Chương 98: Running with wolves

Six wolves meant a pack, and a pack meant coordination, and coordination in the Pale Corridor meant whoever had designed this place had opinions about his week that he did not share. The undergrowth moved wrong at forty meters. Not wind. Not the small-animal scatter he'd catalogued across six months of ambient forest noise.

This was lateral displacement, low and spreading, the kind of movement that meant something heavy was distributing its weight across multiple approach vectors, and Finn's brain completed the threat model before his feet had finished stopping. He read the split in the first two seconds. Two groups, three and three, coming from the northeast and the northwest simultaneously.

Not staggered, not accidental, but the deliberate splitting pressure of a formation that had been doing this since before Finn had a model for it. The compounding point was twelve meters ahead, where the trail narrowed between two rust-bark clusters. If they let the formation close there, the wolves would have a compression corridor and Finn would have a geometry problem with teeth.

"Marco, left. Dez, hold the eastern approach. Move." They moved. He registered that while he was already pulling the first throwing weapon and sighting the lead wolf on the northeast approach. Five weeks ago that instruction would have produced two seconds of processing, one question, and a counter-suggestion from Marco about the left being a bad call.

The counter-suggestion would have been delivered in the slow, considered register of a man who had decided he knew better about spatial positioning on the basis of being larger than the space, which was not how spatial positioning worked and had required four separate conversations to establish. Now it produced movement. Immediate, directional, correct. Marco was already shifting left and settling into the position Finn had called before Finn had finished the visual check to confirm it.

Not because Marco had done the same spatial read, but because Marco had done enough fights now to know what Finn's calls sounded like when they were right, and this one sounded like that. The trust was in the compliance speed. Five weeks ago the gap between instruction and execution was two to three seconds.

Now it was less than one, and the difference between those numbers was the difference between a formation holding and a formation arriving at the compression point two steps late, which was a difference measured in HP. Finn had four seconds of something adjacent to satisfaction before the wolves arrived and satisfaction became the wrong thing to be running. The first three came through the northeast gap in a staggered line. Lead wolf at speed, two trailing at angles designed to cut off lateral retreat.

Standard pack formation, the kind of thing he'd been watching since the first Crimson Dire-Wolf had introduced itself on Day 3 by killing him, which remained the most efficient learning experience of his life. The formation was good. The lead wolf's execution was good. The timing on the trailing pair suggested coordination that had been built through repetition rather than instinct, which meant this pack had been hunting this corridor for a while and had gotten comfortable with the results.

He threw the first weapon into the lead wolf's shoulder at eighteen meters. Not the kill zone. The disruption zone. The throw was designed to break the lead wolf's stride by half a second, which was enough time to collapse the stagger between the trailing pair. At DataVault, the quarterly planning meetings had followed a similar dynamic: the VP of Engagement would open with a statement designed to set the pace, and whoever followed first would set the formation for the rest of the room.

If you disrupted the opener, the followers bunched. Wolves were, in this narrow respect, organizationally similar to mid-level Sacramento executives, except the wolves had better follow-through and nobody had ever described their approach as "synergistic." The disruption worked. The lead wolf stumbled, the trailing pair closed the gap between them by a body length, and Finn had a window where two wolves were occupying the space meant for one. He threw the second weapon into that window.

From the left, Marco met the western approach. Marco held the line the way Marco held everything: with his body, planted, making the geometry work through sheer mass and the controlled application of violence that had gone from desperate to professional somewhere around week three. The first wolf hit him at the shoulder and Marco absorbed it, redirected the momentum into the rust-bark trunk behind him, and used the rebound to create separation for the counterstrike. That was new.

Three weeks ago Marco would have braced and tanked the hit flat. Now he was using the environment, redirecting force rather than just absorbing it, and the difference was visible in the efficiency of the exchange. Less damage taken for the same result. Marco was quieter than he'd been.

Not good quiet, not the noise discipline of someone who had drilled it, but the functional quiet of a man who had been told enough times, in enough variations of Finn's patient repetition, that sound traveled in thin canopy and that a wolf would redirect toward a locatable noise before it would redirect toward a large obstacle. The quiet had started as performance and become reflex somewhere Finn hadn't been watching closely enough to mark the transition, which was itself a sign that the transition was real.

The things you stopped monitoring were the things that had stopped needing it. Dez held the eastern flank. That was the whole thing, right there, and Finn registered it while cutting the fold-through vector on the northeastern pair. Dez held the flank without needing to be told to hold it a second time. No second-guessing, no micro-adjustment from behind him when the third wolf probed the eastern approach and found Dez already there.

Slightly too far left for ideal coverage, about a meter and a half off the position Finn would have chosen, but there, which was the entire fight Dez had been having with himself for five weeks. The one where he tried to get his body to the right place before his brain finished arguing about whether the right place was really the right place. He was winning that fight. Quietly, without announcing it. Finn pulled the third throwing weapon and ran the field.

Northeast pair: one down from the second throw, one circling with the shoulder injury from the first, moving at maybe seventy percent. Western approach: Marco had the first wolf against the rust-bark and was working the second into a position where it couldn't flank without crossing Dez's sight line. Eastern approach: Dez holding the probe wolf in a containment pattern that wasn't elegant but was consistent, which was better than elegant. This was the part that had changed. Not the mechanics.

Not the throws or the positioning or the callouts. The part that had changed was the cognitive load. Five weeks ago, Finn had been running four simultaneous processes during combat: his own engagement, Marco's positioning, Dez's decision-making, and the ambient threat scan. Four channels of continuous processing, each one demanding enough attention that the total load sat at the edge of what he could sustain without losing resolution on at least one of them. He'd made mistakes in that configuration.

Small ones, recoverable ones, but mistakes born from having more inputs than bandwidth. Now he was running two. His own engagement and the ambient threat scan. Marco's positioning and Dez's decision-making had moved off the active-monitoring board and into the trust column, which was a column he had not built deliberately and could not point to the moment it had been added, but which existed now and was doing load-bearing work. He wasn't watching them because he needed to.

He was aware of them because they were there, the same way you were aware of a wall behind you without turning to check if it was still a wall.

It had, in Finn's professional assessment, taken approximately thirty-seven days to build a team that functioned like a team, which in his previous professional life at DataVault Analytics would have required a six-month onboarding process, a mandatory team-building retreat to some winery in Napa where everyone pretended the trust falls were meaningful, and at least two rounds of the personality assessment that Brenda had been convinced would "unlock collaborative potential" but had primarily unlocked Marcus's discovery that he was an ENTJ, which he had mentioned in every meeting for the following three weeks.

The wolves, it turned out, were a more efficient onboarding mechanism than Napa. The wolves did not require a personality assessment. The wolves were the personality assessment. From somewhere behind him, Yuki was calling meters in the even cadence she'd developed for in-motion triage.

"Marco at sixty, clean. Dez at seventy-two, left shoulder." Then, without changing register: "Northeast lead is favoring the right foreleg. Shoulder wolf is circling wide, blood on the flank, it's compensating. Western pair are both still mobile but the one Marco redirected is slower on the turn." She couldn't see their meters. She didn't need to. She was reading the wolves the same way she read the group: movement quality, posture changes, the visible tells that something underneath the surface had shifted.

She had started doing this around week two, unprompted, calling out enemy condition alongside party status as if the combat space were a single patient with multiple systems and she was running triage on all of them simultaneously. He had never asked her to do it. He had never told her it was useful. It was extremely useful.

The northeast lead's foreleg data changed his throw priority by one position, and the shoulder wolf's compensation pattern told him the rust-bark funnel was going to work because a wolf compensating on its flank would drift toward the wider path, which was the path he'd designed to close behind it. He'd stopped consciously hearing the callouts around week three and started just knowing them, the way you knew the hum of a refrigerator in a house you'd lived in long enough.

Not foreground, not background, just part of the air. She was running that process constantly and had never once asked if it was wanted, which was how he knew it was something she'd decided rather than offered. The distinction mattered. Offered help could be declined. Decided help was infrastructure, and you didn't decline infrastructure any more than you declined the fact that the walls were holding up the ceiling. Two wolves left.

He'd funneled them into the rust-bark backstop from the opening, that part of the design had worked, and the pack's coordination fractured when the formation lost its numbers. A six-wolf pack ran on distributed decision-making. At four it was still a pack.

At two it was two animals with a shared history and no formation to fall back on, and the behavioral shift was visible in real time: the injured wolf from the northeast stopped circling with purpose and started circling with hesitation, and the wolf Dez had been containing broke its probe pattern and pulled back toward its packmate. He pulled the fourth throwing weapon.

The throw landed clean in the shoulder wolf's neck, and it went down with the kind of sudden finality that Finn had learned to associate with clean hits on already-compromised targets. The math was simple: a wolf at seventy percent with an existing injury and a second impact to a load-bearing joint did not have the structural reserves to keep standing, and the math was right, and the wolf was down, and that was five. If you discover this narrative on Amazon, be aware that it has been stolen.

Please report the violation. The sixth looked at Finn. Then it looked at Marco, who was standing over the two it had arrived with, breathing hard but upright and uninjured. Then it looked at Dez, who was holding a position it had already failed to break through once. The wolf conducted what Finn could only describe as a real-time risk assessment, arrived at the correct conclusion, and retreated toward the tree line with the intelligent urgency of an animal that had revised its evaluation of the situation.

He let it go. A lone wolf in retreat was not a threat worth the energy. He was close enough to Haven's Reach that he had started thinking about energy in those terms: expenditure against remaining distance, a budget that included not just meters but fights, sleep quality, the cumulative fatigue load that didn't show in the HP bar but showed in reaction time and decision quality. The retreating wolf vanished into the undergrowth.

Finn briefly considered the possibility that it was going to file a report with whatever passed for wolf management in the Crimson Wilds and return with reinforcements. Then he considered the possibility that he had just anthropomorphized a wolf's retreat into a workplace analogy, which was either a sign that he was spending too much time thinking about organizational dynamics or a sign that organizational dynamics were, at a sufficiently abstract level, universal. He chose not to examine which.

He stood in the aftermath. The rust-bark clearing was quiet in the way clearings were quiet after violence: not peaceful, but emptied, the ambient sound stripped out by four minutes of activity and not yet refilled. Marco was checking his weapon against the nearest trunk, testing the edge. Dez was standing at the eastern position, looking at the spot where the probe wolf had been, with the expression of someone replaying the last ninety seconds and arriving at a conclusion about them.

Yuki was already moving toward Marco, not because Marco needed attention but because post-combat assessment was part of the sequence she ran and she ran it regardless of whether the data confirmed what she already knew. They looked like a group that had done this before. Because they had. Finn read the notification.

【 AETHERFORGE NEXUS 】 LEVEL UP — Level 14 XP Cleared: 115, 000 | 5 Free Attribute Points Available HP: 570 → 615 XP to Level 15: 175, 000 Level 14 fired at the fifth confirmed throwing weapon strike, mid-engagement, surrounded by three people. The Watchers have records going back to Level 2. This is the first level that cannot be logged under the solo column. They are not assigning weight to this. They are simply noting the column has changed. He read it twice.

Fourteen was a number that required two reads to fully land, not because the number was surprising but because it represented something about the distance traveled that landed differently when you were standing in the aftermath of a six-wolf engagement that had gone functionally rather than catastrophically.

Which was a category of result he had been building toward for five weeks and had arrived at with, he noted, the same anticlimactic competence that had characterized most of his genuine professional achievements.

The Q3 churn model at DataVault had worked the same way: three months of methodology, and when the numbers finally came in clean, the dominant feeling was not triumph but the quiet recognition that of course it worked, because he'd built it to work, and the building was the hard part and the working was just the building's receipt. Although the churn model had not involved wolves.

The churn model's primary threat had been Marcus accidentally sorting the data by the wrong column, which had happened once and had produced thirty minutes of confusion and zero casualties, which put it meaningfully below the current engagement on the stakes axis. He thought about the allocation in transit while the camp sequence started around him. Five points.

PER and STR as co-primaries: the corridor ahead needed eyes before it needed anything else, and he was carrying four people's worth of operational weight now, which STR handled with diminishing elegance at its current floor. Three points there. Not complicated. WIL got the last one. He looked at that decision for longer than the decision itself had taken. He could argue INT. The pattern models, the cross-reference work, the schedule hypothesis he was still building with Yuki.

He knew the counterargument and it wasn't weak. But WIL was what held a decision-making apparatus together when the model had gaps in it, when the inputs were incomplete, when you were transiting toward something you couldn't fully project. He had a model with a gap in it.

A transit ending at a place he'd never stood, with a channel name he'd been reading for months, and he knew that channel name by its posting rhythm and timestamp habits and the way approach conditions were written in the early morning like someone who was already awake long before the post. And he did not have a model for what he was going to do about any of that when he arrived. WIL was the stat for that. The stat for holding steady when the data ran out and the decision hadn't. He told them at dinner.

Not forty minutes later, not after the debrief had been filed and the appropriate distance created. Before the food was fully heated, at the point in the evening where the fire was at its best and everyone was running the tired-but-functional that came after a fight that had cost less than expected. He told them because the number was good and they'd been in the fight that produced it and it felt like information that belonged in this room rather than in a column.

"Level fourteen," he said. There was a beat. Small, warm, unmanaged. Not the silence of people who didn't know what to say. The silence of people absorbing good news at its natural speed.

"HP jump?" Marco asked. Because Marco's first question was always load-bearing practicality, and Finn had stopped finding this predictable and started finding it reliable, which was a different word that meant a different thing.

"Forty-five." Marco nodded with the satisfaction of a man filing a useful number.

"What'd you do with the allocation?" Dez asked. And then it was a conversation. An actual one, the kind that moved on its own without requiring management. He gave them the breakdown: PER and STR as primaries, WIL for the remaining one, and his reasoning for each, which he delivered the way he delivered briefings but looser, less structured, and nobody had to work to follow it. Yuki asked about the XP threshold to Level 15. He told her.

She filed it somewhere behind her eyes, in the same place she filed world boss schedules and treatment outcomes and whatever parallel dataset she was building about the group that she would share precisely when she had finished deciding what it meant. Marco asked whether the HP gain was proportional to VIT or a flat level bonus, which was a better systems-mechanics question than Marco usually asked, and Finn answered it and watched Marco incorporate the answer with the expression of someone building a longer model.

Marco building models was a relatively recent development. Three months ago Marco had treated the stat system the way Finn's old roommate Travis had treated the thermostat: something you hit when you were cold and ignored the rest of the time. Now Marco was asking proportionality questions. Growth was frequently invisible to the person experiencing it. Then Dez tilted his head.

"Why not INT?"

"Corridor terrain," Finn said.

"I need to be reading it before it decides anything about me. PER's doing more work right now."

"Sure." Dez turned his improvised fork, bark-stripped stick, fork by consensus and repetition, between his fingers.

"But you've been running pattern models, channel cross-references, the schedule work from yesterday. That's INT infrastructure. You're deferring points off a stat that's actively in use."

"Deferring, not skipping."

"That's a methodology choice that could also be described as..."

"Leaving them on the table, yeah." Finn looked at him.

"You're not wrong about INT being underweighted. PER's what the next six days need. After that the calculus changes." Dez considered this for a moment.

"Fair," he said, and went back to his food, and that was the end of it. Finn sat with the aftermath of that exchange. He'd had a person push back on his allocation reasoning, explained the reasoning, received fair as a conclusion, and the whole thing had taken ninety seconds and required no management whatsoever. No defensiveness from either side. No subtext requiring navigation. No performance review framework where fair meant we're going to revisit this in Q4 with a development plan attached.

Just: push, explanation, resolution. He was choosing not to examine what it said about the preceding six months that this qualified as one of the better conversations he'd had in months. The examination would produce a number, and the number would be too honest, and he was running a limited supply of honesty tonight and had already allocated most of it to WIL. 【 AETHERFORGE NEXUS 】 CHA +0.

1 — Voluntary Status Disclosure Trigger: Level-up shared with group before tactical debrief, allocation reasoning subjected to peer review without deflection. Player 4782 announced a level-up at dinner. Voluntarily. Without being asked. Without converting the announcement into a briefing, a data point, or a transition to the next topic. The Watchers are checking the file. The file confirms this is the first time. They are adding a sub-column. The sub-column does not have a name yet.

They are confident he will not name his either. He closed the bark-sheet. Level 15 was 175, 000 XP and he had no model for what Haven's Reach generated, which was a real gap in the projection column and he was filing it as such and then leaving it there. Some data arrived in its own time. He'd learned that somewhere around Day 40, in a different column, and the lesson had apparently carried. Haven's Reach was six days north.

He ran the channel one more time before camp routine closed for the night, the way he always did, scrolling through the Ring Channel's accumulated traffic for the day. Most of it was noise: arrival reports, trade inquiries, a brief argument about stall pricing that had the energy of two people who were going to have this argument every day until one of them died or left. Normal. The mundane texture of a functional settlement filtering through a channel built for survival.

Then a post from an account he'd been tracking for eleven days. Not LiraT. Not the helpful-type accounts or the neutral-report accounts. This one posted in a rhythm that was harder to read, timestamps clustered in a way that suggested movement rather than settlement, and the content was route-adjacent without being route data. The kind of posts you'd write if you wanted to know where people were going without telling anyone where you were. He'd flagged the account eleven days ago.

Cross-referenced the posting pattern against the PK behavioral taxonomy he'd been building since the first month. The match wasn't conclusive. An eighty-two percent confidence interval that the account was associated with the northern corridor patrol, and eighteen percent that it was just a cautious player with bad posting habits, and he could not close that gap from channel data alone.

The corridor ahead was six days of trail through terrain that narrowed in two places: the Pale Corridor's western bottleneck, which he had route data for, and a secondary approach channel that the channel maps marked as lower-traffic but that his PK movement analysis suggested was monitored. Not continuously. Not consistently. But monitored in a pattern that had a schedule, and the schedule had a gap at the 0900 transit window, and he had been building the case for that gap for two weeks. He looked at the fire.

Marco was settling. Dez was doing the thing where he stared at the coals and worked something out. Yuki was writing on her own bark-sheet, the one she didn't share, the one that contained whatever parallel dataset she was assembling about things she hadn't finished deciding were true. Four people. His people, by a definition he was still not putting in writing.

Transiting through a corridor that had at least one confirmed PK patrol and possibly a second one he hadn't mapped yet, because the coverage model had a gap in it, and gaps in coverage models were the thing that kept him awake at night, and he was going to keep the gap to himself until he'd closed it or confirmed it, because sharing an incomplete threat assessment would produce exactly the kind of anxiety that made people move wrong in exactly the kind of terrain where moving wrong had consequences measured in Echoes.

Six days. The route was solid. The group was solid. The corridor was the problem, and the corridor was the kind of problem that didn't announce itself until you were inside it, and the only preparation available was the preparation he'd already done. Finn checked the channel one more time.

The flagged account had posted again: northern approach, route-adjacent, the same not-quite-useful data that could be reconnaissance or could be caution and the difference mattered in ways that would become clear approximately four days from now, when the trail narrowed and the rust-bark thinned and the corridor became the kind of place where visibility was a resource and he intended to have more of it than anyone else present. He closed the thread.

Updated the bark-sheet: approach window confirmed, PK confidence 82%, gap at 0900 holding. The fire popped. Marco's breathing slowed. Dez's coals-staring resolved into whatever conclusion Dez had been building, and he lay down without sharing it, which was Dez's way of saying the conclusion was his and it would arrive when it was ready. Six days north. A corridor. A gap in the model. He was going to close it.