Chương 21: Padding the performance review.

He laid the bones out on the flat root beside the fire before dawn, which was either the sign of a man who had his priorities in order or the sign of a man who had been alone for twenty-three days and had developed a relationship with his material inventory that a therapist would find concerning. Six wolf femurs.

Two boar fragments — he'd found them on the retreat path from the rust-bark creek, shed or knocked loose from something the boar had been working on before he arrived, dense and curved and heavier than the wolf bones by a noticeable margin.

A pile of smaller material: rabbit tibias, the unclassified small prey he'd been accumulating without a clear application for, various fragments from processing work that he'd set aside rather than discard because waste was a resource allocation problem rather than an inevitability. He looked at the bones the way he looked at restructured data — not at the individual values but at the whole, trying to let the pattern surface rather than imposing one on it. The answer was in there. He needed to see it.

He started with the knife. The existing bone knife had been working — he wasn't dismissing it, it had earned its place in the inventory through consistent performance — but it had limitations he'd been noting for a week. The edge held for six hours before needing maintenance, which was acceptable.

The handle geometry was functional but had a tendency to shift during sustained precision cuts when his grip fatigued, which was not acceptable because precision cuts were exactly the situation where a shifting grip caused the most damage. And the tip profile was optimized for processing work rather than penetration, which was a constraint he'd been willing to accept until the boar encounter had introduced the concept of things that require actual penetration force at a distance.

He reknapped the blade first — the hammerstone approach, controlled percussion at a shallow angle, removing material in small deliberate flakes rather than the large fractures of early-arc knapping. The edge that resulted was finer, the bevel more consistent, the tip geometry corrected toward the penetration profile he wanted. 【Crafting Progress — Bone Tools】 Blade reknapping: Hammerstone technique applied.

Edge consistency: Significantly improved over base knife Tip profile: Revised toward penetration geometry INT +0. 1 (precision material work — sustained attention under fine-motor load) The Watchers observe your hands. They find the precision unexpected. The crossguard was the innovation. He'd been thinking about it since the first time the original knife's handle had shifted during a processing session and he'd taken a small cut on his index finger as the price of inadequate design.

A crossguard — a short perpendicular piece of bone set into the handle just above where the blade met the grip — prevented the hand from sliding forward during thrust or precision cuts. It was the obvious solution in the way that obvious solutions always were once you'd arrived at them: inevitable in retrospect, invisible until it wasn't. He had spent four days arriving at an idea that a medieval craftsman would have considered basic.

He chose to measure this against Day 1 Finn rather than against a medieval craftsman, because the comparison that produced useful data was the one that started from where he'd actually been. The boar fragment provided the material. The heavy curve of the fragment worked in his favor — he split it along the grain with careful hatchet work, selected the segment with the right geometry, and fitted it into a notch he'd cut into the ironwood handle.

Forty minutes of incremental work that produced four failures before the fifth iteration seated correctly. He lashed it in place. Resin-sealed. He held the completed knife and tested the grip through a dozen pressure variations — overhead, underhand, lateral, the extended precision grip he used for plant processing. The crossguard did exactly what it was supposed to do. The handle stayed fixed regardless of the pressure distribution.

【Crafting Success】 Bone Knife (Improved, Common) Damage: 14 (precision strike bonus +22%) Skinning/processing efficiency: +18% Durability: 40/40 Edge retention: 6 hours (improved from 4) Handle stability: Significantly improved (crossguard design) Crafting Tree: Bone Tools branch — ACTIVE. The Watchers revise upward. Again. They have now revised upward three times in five days. Water source: revised upward. Boar encounter outcome: revised upward. Knife quality: revised upward. They would like the revisions to stop.

The Nexus has no mechanism for requesting this. A new crafting tree branch is now active. This means the improvements are not random. They are directional. Players who improve directionally in the first plateau phase tend to keep improving. Past a certain point, directional improvement stops being a playstyle. It becomes a trajectory. The Watchers are tracking the trajectory. This is becoming a trend that concerns them. They are not yet specifying why. The knife is excellent. This is also part of the problem.

The Watchers had been bored before. This read differently — the same way a performance review that said "we have concerns" read differently than one that said "could try harder." Bored was one thing. Concerned was a variable worth tracking.

"Intellect gains from precision crafting," he said, fitting the knife back into the waistband position and testing the draw.

"I got smarter by making things carefully. My previous career also rewarded careful attention to detail and did not provide Intellect stat improvements. I'm noting that for the performance review that will never happen." The spear was the afternoon's project. He'd been sketching the concept on the supply log's reverse side for two days — the geometry of the tip, the lashing configuration, the shaft material requirements.

The ironwood sapling selection was the first constraint: a straight run of approximately two meters, consistent diameter, dried rather than green. He'd been drying three candidates in the elevated storage for a week, anticipating exactly this application. The straightest one came down from the storage crossbeam and passed his flex-test — minimal deviation along the full length, the grain running clean. Unauthorized reproduction: this story has been taken without approval. Report sightings.

The bone splinter for the tip came from the largest wolf femur. He'd been looking at it for days: a natural fracture line that would produce a long slender piece if he worked it correctly, dense at the base where it would meet the lashing and tapering to the penetration tip he'd developed on the knife's reknap session.

He worked it with the hammerstone for ninety minutes, slower than any previous bone work, because the geometry was less forgiving and a fracture along the wrong axis meant starting over with a smaller bone and a shorter tip. The fracture went correctly on the third controlled strike sequence. 【Crafting Progress — Bone Tools】 Bone splinter shaping: 90 minutes, hammerstone technique Tip geometry: Penetration profile (narrow, reinforced base) Result: Viable — suitable for spear-tip application INT +0.

1 (sustained precision work — second session) The Watchers note the patience this required. They had not previously categorized you as patient. Lashing the tip to the shaft took another hour. The tip seated into a prepared split in the shaft's end, vine-lashed in overlapping figure-eight passes that locked the base from lateral movement, then resin-sealed across the entire lashing zone.

The resin needed two hours to cure before the joint reached full integrity, which meant the afternoon was officially a waiting period, which meant he tested it early anyway because two hours was a long time to look at a completed weapon without throwing it. The tip held on the first three throws at five meters.

Clean penetration into the dead log he'd set up against the iron-thorn cluster — the bone punching through the outer bark and lodging in the softer interior wood at a depth that, on a living animal, would have reached something important. He moved to eight meters. The tip shattered on the second throw at that distance — a bad-angle contact where the tip hit lateral surface rather than leading-edge-first, the bone fracturing along a grain line he hadn't been able to see.

He retrieved the shaft, examined the damage, and made a note: effective range was defined by the reliability of the penetration angle, and his current accuracy could not consistently guarantee a clean-angle impact beyond seven meters. Seven meters was his range. He knew it now rather than estimating it.

【New Recipe Registered】 Bone-Tipped Spear (Novice) Damage: 18 (direct hit, clean angle) — 4 (glancing/poor angle) Effective range: 7m (accuracy-dependent) Durability: 22/22 (tip) │ 35/35 (shaft) Note: Tip fragility at poor impact angles. Replacement tips required with sustained use. The Watchers note you have achieved ranged capability. They note it with the specific interest of people who know what's coming. They know because they arranged it. Seven meters should be enough. Probably.

"I now have a spear," he said, retrieving it from the bark-target he'd mounted at the seven-meter mark.

"Seven-metre effective range. The boar is large enough to see from seven metres. I'm not saying this solves the boar problem. I'm saying this is the beginning of solving the boar problem." He threw it again. Seven meters, center of the bark-target. Retrieved it. Walked back. Threw it again. The practice session ran until the light started failing. Forty-three throws by the time his arm developed opinions. Thirty-one at the center target zone, eight off-center but in the acceptable range, four genuinely bad.

Seventy-two percent accuracy at seven meters, stationary target, no threat pressure, adequate light. Against a moving target, adrenaline running, the boar already in a territorial response — that number was going to drop. He didn't know by how much. He knew the only way to find out was practice rather than field experience, and he had strong feelings about the field-experience method of accuracy calibration. Seventy-two percent was the number he'd established by not dying.

The boar encounter would require a number he had not yet established. He was going to need considerably more than forty-three throws. Five hundred was the target. He'd said it to himself while setting the throw mark and it was right in the way that data targets were right — not precise, but correctly ordered-of-magnitude. Five hundred practice throws before he trusted his accuracy in an actual engagement. He had the evenings. He had the clearing. He had replacement tip material if the current one broke.

【Training Progress — Ranged Combat】 Practice session: 43 throws (7m, stationary target) Accuracy: 72% (center zone) │ 91% (acceptable zone) Adrenaline-adjusted estimated accuracy: 45–55% Practice target: 500 throws before field deployment Current: 43/500 The Watchers observe you doing math about shooting things. They find this more entertaining than they expected.

He cooked dinner with the improved knife doing the processing work — the efficiency gain was immediate and measurable, the rabbit broken down in seven minutes rather than the eleven the old knife had needed, the portion yield visibly better. The herb-wrapped roast came out better than the previous evening's — not dramatically, but in the particular way that small improvements in preparation quality produced compound improvements in the result.

He ate it and noted it and felt the quiet satisfaction of a system getting better at its own operations. The Ring Channel had been quiet in the background, running its passive register the way it always did. [KaiS] had posted once, in the late afternoon, in a thread about the northern territory. [KaiS]: Haven's Reach has walls now. Player-built. Not just a safe zone. Finn read it twice. Then he put the bowl down. Walls.

Not just an Anchor Stone with its blue dome — actual constructed walls, which meant players working collectively, which meant a community with enough coordination to run a building project. He sat with that for a moment in the way he rarely let himself sit with things — not filing it, not running the analysis, just letting the fact of it exist at full size before the processing kicked in. People had found each other. People were building something.

Somewhere north of his triangle symbols, on the other side of the rust-bark forest and the boar's territory and all the unconfirmed blank space on his map, there was a place with walls that other people had made. Twenty-three days of talking to dead wolves and empty clearings, and there was apparently a town. He added the update to the bark-sheet's northern notation: Haven's Reach — confirmed walls (KaiS source). Player-built. Active community development.

He looked at his own shelter — four walls of his own construction, the thorn-perimeter, the map on the interior wall with all its annotated quadrants. Then at the spear leaning against the east wall, waiting for its five-hundredth throw. The boar didn't know about the five-hundred throw target yet. He picked up the spear, walked to the marking stone at seven meters, and threw it again. Forty-four.