle — December 13, 2007, 1:45 PM]
The handler was supposed to be in the building at two o'clock.
I'd modeled the pattern through the Library — Monday, Wednesday, Friday visits, arrival times between one-thirty and two-fifteen. Today was Thursday. Not a scheduled day. But the Library's Pattern Recognition had flagged an anomaly in Vasquez's morning behavior: she'd canceled two afternoon meetings, cleared her calendar through four PM, and sent Kwan to a off-site procurement review that removed him from the building.
Calendar clearance plus Kwan's removal equaled handler meeting. The pattern shift suggested urgency — something that couldn't wait for the regular Friday slot. I adjusted my prediction: handler arriving today, two PM, northwest stairwell where the camera coverage had a gap between the second and third floor landings.
I positioned myself in the second-floor break room at one-forty-five. The break room's window overlooked the northwest stairwell entrance. A cup of coffee from the building's machine — bitter, institutional, the kind of coffee that existed to fulfill a contractual obligation rather than a culinary one — served as my cover for the extended stay.
Two o'clock. The stairwell door stayed closed.
Two-fifteen. Nothing.
Two-thirty. A janitor passed through, pushing a cart. Nobody else.
By two-forty-five, the coffee had gone cold and the prediction had gone wrong.
I ran the Library's analysis again: Handler meeting indicators. December 13. Vasquez calendar anomaly. Kwan absence.
The cross-reference returned the same probability — seventy-eight percent likelihood of handler meeting. But the meeting wasn't happening at the northwest stairwell. It wasn't happening anywhere I could see.
My phone buzzed. Sarah: V left building at 2:10 via parking garage. Currently at cafe two blocks east. Meeting someone. Not our handler — different man. Can't get close enough for photo.
The meeting was off-site. The calendar clearance hadn't been for an in-building handler visit. It had been for an external meeting with someone I hadn't identified. My prediction — built on a pattern model extrapolated from five days of observation and filtered through meta-knowledge assumptions about Fulcrum scheduling protocols — had missed the target entirely.
I'd been watching the wrong door.
---
The security flag arrived in my inbox at four-twelve PM. An automated notification from Westfield's HR system, cc'd to Elena Vasquez as my direct supervisor.
RE: Employee Attendance — David Chen This is an automated notification. Employee David Chen was absent from his workstation for a period exceeding forty-five minutes during work hours on December 13, 2007. This absence was not logged in the time management system. Please address with your supervisor.
The break room vigil. Forty-five minutes of watching an empty stairwell, logged by the building's passive movement tracking as an unexplained absence from my assigned workstation. The system was automated — badged entry to the break room, no corresponding badge swipe at my desk for forty-five minutes, flag generated.
An hour later, Vasquez stopped by my desk. Her approach was casual — the body language of a manager performing routine oversight. But her eyes held something sharper.
"David. Long lunch?"
"Coffee break went overtime. Lost track of time reading the news." The lie was thin. I knew it was thin. She knew it was thin. The exchange was theater — two professionals recognizing that something was off and agreeing, through mutual performance, not to acknowledge it directly.
"Happens to everyone." She smiled. The Stanford mug was in her hand. "Just log it next time. HR gets fussy about the tracking system."
She walked back to her office. The glass walls framed her as she sat down, opened her laptop, and typed something I couldn't read from my position.
The Library flagged the interaction: Vasquez — noted MC's unexplained absence. Filed internally. Possible counter-surveillance assessment in progress.
I'd made a mistake. A stupid, avoidable mistake born from the assumption that the pattern I'd modeled would hold — that the handler would follow the schedule I'd predicted, that the meeting would occur where I expected, that the show's Fulcrum was still the Fulcrum I was fighting.
The show's Fulcrum was dead. The Fulcrum that existed now — reshaped by Tommy's capture, by the cascade of California operations, by the Prophet dossier that had taught them their enemy could predict their movements — had adapted. Meeting schedules had shifted. Protocols had changed. The playbook I'd memorized across ninety-one episodes was a document written about an organization that no longer existed in the form the writers had depicted.
---
[Cover Apartment — December 13, 2007, 8:30 PM]
"What happened?"
Sarah stood at the kitchen island, arms crossed. Not angry — analytical. The distinction mattered, because angry Sarah deflected and analytical Sarah dissected, and right now I needed dissection more than I needed comfort.
I sat at the table. The prediction notes — handwritten, on a legal pad I'd been using to model the handler's schedule — lay between us. Four pages of analysis. Timing charts. Probability matrices. The careful architecture of a predictive framework that had failed its first real test.
"I was wrong." No qualification. No deflection. The raw admission, delivered the way I'd learned to deliver truths in this life — directly, without the crisis management polish that my previous career would have applied.
"Wrong about what, specifically?"
"The meeting location. The timing. The pattern." I pushed the legal pad across the table. "I built a predictive model based on five days of observation and... extrapolation. The model said the handler would be in the northwest stairwell at two PM today. He wasn't. Vasquez left the building for an external meeting I didn't anticipate. And my forty-five-minute stakeout of an empty stairwell got flagged by the attendance system."
Sarah picked up the legal pad. Read the notes. Her expression didn't change — the professional mask that concealed the analysis happening underneath. Through the bond, I sensed the rapid processing: assessing the damage, evaluating the exposure risk, calculating the operational adjustments required.
"The extrapolation," she said. "What was it based on?"
The truth pressed against my throat. A television show I watched in another life. A story I memorized like scripture. Five seasons of Fulcrum operations that I assumed would hold despite two months of changes I'd personally inflicted on the organization.
"Historical patterns from the Intersect's intelligence files," I said. "Fulcrum scheduling protocols. Handler rotation standards. I assumed the protocols would be consistent."
"And they weren't."
"No. The changes we made in California — Tommy's capture, the cell dismantlements — they've propagated. Fulcrum has adapted. Their protocols have shifted in ways the historical data doesn't reflect."
Sarah set the legal pad down. The bond carried her assessment as it crystallized — not the words, just the shape of the conclusion forming in her mind. She was recalibrating. Not just the mission. Her understanding of me.
"You've been doing this," she said. Quiet. Not accusatory. Observational, the way a scientist observes an experiment producing unexpected results. "Since the beginning. The intelligence you've provided — the agent identifications, the operational predictions, the Graham assassination warning. You've been working from a model. A comprehensive model that predicted outcomes before they happened."
"Yes."
"And the model is breaking down."
The words landed with the precision of a thrown knife. Sarah Walker, intelligence professional, had just described my meta-knowledge without knowing what it was — had identified the architecture of my impossible predictions through pure analytical reasoning, arriving at the correct structural conclusion while missing the foundational cause.
"Yes," I said. "The model is breaking down."
"Because your changes broke it."
"Because I changed the inputs. Every intervention I've made since September has shifted the variables. The further we get from the original parameters, the less reliable the predictions become."
She studied me. The bond was wide open — not because she'd chosen to lower her defenses, but because the analytical engagement was so intense that her walls thinned by default, the way a castle's defenders thin when all attention is directed at a single threat.
"You're not going to tell me what the model is."
"I can't. Not all of it."
"Can't, or won't?"
"Both."
She held the silence for six seconds. Then she picked up her wine. Drank. Set it down.
"Then we work without it." Simple. Decisive. The pragmatism of a woman who'd been given incomplete information and chose to build forward from what she had rather than backward from what she didn't. "No more predictions. No more pattern extrapolation from whatever source you've been using. From this point, we run real-time intelligence. Classic methodology. Surveillance, analysis, dead drops, shoe leather."
"Agreed."
"And the attendance flag?"
"I'll manage it. Extra hours logged this week. Overtime justification. The cover can absorb one anomaly."
"One." She held up a finger. "Not two. If Vasquez flags you twice, the cover degrades past recovery."
I nodded. The legal pad sat on the table between us — four pages of predictions that had been built on a foundation I could no longer trust. The meta-knowledge that had been my greatest asset since September had just demonstrated its greatest liability: it was wrong. Not occasionally wrong, the way any intelligence model produced false positives. Structurally wrong. Built on a timeline that my own actions had dismantled.
Sarah gathered the pages. Carried them to the kitchen sink. Turned on the faucet. Held the corner of the first page to the gas burner on the stove until it caught, then dropped the burning paper into the wet sink. The fire consumed the predictions in curling orange light, the ink blackening, the careful handwriting dissolving into ash.
She burned all four pages. Ran the water until the ash washed down the drain.
"Fresh start," she said. Turned to face me. The bond carried something new underneath the analytical focus — respect. Not for the predictions that had failed. For the admission that they'd failed. "You were wrong, and you said so. You didn't deflect. You didn't rationalize. You just said it."
"Being wrong isn't the problem. Pretending you're right is."
The corner of her mouth moved. That micro-expression — the ghost of the smile from the diner, from the bourbon sessions, from the moments when something I said crossed the distance between us and landed on the right frequency.
"The Bryce I knew would have blamed the intelligence. Would have said the data was corrupted, or the analysis was someone else's fault." She turned off the burner. "You burned the data yourself."
I didn't respond. The moment didn't need words. It needed the acknowledgment of failure processed, accepted, and converted into forward motion.
The rain picked up against the windows. Seattle's perpetual percussion, steady and indifferent.
Read the raw, unfiltered story as it unfolds. Your support makes this possible!
Find it all at patreon.com/Whatif0
Timeline Viewer ($6): Get 10 chapters of early access + 5 new chapters weekly.
Timeline Explorer ($9): Jump 15-20 chapters ahead of everyone.
Timeline Keeper ($15): Get Instant Access to chapters the moment I finish writing them. No more waiting.