Chương 79: Raptors are not small.

The raptor changed direction at the worst possible moment, which in Finn's experience was the only moment raptors ever changed direction. He'd been routing them around it for the better part of an hour. The track signs were fresh enough that the math was simple, the geometry was simple, just a question of staying upwind and adding the distance. He'd explained this to Dez at the last break.

Dez had said "yeah, okay, the load on the other side of the path is the raptor problem, we're moving weight off it," which was not how Finn would have framed it and was completely accurate. The group had adjusted. Simple. Except the raptor had apparently not received the itinerary. At DataVault, they'd had a project management tool that sent automated route notifications whenever a timeline changed.

Estimated reading rate among the Sacramento office: eleven percent, and that was generous because Finn had counted himself in the denominator. He had once watched Marcus walk into a client call that had been moved to a different conference room, rescheduled by forty-five minutes, and reassigned to a different account lead, all of which had been communicated via three separate notifications Marcus had marked as read without reading. The raptor had at least not pretended to read the notification.

Finn respected the honesty, if not the resulting trajectory. He heard it before he saw it. The sharp crack of something heavy moving fast through undergrowth that didn't care about being heard because it had already committed to the approach. He had maybe four seconds between that sound and the moment it broke the treeline, and he spent the first one doing a rapid calculation that arrived at: the routing assumption was wrong, update the routing assumption, move.

"North," he said, not loudly. The group moved. This was something. Three weeks ago he would have had to say it twice. The problem with north was the terrain. The ground ahead sloped upward for maybe eighty meters before it hit a shelf where years of water runoff had carved the soil away from the underlying rock, leaving a raw lip of exposed stone and compacted earth that dropped off at an angle he hadn't surveyed yet.

From this distance it looked like the forest floor simply ended, a clean edge where the trees thinned and the root systems gave way to open air and whatever was below it. He'd mapped it two days ago on the approach, had flagged it as a variable he needed to account for, had not fully accounted for it yet because the accounting was scheduled for tomorrow's route planning when they'd have better sightlines from the eastern ridge.

Moving fast toward a drop he couldn't fully model was not the outcome he would have chosen. The raptor was faster than his preferred outcome. He was running the angles, the predator's likely intercept line, the shelf geometry, whether there was a lateral option that didn't put them closer to the treeline, and he had a route, a workable route, not optimal but survivable, when Dez said "left, through the outcropping" and veered without waiting for confirmation. Finn's first instinct was to call him back.

His second instinct, arriving approximately half a second later, was to actually look at what Dez was pointing at. The outcropping was a limestone formation, wind-cut, the kind of thing that looked like a jumble of pale rock from a distance until you read the geometry of it and realized the gaps between the slabs weren't random. They were channels, carved by centuries of drainage flowing off the shelf above and finding the path of least resistance down through the stone.

The result was a narrow passage, maybe a meter and a half wide at the tightest point where two angled slabs leaned into each other like a doorframe built by erosion, opening on the far side into a shallow depression where the runoff had pooled and carved out a basin over decades. Passable for people. Not passable for something with a raptor's shoulder width and the momentum it had built coming down the slope. Dez was already through the narrow point. Yuki went next.

Marco, who was wide but not raptor-wide, turned sideways and made it with a breath to spare. Finn went last, catching his pack on the limestone edge, losing two seconds, coming out the other side at a run. The raptor hit the corridor entrance at full speed and didn't stop, exactly. It checked. He heard the scrabble of claws on rock, the frustrated percussion of its body trying to force the stone to accommodate a shoulder span the stone had no interest in accommodating.

He kept moving another sixty meters before he let himself slow, and then he stopped, and the rest of the group stopped, and the sound behind them was the raptor pacing the entrance to something it wasn't going to fit through. For a moment nobody said anything. 【Evasion — Route Optimization】 Dez's routing decision: optimal (confirmed via replay) Finn's routing: would have been 2nd-best outcome Exploration XP: +90 | AGI: +0. 2 (Dez) | CHA: +0. 1 (Finn)"North." Three people ran north.

This continues to happen despite the available evidence regarding route-planning infallibility. The engineering student's call was acknowledged correctly, without qualification, and without the Watchers having to watch anyone die in the process, which they suppose counts as progress. You might be reading a pirated copy. Look for the official release to support the author. He ran it back while they caught their breath.

His route, the lateral option he'd been calculating, would have kept them ahead of the raptor and gotten them clear, probably, with a margin he would describe as acceptable and Yuki would describe as uncomfortable and Marco would describe as a good story. Dez's route had ended it in thirty seconds by making the predator's momentum work against it. Finn had spent four months reading this forest as a survival problem. Threat density, patrol routes, coverage arcs, ambient noise profiles.

Everything filtered through the question of what is trying to kill me and from which direction. It was a good lens. It had kept him alive for a hundred and thirty-eight days, which was a stronger performance review than anything DataVault had ever given him, and the forest hadn't even needed a rubric. Dez had looked at the same rock formation and seen a load-bearing structure with a width constraint. Not terrain. Not cover.

A physical system with properties that could be exploited if you understood how forces moved through space. The gap between the limestone slabs wasn't safety. It was an engineering argument, and the raptor had lost the argument at speed. The output had been the same. The methodology had not. And Finn, who had been the methodology person in this group for three weeks, was now sitting in a drainage basin looking at the back of an engineering student's head and recalibrating what "reading terrain" actually meant.

It was the kind of moment that, at DataVault, would have resulted in someone saying "cross-functional synergy" in a tone that made it sound like they'd invented the concept, and Finn would have wanted to leave the room. Here, in a drainage basin, sixty meters from a frustrated raptor, the concept was doing actual work. He looked at Dez, who was bent forward with his hands on his knees, getting his breath back.

"That was optimal," Finn said. Dez looked up.

"Yeah?"

"Your route. Faster and lower risk than mine." He paused.

"I didn't see the corridor." Dez straightened. Something moved across his face, surprise mostly, with something underneath it that was going to take a few minutes to resolve into whatever it actually was.

"I mean, it was just the geometry. The gap was there, the thing was wide, it's basically a load problem..."

"I know what it was," Finn said.

"It was still the better call." Dez opened his mouth and then closed it again. He looked at Marco, who was giving him the expression of someone watching an unexpected thing happen and deciding not to comment on it, and then at Yuki, who had already moved on and was checking the sightlines on the far side of the depression.

"Okay," Dez said, after a moment.

"Okay. Cool." They moved on. Finn fell into pace and thought about the corridor and the way Dez had seen it not as terrain but as a constraint problem. This thing has a width, that thing has a width, the difference between them is a solution. Four months of reading this forest and he had been reading it as a survival environment. Dez had looked at the same rock formation and read it as a structure. The output had been the same. The methodology had not. Camp went up without incident.

The terrain was good, the threat density was manageable, and Marco's injury from two days ago had dropped below the threshold that required Yuki's active monitoring, which was its own small win he logged without making a thing of it. Dinner was the smoked boar strips, which were holding, and which continued to represent approximately the largest single improvement to his quality of life since he'd started moving north. He still thought about the meat he'd left in the field for four months.

He was going to be thinking about it for a while. Dez was quiet over dinner. The processing quiet, not the worried kind. The kind he got when something had gone into the part of his brain that didn't talk until it had finished working through the problem. Finn had learned to read the difference. He let it be. After, when Marco and Yuki had drifted into their pre-sleep routines, Dez said: "Do you always know when your route isn't the best one?" Finn considered this.

"No. Today I didn't."

"But you ran it back after."

"Yes." Dez was quiet for another moment.

"Does that... does it bother you? When someone else gets there first?" He thought about this honestly.

"Not really. That's not the part that bothers me."

"What's the part that bothers you?"

"That I had four months to learn to read terrain as a structural problem and I didn't." Dez laughed, short, not unkind.

"Okay so you're annoyed at yourself."

"I'm noting a gap in my methodology."

"Right, yeah, that," Dez said.

"That's the same thing. You know that's the same thing." Finn didn't say anything. Dez made a sound that was approximately exactly, and let it go, and they sat with the fire for a while before the watch rotation started. He pulled out the bark-sheet at the end of his watch. Wrote: Dez. Spatial reasoning, terrain application via structural geometry, strong. Methodology distinct from mine. Outputs comparable. Asset. The entry was accurate. He looked at it for a moment. Professional. Clean.

The kind of assessment he'd have written at DataVault, the kind that went into a quarterly review and said everything that was measurable and nothing that required him to admit he'd been measuring wrong. He could close the bark-sheet here and the record would be complete and the record would be insufficient and he knew it.

He wrote one more word after asset, and the word was harder to write than the rest of the entry combined, not because it was complicated but because it was honest in a direction the bark-sheet hadn't previously been asked to accommodate. Underestimated. He closed the bark-sheet and wondered how many other things he was underestimating.