107: Phase 2

Wenqing's briefing on Scattered Fan Phase 2 took thirty-eight minutes and used four whiteboard pages.

I've known people who explain things by making them larger and people who explain things by making them smaller. Wenqing was the second kind — he started with the thing that was wrong, reduced it to its essential mechanism, and then showed you what happened if you changed that mechanism. It was the same way he wrote his patch-prediction columns under the NeedlesAndSilk pseudonym: one observation, one mechanism, one implication, full stop.

The virtue of that approach was that everyone in the room understood the same thing when he was done. Not a version of the thing filtered through their own assumptions, not an approximation they'd filled in from the parts they'd followed — the same thing, exactly. I'd been in enough briefings with commanders who explained badly to know what the alternative looked like: the formation reaches the field carrying six different versions of the plan, and the version that wins is whichever one the leader happens to be running when the match starts.

The whiteboard was a portable model he'd brought from his apartment — one of his specific requirements when we'd set up the Greenleaf Inn room as the permanent war room. He'd hung it on the east wall with two wall anchors he'd installed himself, which the innkeeper had either not noticed or had decided not to comment on. It was the kind of accommodation you could make in a game world that a real-world meeting space wouldn't allow, and Wenqing had used it for every briefing since December. By now the board had seen seventeen major strategy sessions and the cleaning rag had started to leave ghost marks that showed traces of previous diagrams under the current ones — a palimpsest of Severing Light's tactical development over three months.

The thing that was wrong with the Scattered Fan against the layered tank formation: the collapse mechanic hit the first layer and stopped. The first layer was designed to stop it. The second and third layers behind it absorbed the converted pressure and the formation held.

"So we break the first layer differently," I said.

"We don't break the first layer at all," Wenqing said. He tapped the third whiteboard page. "We go under it."

He'd drawn the layered formation as a set of horizontal bars — three tanks in the first layer, two in each of the second and third. Between each layer, a 4-meter gap. The standard approach assumed the attacking formation had to break the first layer to access the DPS line behind it.

"The gap between the first and second tank layers is 4 meters," Wenqing said. "At our current level differential, no single melee fighter can cross it under tank fire without taking 40 percent HP loss. But a Berserker at Lv 32 with Ironbody Stance active and Rending Fury pre-charged has approximately 3.2 seconds of burst survivability in the gap zone before the damage from both layers compounds past the healer's correction speed."

"3.2 seconds," I said.

The number landed in the quiet of the briefing room with the specific weight of a number that determined everything downstream of it.

"3.2 seconds at Lv 32. 4.1 seconds at Lv 33. 5.7 seconds at Lv 34."

"We have 6 days until Round 3."

"You will be at Lv 33 by Thursday. Possibly Lv 34." He looked at his notes. "At Lv 33: 4.1 seconds in the gap zone. You use 0.6 seconds of that to cross the gap. 3.5 seconds remain. The DPS line behind the second tank layer has no direct tank cover. Your Crescent Moon Slash has a 3-meter arc range. In 3.5 seconds at Lv 33 Berserker speed, you can clear a 6-meter section of the DPS line."

I did the geometry in my head. Six meters of DPS line in 3.5 seconds was aggressive. The Crescent Moon Slash arc was 3 meters on each activation, 8-second cooldown, but I'd have Rending Fury in the rotation too — the two-hit strike covered forward ground that the arc didn't. The math was tight. The margin for error was the difference between Lv 33 and Lv 34: 4.1 seconds versus 5.7. That 1.6 seconds was the buffer I'd be working to create by Saturday morning.

Old Wolf had been leaning against the wall since the briefing started. He hadn't said anything yet. Now he said: "You're sending him into the gap alone."

"Yes."

"What happens to the first layer when he goes under."

Wenqing turned to the fourth whiteboard page. "The Scattered Fan formation's west and east points engage the first layer simultaneously. Not to break it — to hold it in place for the 3.5 seconds that Bladeless is inside. The first layer can't rotate to cover the gap because the convergent pressure from both sides keeps it anchored. We call it Scattered Fan Phase 2 because it uses the same initial spread as Phase 1, but the collapse mechanics are different — the two points don't converge toward the center, they converge toward the flanks of the first layer and hold."

"So the Scattered Fan is a distraction," Wanqing said.

"The Scattered Fan is a cage," Wenqing said. "It cages the first layer while Bladeless works inside."

She looked at the whiteboard for a moment. Then she looked at me. The expression was the one she used when she'd assessed a plan and found that it was correct and also that she was not going to be the one taking the risk it required.

"All right," she said. "Then the east and west points need to hold the cage for the full duration without break. What's our tolerance for error on the hold."

"Half a second on either side," Wenqing said. "Beyond that, the first layer can begin rotation and the entry point closes before Bladeless is back through it."

"Half a second." She made a note. "I'll need to practice the hold timing. The east point is precise work."

"You're on the east point because you're the most precise person on the roster," Wenqing said, with the flat candor he used for facts.

She didn't argue it.

Old Wolf pushed off from the wall. "What happens when the 3.5 seconds is up."

"Bladeless exits the gap back through the entry point. The west and east points release the first layer simultaneously. The first layer, which has been stationary for 3.5 seconds while its DPS line was being pressured from inside, will need at least 6 seconds to reorganize and reestablish a cohesive response. We use that 6 seconds to complete the formation reset and address whatever remains of their DPS line."

Silence.

Zhu Yuhan, who had been sitting at the back of the room with her data pad, said: "The entry and exit point for the gap is the same point."

"Yes."

"The first layer will anticipate exit through the same entry point. Two of their three tanks will try to close the entry point when they register that Bladeless has crossed the gap."

"Yes. The exit has to be faster than the entry. Entry at 0.6 seconds — exit needs to be at 0.4 seconds or less."

"The Ironbody Stance timing at exit."

"Full activation. At Lv 33, Ironbody Stance plus the Rending Fury damage-received modifier gives a 1.8-second window of reduced-damage exit that overlaps with the tank closure attempt."

She made a note. "I'll practice the healer timing from the outside. The 1.8-second window is where my correction coverage needs to be pre-positioned."

Old Wolf looked at her. "You've done this before."

"I've done the calculation. Not the real thing."

"You're good at calculations," he said. It wasn't quite a compliment and not quite an observation. "The real thing is messier."

"I know," she said. "I'll adjust."

He nodded.

We ran the drill that afternoon on the flat practice instance, and it was exactly as messy as Old Wolf had predicted. The first four runs I either exited too slow or entered too fast and burned through the survival window before I'd done anything useful in the gap. The problem was the head — when I thought about the entry timing, the body hesitated. When I thought about what came after the entry, the entry went wrong because I'd projected past it. The first four runs were an argument between the part of me that knew the plan and the part that had to execute it, and the part that had to execute it was losing.

In the old timeline I'd run a dozen solo breach maneuvers of similar geometry — smaller-scale, different context, but the same fundamental problem: a narrow window, a pre-positioned response, and a mind that kept arriving at the window before the body did. The solution had been the same then as it needed to be now. The body had to know the entry the way it knew its own weight distribution at rest. Not as a step in a sequence but as a thing that had already happened before you decided to do it. The problem with that level of physical knowledge was that it required either months of repetition or a very specific kind of drilled attention, and we had six days.

On the fifth run I got the entry timing right but accomplished nothing inside — the drill opposition set moved differently from a real AI, and the gap between how I'd modeled the DPS line's response and how the drill set actually moved cost me 1.8 seconds I couldn't recover.

Wanqing said: "Stop thinking about the entry timing. You know the entry timing. Think about what you're doing inside."

"I know what I'm doing inside."

"Then let your feet know the entry."

She said it the way she said things that were obvious once said and not obvious before. I went back to the starting position.

I ran it again.

On the seventh run the entry was clean, the 3.5 seconds inside were productive, and the exit hit the 0.4-second mark with Zhu Yuhan's correction pre-positioned exactly where Wenqing had specified. Old Wolf, from the wall, said nothing. He watched.

"Again," he said.

We ran it six more times.

By the tenth run it felt like something I'd been doing for months. Not comfortable — the gap zone was never going to be comfortable, and making it comfortable would be the wrong goal anyway. But calibrated. The way a piece of work feels when you've done it enough times that the errors are known and accounted for rather than unknown and waiting.

There was a difference between mastery and familiarity that most players conflated. Familiarity meant you'd done it enough that it felt known. Mastery meant you'd done it enough that you could do it differently if the situation required it. The gap-crossing needed to be mastery rather than familiarity — if the match presented a variation, a timing shift, a variable Wenqing hadn't modeled, I needed to be able to adjust inside the sequence rather than run the sequence and hope the variation didn't occur. By the tenth run I was beginning to have that. Beginning. The full version would take another five days and whatever the match on Saturday provided.

***

That evening at the Crimson Ridge I hit Lv 33 at eleven-fifty PM, forty-five minutes ahead of Wenqing's Thursday projection.

*Ding!* [System: Level Up! Lv 32 → Lv 33] [Berserker Class Bonus: STR +4, VIT +3, AGI +2. Stat cap for Round 3 gap-crossing: survival window now 4.1 seconds.]

The level adjustment registered in the haptics as a slight shift in weight distribution — the way the character model's mass redistribution worked at each class tier, the Berserker gaining density at the shoulders and arms where the STR bonus manifested physically. I'd gotten used to reading level-ups by feel rather than by the UI notification. This one felt like more room in the chest than there'd been before. The survival window was now 4.1 seconds.

Zhu Yuhan logged the time in her data file without comment.

"Six days," I said.

"To the match."

"I want to hit Lv 34 by Wednesday."

"Thursday is the projection."

"Wednesday."

She looked at her data file. "Extra session Tuesday night. Three-to-seven, then a two-hour evening cycle."

"Can you do that."

"My father's on call Tuesday night. The house will be quiet."

I didn't say anything to that.

"It's a resource," she said, not looking up. "Not a problem. I'm used to it."

"I know. Thank you."

She went back to her healer rotation log. Not pointedly — she just picked up where she'd left it, with the focused economy of someone for whom work was the natural state. The log was something she'd built on her own, outside the guild's formation data system — a running record of her own correction timing, organized by encounter type and damage source, cross-referenced with the class of mob and the phase of the fight. Wenqing had examined a copy and noted that it was more systematic than anything he'd built for the same purpose. She hadn't built it to impress Wenqing. She'd built it because she needed to know why a correction had landed at the wrong time, and to know that you needed the data, and to use the data you needed to record it, and that chain of reasoning was self-contained and didn't require anyone else's approval to be worthwhile. I'd noticed, over fourteen sessions, that Zhu Yuhan had no comfortable middle gear. She was either working or she was asleep. The space between those two states seemed genuinely unfamiliar to her, the way it was unfamiliar to people who'd learned early that the waiting time was better used.

I cleared the next pack.

*Ding!* [System: EXP +524. Level progress: Lv 33 → Lv 33 (1.2%)]

Two levels to Lv 35. Six days to Round 3. Six days to figure out if the Phase 2 gap-crossing worked against a real sixty-seven-member guild instead of an opposition drill set.

Four days to figure out if Wang Jian was going to counter it. And if he did — whether what I had was enough, or whether I'd reach the edge of what foreknowledge bought and find out what three months of careful work was worth on its own.

That question didn't have an answer tonight. It would have one on Saturday morning. I cleared the next pack and let the work do what it did.