Chapter 42: The 67 Percent Calculation
The calculation began at 6:47 AM.
Adam sat at his desk with his notebook open, running the weekly probability update he had maintained since the spy hypothesis first emerged. The inputs had changed. Entry 9 had been added to Sheldon's calibration log. The access log correlation had tightened — five of nine anomalies now adjacent to Adam's building presence. Folder A had reached nine items. And something new: Sheldon had requested a record of "all instrument calibration anomalies in this building over the past six months" from the facilities department.
That last item had come from Leonard, who had heard about it from someone in facilities.
Adam wrote out the variables:
Calibration log entries: 9 Access correlation: 5 of 9 (55.6%) Folder A items: 9 Spy hypothesis probability (Sheldon's): Unknown, likely 30-35% Facilities records request: Active — new data source
Below that, the model:
Sheldon's threshold system: 25% = passive, 25-40% = active correlation, 40%+ = direct investigation Current state: Transition zone (40%), hypothesis testing mode active Time to correct framework (conditional): Dependent on next data points
He ran the numbers. Not formally — not the way Sheldon would, with precise statistical notation — but the way his Resonance Engine processed probability: pattern recognition across multiple variables, weighted by reliability and cross-correlation strength.
The result populated.
67%.
Adam wrote the number on the page.
Stared at it.
"67%: probability that Sheldon reaches a correct framework for the anomalies within 30 days."
He read it back.
67% was not "he will figure it out." It was not certainty. But it was more likely than not, within a month. The difference between "I might get caught" and "I probably will get caught eventually" had collapsed into a timeline.
I have been managing this for four months. I have been careful, methodical, patient.
The number does not care about any of that.
He closed his notebook. Opened it again. Read the number again.
It had not changed.
---
[CALTECH PHYSICS BUILDING — THURSDAY AFTERNOON]
Sheldon was watching.
Adam felt it the moment he entered the building — the Phase 2 Witness Protocol feeding him Sheldon's behavioral signature at a distance. Not the cataloguing watch of passive observation. Something more focused.
Hypothesis testing mode.
The rhythm was specific. Sheldon was checking Adam's behavior against a prediction model, not building a model from scratch. He had a hypothesis. He was testing whether Adam's actions confirmed or disconfirmed it.
What hypothesis?
Adam walked to his office. Normal pace. Normal route. The secondary notebook was in his jacket pocket, always on person. The main notebook was in his bag. Both contained information that would confirm whatever hypothesis Sheldon was testing, but only if Sheldon could read them.
He passed Sheldon's office. The door was open.
Sheldon was at his desk, apparently working on theoretical physics, but his attention was split. The Witness Protocol registered it clearly: approximately 40% of Sheldon's cognitive resources were allocated to monitoring the hallway.
Adam did not stop. Did not look. Continued walking.
He has a new hypothesis. Not the spy theory — the rhythm is wrong. Something newer.
I need to know what it is.
He could ask Leonard. Leonard had been providing protective information — telling Adam about Sheldon's calibration complaints, about the facilities records request. But asking directly would create its own data point. Leonard would tell Sheldon that Adam had asked, and Sheldon would adjust his model accordingly.
Cannot ask. Can only observe.
He reached his office. Closed the door. Sat at his desk.
The Phase 2 passive range continued feeding him Sheldon's signature through the walls. Still in hypothesis testing mode. Still watching.
More likely than not. Within a month.
---
[APARTMENT 4A — THURSDAY EVENING]
Leonard had invited Adam to dinner. Just the three of them — Leonard, Penny, Adam — at the apartment, while Sheldon was at a physics department lecture.
The meal was takeout Chinese, eaten on the couch with the television on low. The particular informality of people who had stopped standing on ceremony with each other.
"Sheldon asked me something today," Leonard said.
Adam set down his chopsticks. "What?"
"Whether the building's equipment has ever produced anomalies correlated with a specific researcher's access schedule before you arrived."
Penny looked up from her food. She did not say anything.
"What did you tell him?" Adam asked.
"I said no. Which is true — the calibration anomalies started after you arrived. They're documented in Sheldon's log." Leonard took a bite of lo mein. "He filed the response without comment. I don't know what he's doing with it."
"Thank you for telling me."
"Sure."
The conversation moved on. Penny asked Leonard about his latest experiment. Leonard described a problem with his laser array. Adam listened and ate and processed the information Leonard had given him.
Sheldon is correlating the anomalies with my arrival. Not just my presence — my arrival. He is looking for causation, not just correlation.
The spy theory does not explain that question. A spy would have brought suspicious equipment, but suspicious equipment would produce anomalies before the spy arrived — during setup, during testing. Sheldon is asking why the anomalies started when I arrived.
He is getting closer.
Leonard and Penny were watching him. Not obviously — they were too practiced at casual observation for that — but attentively. The way people watched someone they had decided to protect without fully understanding what they were protecting him from.
"I appreciate you telling me," Adam said again.
Leonard nodded. Penny said nothing. The television continued at low volume.
All of them were choosing, in different ways, to protect something they did not fully understand.
---
[ADAM'S APARTMENT — MIDNIGHT]
Adam sat at his desk with the 67% calculation in front of him.
The apartment was quiet. The Phase 2 passive range provided the usual low-level awareness: Sheldon asleep at his scheduled time, Leonard still awake, Penny's television off. The building settling into its nighttime rhythm.
He had managed everything for four months.
He had been careful. Methodical. Patient. He had maintained his cover, encoded the necessary behavioral patterns, built relationships that provided both information and protection. He had deflected Sheldon's spy hypothesis with accurate but misleading explanations. He had given partial truths to the people who asked directly.
The number did not care about any of that.
He wrote in his notebook:
"67. I have approximately 30 days before the most categorically rigorous person I have ever met arrives at a correct framework for what I am."
He read it back.
A correct framework is not the same as a correct answer. Sheldon can understand that the anomalies correlate with my presence without understanding why. He can identify the pattern without identifying the mechanism.
But the framework leads to the mechanism. Given enough time, given enough data, he will arrive at the conclusion.
He added:
"He will probably still get the category wrong. But the framework will be right."
This was the core of the problem. Sheldon's investigation was methodologically sound. His data was accurate. His correlation analysis was correct. The only protection Adam had was the conceptual gap between "something is anomalous about Adam Carter" and "Adam Carter is an esper whose abilities exceed his registered classification."
That gap was closing.
He turned to the back of his notebook. Found a blank page. Wrote the date: 30 days from Thursday.
He did not write what it was a deadline for.
---
The Phase 2 Witness Protocol ran continuously. Sheldon's sleeping signature was precise and predictable. Leonard was working late — the particular rhythm of someone trying to solve a problem before sleep. Penny's apartment was dark and quiet.
Adam lay in bed with the number in his head.
67%.
More likely than not.
Within a month.
He had deflections prepared. The spy theory redirect. The anomalous-physics explanation. The Academy City equipment interference framing. All of them technically accurate. All of them incomplete. All of them designed to buy time rather than solve the problem.
There was no deflection for the correct answer. There was only misdirection.
30 days.
Sheldon will ask questions. I will give answers. The answers will be true enough to pass immediate scrutiny and incomplete enough to maintain the gap.
Until they are not.
Until the gap closes.
Until more likely than not becomes certainty.
He closed his eyes.
The signatures continued. The knowing continued. The number remained.
67%.
He did not sleep.
Get Early Access to New Chapters
Thank you for reading. For those who want to skip the wait, my Patreon is currently 21 chapters ahead of the public sites.
Schedule: 7 new chapters released every 10 days.
Benefit: Gain a significant lead of 7 to 21 chapters depending on your tier.
Support the project and start reading the next arc now: Patreon.com/IsekaiStories