221: CW IV Group Stage

The CW IV group stage ran fourteen days.

Same format as CW III: fixed member cap of 80 for the group stage. The guild's group stage record was undefeated over CW III — four matches, four wins. The four wins had been clean, mostly. Two opponents with genuine defensive structures; two others who'd relied on raw member coordination and hadn't built anything underneath it. Four wins was the baseline expectation going into CW IV.

The group stage had its own rhythm — fourteen days with matches every three or four days, the in-between time spent on the analysis cycles that Wenqing ran after each session and before the next. Post-match review, opponent modeling, the pre-match summary that arrived twelve hours before every scheduled encounter. The rhythm was familiar now, in the way that a three-year-old process was familiar. Not comfortable in the sense of easy — the tactical problems were real — but navigable in the sense of having tools for every layer of the navigation. That was what three years bought: not certainty, but the tools for uncertainty. You walked into each match with a model. The model was tested by what happened. The model was updated. Wenqing filed the update. That was the process.

Group A this cycle: Black Dragon Guild seeded first, plus five other guilds. Two were unseeded entries with no previous CW history — the kind of guilds that made it through server qualifications by being locally dominant without ever facing a top-tier formation. One was a mid-server guild that had finished sixth in CW III, which meant they'd placed behind five guilds that had more than their current resources. And then there was Emerald Banner.

The group stage had the particular energy of competition that is technically serious but not yet the serious it would become later. You didn't carry the weight of the previous match's result into the next one. You carried analysis, you carried the Wenqing pre-match summary, you carried the formation's developed instinct — but not the weight. The knockout rounds had the weight. The group stage had the rhythm.

The mid-server guild had upgraded their formation between CW III and CW IV. Ningxia's analysis: *They recruited three experienced players from disbanded guilds in the off-season. Their formation has more depth than their CW III version. Still below top-tier, but the gap has closed.*

Three players from disbanded guilds. The kind of targeted acquisition that signaled a formation council that was thinking seriously — looking at what they lacked and finding the specific human additions that would close the gap. That was not a passive approach. That was a guild building with intention.

The three matches against the unseeded and lower-seeded guilds ran as expected: 60–0, 60–3, 60–11. No significant complications. The resonance at 80-member cap performed at the 69% efficiency level — consistent with the CW III group stage baseline and with Wenqing's projection. The sessions before the tournament had held.

The Emerald Banner match was the fourth.

***

They opened with a revised formation. Not the same configuration as CW III, which Wenqing had fully documented. The new configuration used a mobility-heavy approach: five rapid-repositioning damage dealers who changed formation positions every twenty seconds, cycling through three preset coordinates in a rotation that looked chaotic from the outside but had the internal logic of a clock hand.

I recognized the design almost immediately. Not from CW III — from something older, a tactical structure that had appeared in the first-timeline server around the period I was now living through. Someone on Emerald Banner's formation council had done real research. The kind of research that required reading competitive records from multiple tournament cycles and extracting principles, not just tactics. They'd done the work. The formation showed it. The first thirty seconds of the match made that clear: the five dealers moved with the precision of people who had practiced this rotation until the positioning was instinct. They knew exactly where they were going and when.

I felt something specific when I recognized the design — not alarm, not concern. A kind of respect, the same quality of attention you gave to a well-constructed argument. The formation had been built by someone thinking seriously. The right response to serious thinking was to think seriously back. Not to dismiss it and not to be intimidated by it. To engage with it at the level it deserved, which was the level of its actual quality.

Ningxia's real-time note: *The mobility approach is designed to prevent sustained resonance application to the same members. If the 80-member cap means only 14 members are in range at a time, constant repositioning means those 14 are always different people — the resonance can't compound on a consistent cohort.*

Compounding. The resonance's highest efficiency came from the same members receiving multiple augmentation intervals in sequence — the effect wasn't additive, it was multiplicative when layered on the same recipients. Constant repositioning broke the compounding dynamic by ensuring no member stayed in range long enough for the second interval to arrive.

*The same counter-principle as Amber Ascent's rotation,* I sent. *The counter to the counter: find the coordination seam.*

*Confirmed,* Wenqing said. *The repositioning cycle requires a timing coordinator. Watch for the signal.*

I watched. The repositioning wasn't random — the five dealers moved in a specific sequence, and the sequence required someone to initiate it. At 20 seconds, the coordinator's signal arrived as a brief targeting-lock pause, a half-second where the coordinator's cursor stopped before the repositioning triggered. At 20-second intervals. Like a clock.

The repositioning cycle was on a 20-second interval, which was faster than Amber Ascent's 38-second rotation. The coordination overhead was proportionally higher — moving five dealers on a precise schedule every 20 seconds required the coordinator's full attention. That level of sustained attention had a ceiling. Attention that was fully occupied couldn't also adapt when the pattern needed to change.

At minute eighteen: the first seam. The repositioning coordinator's signal was delayed by 1.8 seconds — faster to develop than Amber Ascent's seams because the shorter cycle meant less slack in the timing. One moment where the coordinator held instead of moving, because the formation had drifted two meters out of position and the reset required a beat. That was all a seam needed to be. A beat was enough.

I hit the window. Eleven members in simultaneous range. 24% augmentation.

Match result: 60–18. 51 minutes.

The group stage: 4 wins, top of Group A, advancement to the knockout rounds. The resonance had held at 69% efficiency across the four matches, which told me the formation was calibrated correctly for the cap and the travel time between opponents hadn't introduced drift. Clean. The kind of clean that came from the session work Wenqing had been running since the previous cycle.

***

Iron Frost Ascent's group stage results arrived through the server achievement board on November 17.

Group C: four wins, top of Group C. Matches run in 42, 38, 51, and 44 minutes. Four opponents, four different challenges, and a 38-minute clean kill somewhere in the middle that stood out against everything else in the bracket data.

Their fastest match was 38 minutes. The fastest our guild had run in the group stage was 40 minutes, in CW III's second match against a formation that hadn't developed past their previous year's structure. I'd looked at that number when the match ended and thought: the formation is operating well today.

Iron Frost's 38-minute match was against a competent mid-server guild that had placed eighth in CW III. Not an easy opponent. Not the kind of guild you run in 38 minutes unless something about your formation is genuinely ahead of where it was expected to be. I looked at the timestamp on the achievement board entry and then at the opponent's server rank and ran the arithmetic twice because the first time felt wrong.

Wenqing: *Their 38-minute match was against a competent mid-server guild. That's the fastest group stage time in this tournament. Their formation efficiency is higher than ours was at the equivalent CW stage.*

Higher efficiency at the equivalent CW stage.

I sat with the sentence for a moment. Not resistance — assessment. The statement was precise. Not "they're stronger than us" — that wasn't what Wenqing had said. He'd said their efficiency at this stage was higher than ours had been at the same stage. Which meant they were building faster. Which meant either they had advantages we hadn't had at their stage, or their design was structurally more efficient, or both.

They were building faster than we had.

Bai Yueran's message came in the same hour: *MoonShadow advanced from Group B. 4-0. We're in the knockouts.* A pause. *I watched Iron Frost Ascent's group stage kill-cams. The commander and the primary healer are operating at a level I've never seen from a first-year competitive guild.*

First-year competitive guild. They'd been founded in September 2017 — fourteen months. Fourteen months from founding to server rank 7 to CW IV top-of-group in the most competitive format on the server. The trajectory was steep enough to be notable even to Bai Yueran, who had seen what sharp trajectories looked like from the inside.

We'd gone from founding to CW I third place in eleven months. And we'd had the Pioneer's Path data and the first-timeline knowledge and Wenqing's archive from the first week.

*Someone came back,* I sent to Bai Yueran.

The sentence sat between us for a moment.

She was quiet for a long moment.

*From a different future than yours,* she sent.

*Yes. Possibly.*

*Does that mean they know things about us.*

I thought about it. The previous-timeline me had not won three CW championships. Had not activated the Heaven-Severing class. The Iron Frost Ascent's founders were building toward something, not building from knowledge of our specific history. FrostDragon and QingxueTide — they were building with the urgency of people who understood what was possible, who'd seen how high the ceiling could be, who knew that the formation they were building now needed to be something that would survive what they knew was coming. That was different from knowing our specific moves.

*They know their own futures,* I sent. *Not ours.*

*Is that good or bad.*

I considered it. A second returning player, on the same server, was not a threat in the usual sense — we weren't competing for the same objective, weren't working from the same starting conditions. But a second player with foreknowledge of the network's methods, building against the same pressure from a different angle, was something I didn't have a clean category for.

*Interesting,* I sent.

She sent a brief laugh-adjacent message. Then: *Round of 16 is November 24. Our opponent is Jade Horizon — the same guild from CW III. We beat them 60–9.*

*Yes. The 5-meter engagement distance will be even more effective now than in CW III.*

*I know. Ningxia has the analysis.*

***

Wanqing at the November bench.

The maple had passed peak autumn color — the late-November quality where the remaining leaves were the deep orange that stayed longest, the gold mostly gone, the branches beginning to show through. The campus had the November settled quality that accumulated as the semester ran toward its end: the library more occupied, the dining halls quieter after evening hours, the research buildings lit until later than they'd been in October. November had a specific texture on this campus — not the fresh-cold of December, not October's color, but the working cold of a semester in its final weeks. Students on the paths with the particular calculation of people who knew exactly what they still had to complete.

She was reading when I arrived, the bench in the late-afternoon light. She didn't look up immediately. This was the settled version of Wanqing at the bench — when she'd been sitting for a while and the work had its own current going. I sat down and poured from the thermos without interrupting it. The November air was cold enough to make the tea worth drinking rather than sipping.

"Iron Frost Ascent," she said.

"4-0 in the group stage. 38-minute fastest match."

"You mentioned them to me in August. FrostDragon and QingxueTide." She looked at the late-autumn maple. "You said it was interesting."

"Yes."

"Still interesting?"

"Yes." I looked at the maple, the deep orange that would be gone in a week. "The quarterfinal bracket draw is November 18. We could face them at the quarterfinals."

She turned a page. "What would you do."

"Run the match. See what they've built." I looked at the campus — the November light, the long shadows at this hour. "And try to understand which timeline they came from."

She looked at the maple for a long time. Not the processing look — the kind of attention she paid when a thing was genuinely new to her and she was deciding how to hold it. The late-autumn campus around us had the particular quality of a place where a lot of work was being done — the library windows lit, students crossing the paths with the purposeful urgency of end-of-semester, the year gathering toward its close. None of them were aware of Iron Frost Ascent or FrostDragon or the question of which future someone had come back from. That was fine. They had their own work.

"If FrostDragon came back," she said. "The way you came back. What does that mean for them."

The question was specific. Not what does it mean for us — what does it mean for them. She'd gone immediately to the person rather than the strategic implication.

"It means they're also trying to build something that holds," I said. "Just with different starting conditions."

She looked at the maple. "That's not a threat," she said.

"No," I said. "It's not."

A pause. The November bench, the deep orange leaves. She went back to the problem set. The bench in the November late-afternoon was its most structural version — past the color, the branch framework visible, the maple's architecture showing through the last leaves. Three years at this bench and I'd seen the maple go through its full cycle four times. This was what it looked like at the beginning of its winter phase. Stripped down to what it was.