Chương 96: a sense of Urgency.

The notification arrived at 0847, during the second water break of the morning, between the section of trail where the rust-bark started thinning and the section where it thinned into something he'd tentatively named the Pale Corridor. Mostly to give it a column on the bark-sheet, not because the name was right.

【 AETHERFORGE NEXUS 】 World Boss Proximity — Event 5 Blood Mire Stalker | Apex Predator Class Current Trajectory: SSW — Sector 12-F, southern basin approach Projected Impact Zone: Sectors 9-E through 11-G (eastern lowlands — two sectors south of previous position) Distance from Current Position: 41. 2 km Route Conflict Probability: 0% The Watchers note this is the fifth world boss proximity notification Player 4782 has received. They note the impact probability is zero.

They also note he is reading it a second time. They understand the impulse, but they want it on the record that at zero percent, the second read is not doing any additional mathematical work. He read it a third time. Zero was such a good number. Zero was, in fact, a number he had spent an embarrassingly large fraction of the last 172 days wishing was more available as an option.

He had impact probabilities for most things in his life now: thorn-serpent corridor crossings, raptor sign in the undergrowth, the standing probability of Dez asking a question in the next four minutes, which hovered around 94% at any given moment. Zero was not a number that came up often, and when it did he liked to read it twice, slowly, the way you re-read a positive lab result because you don't trust your eyes.

"Southern basin," he said, pulling out the bark-sheet. The map he'd been building for the last two weeks had enough coverage to put a finger on the relevant sector, the wetland approach corridor he'd marked AVOID in bold for reasons that were still very much accurate.

"Well south of us. Impact probability is zero for anything on our current path." He said it the way he'd learned to say it over five weeks of group briefings: direct, no hedging, no attached footnote about how he'd reached the conclusion. Five weeks ago the information would have been processed internally and routed through the bark-sheet alone. Now it went to the group in the same breath. The briefing habit was built. The trust had done that, incrementally, one shared data point at a time.

He had not planned to build it. He had built it anyway by repeatedly being correct. Marco nodded with the efficiency of someone who had received good information and was filing it without needing to comment on it.

"How far south, exactly?" Dez asked. Finn looked at Dez with the expression he was still developing for the question is technically valid but the answer doesn't change the next hour of our lives.

"Far enough that we're not on the map."

"Right, but..."

"Dez." Dez absorbed this with the grace of someone who had been told Dez at varying levels of urgency for the last five weeks and had learned to read the register. This version was category two: complete and sufficient, you don't need the coordinates. He let it go. He was writing world boss. Southern basin, no route adj. required in the operational column, deliberately noting that phrase because he liked to document rare events, when Yuki said, "I've been tracking it." Not the briefing.

She'd stepped slightly sideways of the conversation in the way she did when she was already two beats ahead and waiting for everyone else to catch up.

"All of them," she said.

"The last ten days." He stopped writing.

"The channel data," she said, and her voice had the quality of a person delivering a conclusion that had been fully processed before it left her mouth. No preamble, no warm-up, just the output.

"I've been filtering it for world boss position reports. Two events in the southern sectors, confirmed. One near the eastern ridge, less confirmed, but the movement reports point the same direction consistently enough that I'd put it in the probable column." She paused, which was Yuki's version of emphasis.

"All three are approximately synchronized. By roughly the same interval." Finn looked at her. Synchronized. He'd written that word in his own bark-sheet notes four days ago and then erased it because the sample size was too thin to go on record about. Three data points arranged suggestively were not a schedule; they were three data points arranged suggestively, and the gap between those two classifications was where bad routing decisions came from. He had been sitting on the hypothesis and not committing to it.

She had ten days of work and five data points and she had already found the same word.

"You think it's a schedule," he said.

"I think it's more than coincidence," she said.

"Which I think is the same thing." It was the same thing. And he was noticing, in the careful way he noticed things when they required him to update a prior belief about himself, that she had been running this operation in parallel for ten days without mentioning it because she wasn't sure yet. She'd built the data set independently. Waited for the confidence ceiling to clear. Brought it to the operational layer only when the model was solid enough to be worth the air time.

"What's the interval?" he said.

"Eleven to thirteen days between events. Consistent across all five data points." He did the math before he finished breathing. The last event was today. The zero-percent one, the comfortable one, the one sitting safely in the southern basin where it couldn't touch them. If the schedule held, the next event would fire in eleven to thirteen days. And the trajectories Yuki had tracked were not all going the same direction. They were rotating. Southern basin, eastern ridge, and a probable in the western sectors.

The system was cycling through impact zones the way a spotlight sweeps a stage. He pulled the bark-sheet back out and looked at the map. Their corridor ran north through Sectors 8-D and 7-C. The eastern ridge event had clipped the edge of 8-E. One sector over. If you come across this story on Amazon, be aware that it has been stolen from Royal Road. Please report it.

"Yuki," he said.

"The eastern ridge event. How far did the impact zone extend west?"

"Edge of Sector 8-E. Possibly into 8-D, but the channel reports were inconsistent on the western boundary." 8-D. The sector they were currently walking through. He stared at the bark-sheet for several seconds. At DataVault, there had been a quarterly exercise where the risk assessment team would present scenarios with two bad options and ask the analysts to identify the "preferred outcome." Finn had always found the framing dishonest. The preferred outcome was a third option that didn't exist.

The exercise was really asking you to pick which variety of bad you could tolerate, and then dress it up in enough confidence intervals that the VP would nod instead of asking follow-up questions. This was that exercise, except the VP was a Blood Mire Stalker and the confidence intervals were measured in whether or not they got trampled by an Apex Predator.

"So if the schedule holds," he said, keeping his voice in the briefing register because the alternative was the register where he started making sounds that weren't words, "the next event fires in eleven to thirteen days. And if the rotation pattern holds, the impact zone could clip our corridor."

"Could," Yuki said.

"The western boundary data is soft."

"And if we turn back south to avoid it?" He already knew the answer. He'd looked at the map.

"Southern basin is the current event's impact zone," Yuki said.

"It'll be active for days. We'd be routing into a confirmed threat to avoid a projected one."

"And if we stay put and wait it out?"

"Then we're stationary in a corridor that might be inside the next impact zone, and we've lost eight to thirteen days of travel toward Haven's Reach." Finn nodded slowly. Three options. Forward into a tightening window. Backward into confirmed world boss territory. Stationary in a corridor that might become the next impact zone. This was what the risk assessment team at DataVault would have called a "constrained decision matrix."

What Finn would have called it, privately, in the break room with Marcus, was a situation where all the doors lead to the same hallway and the hallway is on fire.

"We go forward," he said.

"And we don't miss the window."

"How tight is the window?" Dez asked.

"To clear this corridor? Eight days to be safe. Twelve if we're willing to bet on the interval being thirteen instead of eleven." He looked at Dez.

"I am not willing to bet on the interval being thirteen instead of eleven."

"Right."

"And that's just this rotation. Haven's Reach is still close to three weeks out. Every eleven to thirteen days, the schedule cycles again and we need to not be standing in whatever sector it points at next." He let that settle.

"So the short answer is: we move faster, we stay ahead of the rotation, and we do that for the entire remaining route."

"And the long answer?"

"The long answer is the same answer with more math and the same conclusion. We move faster." Marco adjusted his pack straps, which was his version of voting. And then, because the tactical conversation was finished but the other one wasn't, Finn looked at Yuki.

"That is exactly what I would do," he said.

"Run it in parallel until you trust it before you bring it out. Don't hand over a hypothesis, hand over an argument." He heard himself still talking and had the experience, not for the first time with this group, of arriving at the end of a sentence and realizing the beginning was about thirty percent too self-revealing.

"I'm not... I mean. That's not a criticism. That's the correct methodology. I'm saying it as... I'm familiar with it. From having done it. To other people. Repeatedly." He stopped talking. This was the correct move and it was approximately twelve words late. Yuki looked at him with the focused, entirely patient attention she used when she was waiting for a person to locate their point.

"That's exactly what I would do," he said. Simpler this time.

"I know," she said. Not I know, I've been watching you. Not I know, I was more or less copying the approach. Just: I know. Two syllables, carrying information in at least three directions simultaneously, delivered in the exact register she used for things she had thought about before she said them. He parsed all three directions.

He was not entirely sure what to do with the fact that he'd parsed all three directions, but the parsing had happened, and it was accurate, and she had looked at him the same way throughout, the same focused patience, waiting for the rest of the world to catch up, which meant she had watched him parse it and had not been surprised. He was going to have to file this somewhere. He didn't have the header yet. 【 AETHERFORGE NEXUS 】 INT +0. 1 — Cross-Referenced Intelligence Source: Independent parallel analysis (Y. Tanaka) / Shared Intelligence Integration Data: 10 days | 3 confirmed events | 2 probable | schedule hypothesis: active, unfalsified Confidence delta: +3% (second independent analyst, non-correlated methodology) Player 4782 said "this is exactly what I would do" to a person who has been doing exactly what he would do for ten days without his knowledge. He then explained his methodology to her. Yuki Tanaka's response was two words. He understood all of them. Filing him under: sometimes his own slowest data point.

He held the notification for a moment longer than usual. Then he dismissed it and looked at Yuki.

"Walk me through the eastern ridge data," he said. She did. It was clean work. She'd sourced the position reports from four separate accounts with posting histories that cross-validated without coordinating, which was exactly the methodology he would have used, and he was going to stop pointing that out to people internally and externally. Dez asked two questions, both good, one of which identified a gap in the geographic overlay that Yuki had already flagged as a known uncertainty.

Marco stood to the side with his pack on, holding space for the technical conversation without fidgeting, which was his contribution and it was the right one. Finn added her data to the shared intelligence section. Not the operational column. The shared intelligence section, which had previously contained only his own analysis, because that was all there had been.

【 AETHERFORGE NEXUS 】 Shared Intelligence — Updated World boss schedule hypothesis: two independent analysts | non-correlated entry points | converging model Schedule window conflict: active concern — recommend Sector 8-D corridor clearance within 8 days Haven's Reach timeline: ~18–20 days | viable | current path | confidence +3% The shared intelligence section now contains work from more than one person. Player 4782 added it without pause, which the Watchers found notable enough to document.

The section has been his for six weeks. He did not appear to notice he was doing something different. They are not going to be the ones to tell him. The Haven's Reach timeline held. He had it at eighteen to twenty days, terrain-adjusted, which was close enough to the original estimate that the methodology was still sound. But the schedule hypothesis had changed the shape of the problem. It was no longer a straight line north with a destination at the end.

It was a straight line north through a corridor that became a potential impact zone every eleven to thirteen days, which meant the route was viable only if they stayed ahead of the rotation. Clear this sector in eight days. Clear the next vulnerable stretch before the cycle after that. The margin was not in the total distance. It was in the segments. Not comfortably inside the window. Not in the way he'd prefer with margin on both sides and a contingency route that didn't involve the words "Apex Predator."

Inside it by enough, if they maintained pace.

"We're still on track," he said.

"Tighter than I'd like, but on track." Marco made a sound that was affirmative.

"What happens if we miss the window?" Dez asked.

"We don't miss the window," Finn said.

"Right, but if..."

"Dez."

"Yeah." They started walking. He ran the route model in the back of his head while the trail unfolded north, the gait adjustment nearly invisible now on the left side, day four, the shoes doing their work with the quiet competence of something built by a person who understood the problem, and the updated confidence column organizing itself. Three percentage points up on the routing, which wasn't large. The schedule model confidence was up by significantly more.

Two analysts independently converging on a hypothesis from different entry points was a different weight of evidence than one analyst sitting alone on a conjecture and waiting for the world to confirm it. He did not write second independent analyst in the bark-sheet. He wrote cross-referenced: Y.

Tanaka, which was accurate and professional and did not contain the precise thing he was actually thinking, which was that he had been out here 172 days and had not expected to find a person who treated a survival situation the way he treated a survival situation. As a data problem, worked methodically, held in private until the methodology cleared. He had not expected that. He had found it anyway.

He thought about it for the next two hours, which he noted, and filed in the section that still didn't have a header, and kept walking north. Faster now. The math required it.