October 23 evening · Rank changes and candidates

Tonight’s highlights

• Groundwork: US bestsellers 62→34; make the recovery chain legible
• Moon Sweepers: purchasable today, with cleaning feedback under scrutiny
• Escape: a lower rank alongside more reviews

Steam: actions and rewards

Today and the next 14 days

For the next two weeks, Probably Stolen is planned for October 28, linking inspection with pricing; Emberville for October 27, connecting rescue with new town functions. Nomad Drive shows October 27 in the English list and October 28 in the Traditional Chinese detail; that discrepancy remains unresolved. These announced games are direct substitutes, not an empty market.

Probably Stolen - Cyberpunk Shopkeeper Sim · Emberville · Nomad Drive · Groundwork Simulator · Scam With Your Friends · Escape from Playtime · Moon Sweepers · Phantom Blade Zero

Eight games: three future, three recent, one today and one attention phenomenon. Phantom Blade Zero is an attention watch, not a low-cost H5 primary recommendation.

TapTap: limited release observations

Limited TapTap creator-linked evidence verifies first-playable dates of October 23 for 篮球少女：飞跃巅峰 (Basketball Girls, editorial shorthand), rated 8.3/10 from 1,097 ratings, and October 22 for 失落城堡2 (Lost Castle 2), rated 7.2/10 from 2,214 ratings. Rating counts are not new players; an associated vendor is not automatically the developer.

篮球少女：飞跃巅峰 · 失落城堡2

Upstream themes: follow changes, not labels

On Hongguo, 和离后我带女儿横扫魔域 moved 9→3 overall and 4→1 on the comic-drama list, with displayed composite heat 60.94m→67.81m. It remains one work, not two independent diffusion sources. 王妈 moved 3→2 while its heat fell. Autumn collecting and fluffy-animal care suggest short-feedback tests, but popular topics and stories do not establish game demand. The shared design question is whether one interaction produces a visible change.

Hongguo: family-protection series · Hongguo: Wang Ma

Creator evidence: identity and time

Basketball Girls was observed on October 23 at 09:13:39 UTC, with 黑曼波工作室 credited as developer and 广州灵犀互动娱乐有限公司 as publisher. Lost Castle 2 was observed at 09:13:26 UTC; 雷霆游戏 is verified only as an associated vendor. 我不是胖虎：小岛大当家 remains 5.4/10 while ratings changed 297→320 between October 22 at 16:26 and October 23 at 08:52 UTC; CottonGame is likewise only a vendor credit. These belong to limited creator-linked coverage, not comparable cross-platform scores.

我不是胖虎：小岛大当家

Comparable ranks and conflicting signals

All comparisons use the same English/US entrance and first 100 positions. Bestseller observations are October 22 at 16:11 UTC and October 23 at 09:17 UTC, or October 23 at 00:11 and 17:17 China time. Upcoming and wishlist comparisons use adjacent valid snapshots of their respective entrances, with source times retained below. Earlier games leaving a list at release can move others upward; ranks do not measure wishlist growth.

Escape’s global-filtered reviews changed 650→924 and 58%→61% positive between October 22 at 16:20 and October 23 at 08:50 UTC. Bruisers 2D Boxing changed 123→181 and 78%→83% in its comparable scope, but was absent from this observed bestseller top 100. That is not delisting, zero attention or a sales conclusion. Review counts are not player counts or substitutes for normal-review evidence.

Groundwork Simulator · Bruisers 2D Boxing

Dates, scopes and coverage boundaries

The current combined input has 862 records: 812 Steam and 50 Baidu. Steam comprises 797 chart/date-list positions plus 15 automatic details: four charts of 100, a recent-date list of 298 and a future-date list of 99, all bounded coverage. Separately, retained recent-calendar evidence contains 850 displays and 848 distinct AppIDs; future evidence contains 325 displays, 300 within-entrance distinct IDs and 217 cross-entrance AppIDs. Combined metadata screening covers 1,766 distinct Steam AppIDs, not 1,766 detail studies or opportunities. Creator-linked coverage adds 143 game rows; TapTap remains a separate limited entrance.

Date windows remain separate: October 9–22 for the past 14 complete calendar days; October 23 as today; October 24–November 6 for the next 14 days. Retained recent evidence spans October 13–21, while the current input mostly covers October 22–23; October 9–12 remains missing. The current future input reaches only October 29, so later coverage is not complete. Demos, Early Access, full releases and updates are not interchangeable. Reusing an unchanged snapshot does not establish an unchanged market; original observation times are preserved.

Regular CNY prices, wishlist counts, sales and unpublished downloads remain unknown. Dollar conversions, rank and rating counts are not substitutes. An Indie tag does not establish team size. No game was launched: feel, mobile compatibility, ad placement/frequency, full economy values and code/asset licences remain unknown. Complete current global evidence was not obtained for TapTap, itch.io, Bilibili or Kuaishou. Limited creator coverage is not global recovery, and unobserved does not mean unchanged.

Ten editorial research candidates

The ten candidates are editorial priorities, not a cross-platform heat ranking. Screening covers 354 source positions, including 350 ranked ones: Baidu 50; Weibo 53 including one pinned and two promoted positions; Douyin 51 including one pinned position; Hongguo 80 positions across four lists, representing 60 series; and 120 retained Fanqie positions. Only 36 Fanqie titles are verified; 84 are unknown. Its statistical date remains October 22, and an evening date check does not create new data. Metrics are not added across platforms; one work appearing on multiple lists is not independent diffusion.

Upstream completed deep studies: zero; confirmed opportunities: zero. Normal demand comments and four-group supply searches are incomplete, so the candidates below remain hypotheses. Escalation requires the normal first 20 comments from each of two independent sources, first-ten queries for original terms, synonyms, scenes and actions, and named strongest released/announced substitutes. Then test one playable original-asset loop, separating click comprehension from actual completion. Stop if the promise is unplayable, substitutes already cover it without differentiation, or demand does not support the action.

Two alternatives per effective source

Three reading levels