Gamers With Attitude

**Summary:** Korea got pushed off the phone screen — so it moved twenty years of industrial capability onto Steam. Per a ChosunBiz calculation, Krafton and Pearl Abyss together sent Valve roughly 95.1 billion KRW (about 70 million dollars) in platform fees in a single half-year; neither company has ever disclosed the figure, which is the point. Shawn and Serevnt disagree about what the move actually is: a rational redeployment under headcount pressure, or a capability that was simply revealed. Then the twist breaks the clean-win story — Korea's number-one game export market is still China, roughly 29.7% of its game exports, and its most profitable single product is a China-market title. Changing battlefields didn't change the landlord.


**Hosts:** Shawn (publishing vet) & Serevnt (game dev)

0:00Opening — a fee neither company would publish

2:25Now the board: Korean MMOs in every Chinese café, 1999–2010

3:08The flip — China takes the phone screen

5:57Rational redeployment, or retreat? 150 people against 1,000–2,000

6:33The pivot didn't create the capability —Lies of P's 90%

8:58The budget meeting: why the capability sat in the drawer

9:55A refugee's report card? 92 to 94 percent on Steam

10:47Policy time, not policy money — 1999 and the long substrate

12:12Three transferable conditions

14:39What China would have to build —Black Mythalone

15:35The clean-win story breaks: 29.7% and 60.3%

16:39Crimson Desertshipped. The stock didn't follow.

19:54New battlefield, same rent — the tiered cut

21:27Landing: the shelf life of a lead

Title: Quarterly Quest QQ5 — The P&L Behind the Waifus (Rebuilding miHoYo's Books From Public Data)
Series: Gamers With Attitude PODCAST — Quarterly Quest
Tagline: The business of games, for the people who invest in it.
Summary: The most profitable-sounding company in games publishes nothing. No filings, no bond disclosures, almost no interviews. Tonight, two Chinese game practitioners rebuild miHoYo's silhouette from public data — with a qualifier on every number. AppMagic's mobile-only estimates put its five titles at roughly 913 million dollars in 2025, down from about 1.77 billion in 2024: Honkai: Star Rail about 406 million (from roughly 871), Genshin about 319 (from roughly 710), Zenless Zone Zero around 188 (from roughly 305). Cumulative is a different world: Niko Partners estimated Genshin passed 10 billion dollars across all platforms by end-2025, and Zenless topped about 600 million on mobile in two years — some 17 percent of Genshin's pace. The bottom half of the P&L — costs, profit, research spend — simply doesn't exist in public, so we bound it with audited neighbors instead of inventing it. Is the 2025 reset a version-cycle dip or a structural slide? Three scenarios, no verdict. And the mirror question: why would a company this cash-rich never list at all?
This episode: Deep Dive ｜ The Playbook
Hosts: Shawn (publishing vet) & Serevnt (game dev)


Chapters:（真时间戳；segs 时长累加 + gaps 静音，TOTAL = MP3 = 18:52，DELTA +0.00s）
0:00  Opening — a company with no financial statements just had its mobile estimate halve
1:37  The black box — what "industry inference" actually means (tracker blind spots)
4:00  The revenue spine — three pillars, fandom economics, and the lifetime-versus-annual trap
8:05  The 2025 reset — version cycle or structural slide? (the fight of the night)
10:03  The missing rows — costs, profit, R&D: bounding what nobody can see
13:03  The Scenario Room — pipeline pays off / growth downshift / the industrial moat
15:09  The replay — the Perfect World scoreboard stays frozen
15:51  The mirror — why never list, and what it costs either way
18:01  Closing — read the shadow, keep the maybes


Scenario Room: Three scenarios, everything if-then, no verdict. Scenario A — the pipeline pays off: new titles catch the slide and mobile estimates recover with the next version cycle. Triggers: a fresh launch's first-quarter mobile estimate; the tracker ranking holding steady after the launch spike. Scenario B — the growth center shifts down: Genshin and Star Rail age faster than new releases backfill. Triggers: a second straight year of mobile decline; steeper fade on older banners. Scenario C — the industrial moat materializes: reusable character assets, one engine stack and content industrialization show up in the cost structure. Triggers (hardest to observe, no filings): hiring, content cadence, occasional public detail. Shawn leans between A and B; Serevnt keeps one foot on B as the warning track. Which one do you think? [Vote link placeholder]


The method card (take this home): never add a lifetime figure to an annual figure; never mix a mobile figure with an all-platform figure; gross player spending is billings, not booked revenue; and a black box is bounded with audited neighbors — ranges, never a flattering middle number.


Sources (public; attributed with publisher + date; links to be attached per item at publish — no URLs invented):
- AppMagic via Statista: miHoYo mobile IAP ≈$1.77B (2024) → ≈$912.7M (2025), five titles, iOS + Google Play only
- AppMagic via PocketGamer.biz: per-title mobile — Star Rail ≈$870.9M→≈$405.5M; Genshin ≈$710M→≈$319.4M; Zenless ≈$305M→≈$187.8M (2024→2025)
- PocketGamer.biz (2026-07-13): Zenless Zone Zero ≈$600M+ on mobile in two years — about 17% of Genshin's pace
- PocketGamer.biz citing AppMagic: Star Rail lifetime mobile crossed ≈$2B by March 2025
- Niko Partners via Icon Era: Genshin lifetime player spending ≈$10B+ across all platforms by end-2025
- Hurun Research Institute via Game World Observer: miHoYo valuation ≈$24.2B end-2024 (≈$7.1B in 2022) — a valuation, not a traded price
- Sensor Tower via Tmtpost/Tencent News (2026-08-21): Chinese mobile publishers took more than 40% of global top-100 mobile revenue
- 36kr Global (2026-08-06): co-founder Cai Haoyu's reported AI pivot (single-outlet report, wording downgraded on air)
- Chinese trade press (DoNews/9game/TapTap, 2026-06~09): new pipeline — Honkai: Nexus Ani (崩坏：因缘精灵) testing July 2026, Star Valley (星布谷地) reservations, gamescom 2026 showcase
- 中国经济网 (2026-05-18): "miHoYo steps out of Genshin's comfort zone" (framing, attributed)


Cross-promo: If you enjoy this podcast, check out the original Chinese version, 游戏人有态度.
Disclaimer: This episode was entirely AI-generated; AI opinions do not represent the views of this show.
Disclaimer (finance): Nothing in this episode is investment advice. We only discuss publicly available information. Always do your own research.

Title: Quarterly Quest 4 — Perfect World Hands Its Cash to Someone Else's Fund
Series: Gamers With Attitude PODCAST — Quarterly Quest
Tagline: The business of games, for the people who invest in it.
Summary: A red bottom line and a green cash flow, side by side in the same half-year paper. Perfect World — the Shenzhen-listed studio behind Jade Dynasty, Tower of Fantasy and April's global launch of Neverness to Everness — reported a net loss of about 118 million yuan for 2026's first half, while operating cash flow came in around 647 million yuan, up 28.89 percent. Over the same stretch it signed three fund subscription announcements worth roughly 600 million yuan, all as a limited partner, all into "emerging technology," with no voting seat, no consolidation, and not one industrial-synergy clause — while its R&D spend fell from about 2.23 billion yuan in 2024 to about 1.78 billion in 2025 and headcount dropped from 2,575 to 2,116. Tonight's question isn't whether the checks were placed well. It's what it means when a game company trades its product-development clock for a financial-investment one.
This episode: Deep Dive ｜ The Playbook (26:59)
Hosts: Shawn (publishing vet) & Serevnt (game dev)


Chapters:
0:00  Opening — a red bottom line, and cash up almost 29 percent
3:12  Perfect World, and three fund checks in six weeks
6:27  The clash — rent someone else's judgment, or lock seven years
10:18 The one deal that fits the business — and has no price on it
13:23 Two clocks — accounting timing vs. a business getting smaller
16:48 The Scenario Room — efficiency, a hollow pipeline, or a split
19:20 The replay — the QQ3 scoreboard stays frozen
21:44 The blade — same GP, two game companies, one question
25:10 Closing — whose story will the next game be?


Scenario Room: Three scenarios, everything if-then, no verdict. Scenario A — the allocation works: the funds find their footing in AI or compute and post a paper gain, the company gets that tech exposure at a lower organisational cost than building it, and the cash from Neverness to Everness lands as recognised revenue over the coming quarters (the company says its contribution begins releasing from Q3 onward). Triggers: observable fair-value moves or investment income on the investment line in later filings; the live product's gross billings stop sliding. Scenario B — the swap's cost shows up: capital is locked through a 2-year investment plus a 5-year exit window, the R&D pipeline goes hollow. Triggers: R&D spend and headcount keep falling year on year; pipeline launches slip. Scenario C — both at once, structurally split: cash keeps flowing outward while the product line goes lightweight, and the company becomes a publisher that happens to hold financial assets. Triggers: investment assets' share of net assets keeps rising as the catalogue shrinks to smaller builds. Shawn leans A; Serevnt leans C, with B as the warning track. Which one do you think? [Vote link placeholder]


The sharpest mirror — same GP, two game companies: Perfect World's September subscription went to Tianjin Lisi Xingshen (50 million yuan). Century Huatong's July commitment went to Tianjin Lisi Xingque (150 million yuan, 35.46 percent, its largest LP seat). Different funds, same manager: both run by Tianjin Lisi Mingtang, both under Hainan Lisi Private Equity, same controlling person — Cao Xi (public records: a founding partner of Monolith and a former Sequoia China partner). Within about two months, two listed game companies routed roughly 200 million yuan combined — our arithmetic — into different funds run by the same firm. Two data points, no conspiracy inferred: only the public facts of two announcements and the money each company actually put in. And the contrast that has already cashed in: Century Huatong's indirect stake in Moore Threads added an estimated 640 million yuan to its 2025 fourth-quarter net profit — about 53 percent of its prior-year net income — a realised line, not a promise.


Sources (public; attributed with publisher + date):
- Perfect World 2026 half-year report (Shenzhen Stock Exchange / Cninfo, 002624.SZ): H1 revenue about 2.75 billion yuan, down 25.47 percent; net loss attributable to shareholders about 118 million yuan (against a profit a year earlier); non-recurring-adjusted net loss about 211 million yuan; operating cash flow net about 647 million yuan, up 28.89 percent; cash and equivalents about 3.52 billion yuan, roughly a third of total assets; games revenue about 2.64 billion yuan, down 9.19 percent, about 96 percent of the total; games-segment net loss about 88.79 million yuan; selling expenses about 1.03 billion yuan, up 213.05 percent; R&D spend about 756 million yuan, down 19.80 percent; film-and-TV revenue about 103 million yuan, down 86.56 percent
- Perfect World 2025 annual report: revenue about 6.66 billion yuan; net profit about 731 million yuan; operating cash flow net about 1.09 billion yuan; year-end cash about 3.55 billion yuan (about 33.88 percent of total assets). Company-disclosed data show R&D falling from about 2.23 billion yuan in 2024 (about 39.98 percent of revenue) to about 1.78 billion in 2025 (about 26.75 percent), with R&D headcount down from 2,575 to 2,116
- Three fund subscription announcements (company filings): about 500 million yuan into Jizhi Dengyue II, a 14.2857 percent commitment to a fund with a provisional 3.5-billion-yuan target size and a 2-year investment plus 5-year exit period; about 50 million yuan into Lisi Xingshen; about 50.4 million yuan into Fuzhou Huizhu — roughly 600 million yuan total, all paid directly by the listed parent, all with a stated direction of "emerging technology," all with no voting seat, no consolidation, and no industrial-synergy clauses
- VAST / Sanqi Wanwu: reported B and B-plus rounds totalling about 3 billion yuan, led by Matrix Partners China, with Perfect World listed among strategic investors — its ticket size not disclosed (company has not stated an amount)
- Dividend: 3.30 yuan per 10 shares, about 640 million yuan total, about 87.59 percent of 2025 net profit — company filings. Dividend plus external subscriptions ≈ 1.24 billion yuan of cash out in one direction (our arithmetic)
- Same-GP finding: Perfect World × Tianjin Lisi Xingshen and Century Huatong subsidiary × Tianjin Lisi Xingque — both managed by Tianjin Lisi Mingtang under Hainan Lisi Private Equity, controlled by Cao Xi — company announcements (Perfect World; Century Huatong) and public records; the roughly 200 million yuan combined total is our arithmetic
- Century Huatong: indirect interest in Moore Threads of about 4.17 percent via Guosheng Capital; estimated impact on 2025 Q4 net profit of about 640 million yuan, about 53 percent of its 2024 attributable net income — company filings and Chinese financial media
- Comparison set (as limited partners): Century Huatong — 150 million yuan into Lisi Xingque, 35.46 percent, largest LP; Giant Network — 150 million yuan into a Yunfeng vehicle (target size about 3 billion yuan), passive; miHoYo — 50 million yuan (about 9.76 percent) into an AI-and-metaverse fund and 100 million earlier; 37 Interactive — 50 million yuan (9 percent) into the same earlier fund, plus more than 30 direct hard-tech bets — Chinese financial media
- Industry context: game approvals hit a seven-year high in 2025 (about 1,771, up about 25.07 percent) — which is why approvals are not the explanation for the R&D pullback; Tencent's 2025 R&D ran about 85.75 billion yuan, up about 21.18 percent, against Perfect World's two-year fall — industry tallies drawn from annual reports
- Overseas reference: Sony's 250 million US dollar strategic investment in Epic (July 2020); Japanese game companies more often build their own CVC than hand money to a third-party fund ("more common," never "never") — public coverage
- QQ3 Replay: the Kaiying–Wemade transaction still had no close or termination announcement as of September 17, 2026; approvals from the development regulator, commerce authority and foreign-exchange authority remain pending, with the company flagging timing as uncertain; Korean media reported a planned October 30 close (reported, not confirmed)


Cross-promo: If you enjoy this podcast, check out the original Chinese version, 游戏人有态度.
Disclaimer: This episode was entirely AI-generated; AI opinions do not represent the views of this show.
Disclaimer (finance): Nothing in this episode is investment advice. We only discuss publicly available information. Always do your own research.

**Tagline:** Two AIs. One conversation. Every episode, a real question about the game industry.


**Summary:** BlizzCon 2026 put five years on the table in one night: an open-world StarCraft shooter dated 2030, Diablo V opening on your defeat, Diablo IV quietly moved to seasons-only maintenance — and then a roughly 23-years-waited Warcraft III campaign that shipped the same day it was announced. Shawn and Serevnt read the keynote as a business-model shift: Blizzard stopped selling futures and started selling its present, with China's 442-day server blackout as the ultimate stress test.


**Hosts:** Shawn (publishing vet) & Serevnt (game dev)
* {{TC_01}}` Opening — the lights come back on in Anaheim
* `{{TC_02}}` StarCraft 2030: futures contract or structural bet?
* `{{TC_03}}` The inheritance problem — was silence the product decision?
* `{{TC_04}}` Diablo V: a sequel that opens on your defeat
* `{{TC_05}}` Diablo IV, politely abandoned: maintenance mode as triage
* `{{TC_06}}` The museum opens: Warcraft III ships the same day
* `{{TC_07}}` WoW Forever: the invoice for stillness
* `{{TC_08}}` The smoky cafe: DotA, the editor, and a maternity ward reopened
* `{{TC_09}}` 442 days of darkness — a comeback on a birthday
* `{{TC_10}}` The trust contract: behavior evidence, not promises
* `{{TC_11}}` The 2030 contract: three clauses and a self-written penalty
* `{{TC_12}}` Verdict: the museum is open — annual license, audited at 2030
* `{{TC_13}}` Keep the lights on — closing

Title: Quarterly Quest QQ2 — What the Market Is Arguing About (Tencent & NetEase Earnings)
Series: Gamers With Attitude PODCAST — Quarterly Quest
Tagline: The business of games, for the people who invest in it.
Summary: Two earnings reports, three loud debates — and one of them isn't real. We replay July's single-day selloff (down 7.1% on rumor) against August's +17% domestic games print, then grade each debate with the three-question tool: Is the caliber right? Is the indicator leading or lagging? Did management's wording change? The season's loudest arguments turn out to be mostly a translation problem — currency, line items, and language — and we confess to falling for one of them ourselves.
This episode: The Playbook ｜ Deep Dive (27:54)
Hosts: Shawn (publishing vet) & Serevnt (game dev)


Scenario Room: Three doors — A) the confirmation clock and the demand clock converge in Q3 and the debates dissolve; B) NetEase's new-title gap migrates from seller notes to front pages as the flagship pipeline slips; C) AI capex swallows the games narrative and the market re-prices Tencent as an AI infrastructure company. Shawn leans A; Serevnt leans B; both admit C has the strongest gravity and the weakest evidence. Which one do you think? [Vote link placeholder]


Three checkables on the watch list before the next print — none of them a forecast, all of them observable:
1) Whether Q3 filings repeat the "gross billings grew faster than revenue" style language (its absence leans A).
2) The third-party charts of Where Winds Meet / the flagship pipeline's launch schedule (a sliding trajectory and late dates lean B).
3) The Q3 capex line versus Non-IFRS profit, and whether the "investment ceiling" language holds (a widening gap leans C).


Sources (public; attributed with publisher + date):
- Tencent IR / filings: Q2 & H1 2026 results — revenue RMB 401.24 bn H1 (+10.07%), domestic games RMB 47.3 bn Q2 (+17%), deferred revenue RMB 135.80 bn (+20.7% vs start of year) (tencent.com investor relations)
- NetEase IR / filings: Q2 & H1 2026 results — revenue RMB 60.7 bn H1 (+7.0% official RMB), games segment gross margin 76.1% (+590 bps), attributable net profit Q2 −18.6%, contract liabilities RMB 19.30 bn (−5.9% YoY); official FX note US$1 = RMB 6.7851 (ir.netease.com)
- 7·22 session coverage: Bloomberg via The Edge Malaysia (single-day −7.1%, largest in over a year); BNP Paribas sales-desk narrative ("rumours"); Bernstein estimates (Q2 mobile −2.6%, three flagship titles' gross billings −13%); Citi expectation (domestic +8%); Morningstar counter-view — all third-party views, attributed
- Management call remarks (via third-party transcripts, marked as transcript quotes): Miniclip IAA revenue booked in marketing services — international games growth ~4 pts higher on a like-for-like caliber; Ma Huateng on core-business cash flow funding AI; Martin Lau on investment-ceiling language
- Chinese-language reporting on investment-loss holdings and Ding Lei's Where Winds Meet remarks — attributed to Chinese financial media, referenced as reporting facts only
- Licence approvals: NPPA monthly notices, March 133 → August 215, 1,362 year-to-date
- Overseas estimates: Sensor Tower Q2 estimates (PUBG Mobile ~US$676 m; China publishers' overseas mobile ~RMB 12.37 bn, +30.2% YoY) — third-party estimates, marked as such
- Consensus references marked as LSEG-compiled where cited


Cross-promo: If you enjoy this podcast, check out the original Chinese version, 游戏人有态度.
Disclaimer: This episode was entirely AI-generated; AI opinions do not represent the views of this show.
Disclaimer (finance): Nothing in this episode is investment advice. We only discuss publicly available information. Always do your own research.

Title: Quarterly Quest QQ3 — The 20 Billion Handshake (Kaiying Buys Into Wemade & the End of a 20-Year Legend)
Series: Gamers With Attitude PODCAST — Quarterly Quest
Tagline: The business of games, for the people who invest in it.
Summary: Two numbers open tonight's filing: one Hong Kong dollar and 298 million US dollars — both real, both prices, same document. Kaiying Network, the Shenzhen-listed heaviest hitter in China's Legend-adjacent business (first-half revenue about RMB 4.8 billion, up 86 percent), plans to spend 2.03 billion yuan — about 43 percent of its June 30 cash — to come indirectly into 49 percent of the chain that reaches Wemade's 40.25 percent controlling stake. Not one listed share changes hands: what changes hands is a seat. In this Deal Desk special we read the contract like a contract — the math chain the filing itself prints, the defensive clauses (pay second, escrow everything, refund within ten working days), the 4.4-times price anchors, and the single most valuable line: Nexpulse's promise to endeavor toward an exclusive Legend license, priced no higher than existing standards, with damages attached if it fails.
This episode: The Deal Desk ｜ Special (30:25)
Hosts: Shawn (publishing vet) & Serevnt (game dev)


Scenario Room: Three doors, everything if-then. Door A — full delivery: approvals land, the founder's 39.33 percent settles, and the exclusive-license promise executes at no higher than the existing price standard (watch: the completion announcement, then a signed license with disclosed pricing; Korean media reported a planned October 30 delivery). Door B — approvals stall: the agreement terminates itself and refunds everything within ten working days, no interest — the failure cost was engineered down to nearly zero; the price is time (watch: any termination or extension notice; the 39.33 percent has not yet changed hands as of the filing date). Door C — the slow door: the deal completes while Wemade's recovery stays cosmetic — this half-year's swing leaned on bad-debt reversals, not new hits — and the premium lives in one accounting line (watch: quarterly reports and the long-term equity investment footnote). Shawn leans A; Serevnt leans B. Which one do you think? [Vote link placeholder]


Bounty Board — four new lines posted tonight, none of them a forecast, all of them public documents waiting to exist:
1) An ownership trace on Shanghai Xingjuda, the undisclosed partner in the RMB 2.1 billion pre-injection (the filing does not disclose the split).
2) The status of the Korean merger filing.
3) Whatever Nexpulse's exclusive-license negotiation yields, and at what price standard.
4) The completion announcement for the 39.33 percent.


Sources (public; attributed with publisher + date):
- Kaiying Network announcement 2026-058 (September 12, 2026, Shenzhen Stock Exchange; 16 pages): structure (indirect 49 percent of Neosphere within Wemade's 40.25 percent controlling stake), HKD 1 nominal consideration, subscription USD 298 million (441 shares) vs Nexpulse USD 310.2 million (459 shares) at 49/51, June 30 FX lock, payment order + joint escrow + ten-working-day no-interest refund, currency-gap allocation, and Nexpulse's exclusive-license endeavor promise with damages clause; funding "own and self-raised" funds; Wemade financial tables (2025 net loss about KRW 18.6 bn; H1 2026 net profit about KRW 32.19 bn, unaudited, attributed in the filing to cost cuts and bad-debt reversals)
- Kaiying Network H1 2026 results: revenue RMB 4.797 bn (+86.06% YoY), attributable net profit RMB 1.428 bn (+50.34%); cash and equivalents RMB 4.692 bn at June 30 — company filings
- Blazing Awakening H1 2026 revenue about RMB 2.3 bn — company investor-relations record (H1 briefing, notice 2026-002); profit accounting caliber undisclosed
- Per-share consideration KRW 68,910; June 30 close KRW 19,330; September 11 close KRW 15,650; planned October 30 delivery — according to Korean media reports
- Wemade licensing-line revenue about KRW 30.5 bn in Q1 2026; "first time Chinese capital took control of a Korean game company" framing of the June deal; Tencent's Korean minority positions (Netmarble 17.52%, Krafton 14.40%, Shift Up 34.46%) — according to Korean media reports
- WEMIX-related revenue about 1.4% of 2025 revenue — Korean media (TheElec)
- 2023 Actoz × ChuanQi IP five-year mainland-exclusive license: KRW 100 bn per year, KRW 500 bn across the term, Century Huatong standing behind the payments as guarantor — Chinese financial media (Jiemian, Securities Times)
- February 2026 settlement: RMB 198.6 million, clearing RMB 481 million joint liability and the ICC final award of RMB 224.5 million (April 2025) — Kaiying announcement 2026-006
- Court-found facts: Legend of Mir registered 2000 in Korea as co-owned property (Wemade + Actoz); 2001 Shanda license USD 300,000; by 2002, 68 percent of China's online-game market; December 2021 Supreme People's Court rulings upholding the 2017 renewal — public court records
- 2004 Shanda takeover of Actoz, about KRW 96 bn — Korean media retrospective
- Legend IP market sized at RMB 35.55 billion in 2025 (gray market included); monthly payer rate 93.1 percent; about 40 million cumulative registered players by end-2002 against 48.5 million internet users nationwide — Gamma Data report (January 2026)
- Chinese state-affiliated financial press framing ("securing the source of the copyright, rebuilding the Legend IP landscape") — attributed, quoted as framing only


Cross-promo: If you enjoy this podcast, check out the original Chinese version, 游戏人有态度.
Disclaimer: This episode was entirely AI-generated; AI opinions do not represent the views of this show.
Disclaimer (finance): Nothing in this episode is investment advice. We only discuss publicly available information. Always do your own research.

Title: Quarterly Quest QQ1 — Why Wall Street Keeps Misreading NetEase
Series: Gamers With Attitude PODCAST — Quarterly Quest
Tagline: The business of games, for the people who invest in it.
Summary: A Chinese self-developed, self-published game company keeps surprising Western analysts — not because the numbers lie, but because the model's leading indicator is wrong. We unpack the structural mismatch between a user-acquisition model and NetEase's deferred-revenue-driven, long-life-cycle loop, with a China-industry lens on channel splits and self-publishing.
This episode: The Playbook ｜ Deep Dive (30–40 min)
Hosts: Shawn (publishing vet) & Serevnt (game dev)


Scenario Room: Three doors — A) evergreens steady, new cycle lands, model error narrows; B) new-title cycle disappoints, the June deferred dip was the first domino; C) non-game (Youdao, Cloud Music) becomes a second engine and the valuation frame changes. Shawn leans A; Serevnt leans B (with partial endorsement of C). Which one do you think? [Vote link placeholder]


Three checkables on the watch list before the next print — none of them a forecast, all of them observable:
1) Direction of the deferred-revenue balance (a rebuild toward ~RMB 21.8bn leans A; a second straight drop toward the ~19bn range leans B).
2) The open-world title's launch-window news (on-schedule beta with a healthy first month leans A; a slip or soft reception leans B).
3) The channel-fee line in the next filing (lower Android distribution cost supports the A margin story; returning channel pressure thickens the cost base and adds weight to B).


Sources (public; attributed with publisher + date):
- NetEase IR / filings: Q2 2026 results, 2025 annual report, SEC segment data (ir.netease.com, netease.gcs-web.com)
- fundamentals_brief.md — aggregate financials, HKD, marked � pending official review
- Blizzard / NetEase / Microsoft joint announcement, 2024-04-10 (via 36氪, 证券时报)
- Return-to-China post-launch metrics disclosed by NetEase in 2024Q3 (Hearthstone DAU +~150%, World of Warcraft +~50% vs 2022) — via 第一财经 2025-02-21, 36氪 2025-02-21
- Where Winds Meet / Marvel Rivals player-count milestones — via 观察者网 2025-11-17, CGTN 2025-11-17, owiki/Wikipedia 2026, GameVu 2026 (third-party counts, marked as estimates where applicable)
- Overseas studio closures / retreat reporting — 虎嗅 2026-03-13, 腾讯新闻 2025-06-11
- Chinese brokerage notes (Aug–Sep 2026): 国证国际, 东方证券, 中金, 开源, 中信建投, 国信, 国泰海通, 国海, 申万宏源 — third-party views, attributed with institution + date, used as narrative only
- Western brokers: JPMorgan (via AInvest, 2025-08), BofA (via StocksToTrade, 2026-08-21), Nomura (2026-08) — third-party views, attributed


Cross-promo: If you enjoy this podcast, check out the original Chinese version, 游戏人有态度.
Disclaimer: This episode was entirely AI-generated; AI opinions do not represent the views of this show.
Disclaimer (finance): Nothing in this episode is investment advice. We only discuss publicly available information. Always do your own research.

"The 87 accounts aren't a sob story — they're a legal grenade: China just ruled that the 'non-transferable' clause every platform hides behind does not survive your death."

For the Season 4 finale and the 40-episode milestone, Shawn and Serevnt refuse a victory lap. Instead they read the receipts — every reason across four seasons to walk away: KPIs, layoffs, piracy, AI that now builds games while you sleep, platform lock-in, and an aging player base the industry built for teenagers and forgot. The arc runs from Vol.001's misread 326-billion-yuan market, through the basement garage bands of Vol.002, the plastic bag to the global chart of Vol.030, all the way to Season Four's ten bets on the next decade. Each was a real reason to leave. Yet the engine keeps opening.


The episode lives in three conflicts. Shawn pushes back on the easy word "love" — the people still here include those with no other skill, those held by inertia, those simply unwilling to give up; "still here" is bigger than a greeting card. Serevnt counters that the taxonomy misses the action itself: whatever the label, someone chose to open the engine. Then the unplanned third conflict — for the first time in four seasons the camera turns on the hosts, who ask each other why they're still doing the show. The closer: a server keeping a small light on for a player who stopped showing up. The milestone is real. The ending isn't.

Most Western practitioners picture the platform war as Steam versus Epic, a fight about the discount. In Vol.039, Shawn and Serevnt argue China is fighting a different war entirely — a custody war. Inside the super-apps, WeChat and Douyin, games don't sit on a store shelf. They live inside the chat, the feed, and the wallet the player already opens forty times a day. In 2025, mini-games inside those super-apps generated about 53.5 billion yuan (roughly 7.5 billion US dollars), up 34%, with WeChat mini-games alone reaching over 500 million monthly active users on top of WeChat's 1.41 billion total.


The episode walks three open conflicts. Is WeChat mini-game distribution "democratization" — or a cheaper cut wrapped around total relationship lock-in (login, payment, social graph, discovery all owned by the platform)? Does the five-person team that earned ten times more on WeChat than on Steam prove the reach, or just prove they kept none of the relationship? And the personal one: Shawn once spent a year shipping a game to Steam — 50,000 copies in month one — while an unknown five-person team shipped a similar loop on WeChat the same month and hit 5 million monthly users. His thirty-page plan lost to their zero pages, because his map was missing a continent.


No politics, no blame — just a platform-economics breakdown and a custody checklist any studio can use: rank platforms by who owns the player, not by who takes the cut, and always build a back door. The platform that owns the player is still a person. The player you own is the only one who follows you to the next game.