.claude

.kimi

062026

a327ex-site

a327ex.com-blog

ai-assets

Anchor

Anchor-lua

anchor-site-archive

Anchor-YueScript

anchor.blot.im

Anchor2

artifact

artifact-opus

artifact-v2

BYTEPATH

cratebox

Danpenzero

data

edition-template

emoji-aimer

emoji-ball-battles

emoji-ball-battles-generated

emoji-ball-bounce

generation-loss

Invoker

Invoker-old

it-follows

lane-battle

ld54-remake

... [21 more lines]

[
  {
    "type": "text",
    "text": "I have a thorough picture now. All Python files that touch networking are either website tooling (`fetch_tweets.py`, `fetch_embeds.py`) or third-party `.venv` packages \u2014 none do Steam data collection. Here is the structured summary.\n\n---\n\n## Summary: a327ex \u2014 game-making & marketing philosophy, and existing Steam-data work\n\nSources are all under `E:\\a327ex\\a327ex-site\\` (his website `a327ex.com`). Older essays live in `posts\\`, short-form current thinking in `posts\\2026-*.md` (homepage \"messages\"), work-session writeups in `logs\\`, and long-form research in `E:\\a327ex\\reference\\` and `E:\\a327ex\\private\\`.\n\n### (a) How he approaches MAKING games\n\n- **\"Small long games\" is his signature thesis.** Short *development* time, long *player* time \u2014 an under-explored quadrant. \"Most indiedevs inherently assume that because a game is small, it has to be short. I reject this notion entirely.\" He calls these \"spacious\" games (BYTEPATH: 900-node skill tree, 40 classes, built in 4 months). Cite: `posts\\small_games.md`, and `posts\\lessons_second_game.md` (the big/small \u00d7 short/long chart).\n- **Time budget: 1\u20132 months per game, initial playable in \u22641 week.** \"The right duration is between 1 and 2 months\u2026 2 months [is] my limit\" before scope-creep. In hindsight on SNKRX: \"In the future I should probably take less than a month to finish the initial version\u2026 then have the next 6-7 weeks just for adding content + long hooks.\" Cite: `posts\\small_games.md`, `posts\\lessons_second_game.md`.\n- **He gravitates to \"lots of builds\" games** \u2014 roguelites, deckbuilders, auto-battlers, skill-tree games. \"I'm particularly interested in games where you can try lots of builds\u2026 as long as you can play it in lots of different ways.\" This is also what makes him overscope. Cite: `posts\\lessons_second_game.md`.\n- **\"Artifact accumulation\": reuse engine code across projects, never rewrite.** He settled on one way of doing things (Lua + L\u00d6VE) and documents reusable modules (e.g. his \"Nodemap\", timer/trigger `every(condition, action)` local-coding constructs). \"Everything I code for a project now I'll try to see if it can be made into reusable code.\" Cite: `posts\\lessons_second_game.md`.\n- **Stack is fixed: Lua + L\u00d6VE** (now his own \"Anchor\" framework, which exports to web). \"Stick with what works.\" Cite: `posts\\lessons_second_game.md`; framework/engine work in `logs\\anchor-*.md`.\n- **Planning beats grinding; follow intuition ruthlessly.** Wasted time came from not knowing the next task; he keeps a devlog. Deeper: he drops projects the instant his \"body\" removes his will to work \u2014 \"whenever my body thinks an idea for a game is bad, it will communicate this to me by removing my will to keep working on it.\" \"Doing things you don't want to do is incredibly bad for you\u2026 all the advice that says to grind is bullshit.\" Cite: `posts\\lessons_second_game.md`, `posts\\son_of_a_serpent.md`.\n- **Design values simplicity/transparency over opaque stat systems.** SNKRX's multiplicative %-stat system created an unwinnable \"you either pass the DPS/defense check or you don't\" problem; he wants integer, legible stats (Artifact-like) going forward. Cite: `posts\\lessons_second_game.md`, `posts\\snkrx_log.md`.\n- **The \"auto chess formula\" as a reusable engine of fun:** shop + rerolls + interest + set bonuses + unit upgrades, portable to any genre (TD, ARPG, city-builder, deckbuilder). He believes auto-battling itself is a *negative*; the 5 economy mechanics are the hook. Cite: `posts\\auto_chess_formula.md`.\n- **Self-assessed weaknesses:** poor discipline for long projects, chronic overscoping, and \u2014 per his own 2026 audit \u2014 building beautiful *infrastructure* instead of testing whether a game is fun. Cite: `logs\\gamedev-progress-assessment.md` (see (d)).\n\n### (b) His MARKETING approach & beliefs\n\n- **Marketing is a multiplier on base quality; fix quality first.** \"You should think of marketing as a multiplier on a game's base quality\u2026 very few indiedevs are making games of high base quality, and so that's not the thing they should focus on.\" For low-quality games, do only \"low effort high potential gains activities, like posting it on reddit, Hacker News.\" Cite: `posts\\marketing_skills.md`.\n- **Anti-wishlist-building (for small games).** Building wishlists for a mediocre game is \"a huge cope\"; the correct move is \"0 or 1\" (either minimal marketing on short games, or 1+ year of wishlist-building only if the game is genuinely high quality \u2014 avoid the \"toxic middleground of risk aversion\"). \"For these small releases all that matters is driving traffic to the store page on release day.\" Cite: `posts\\marketing_skills.md`, `posts\\lessons_second_game.md`.\n- **Screen composition / visual quality is the highest-ROI lever**, above marketing. A screenshot passes or fails a viewer's \"visual quality filter\" in 1\u20132 seconds, which drives shareability. He openly says his own games \"look like garbage.\" Cite: `posts\\marketing_skills.md`.\n- **Put up the Steam page early with a minimal trailer.** For a 1\u20133 month game: as soon as you have 1\u20132 min of gameplay, cut a simple trailer (more gameplay, fewer cuts \u2014 links Derek Lieu's \"simplest trailer\" article) and publish the Steam page week 1. Cite: `posts\\lessons_second_game.md`.\n- **Reddit >> Twitter for driving sales.** \"Twitter is notoriously poor at driving sales\u2026 sites like reddit are much better.\" Cite: `posts\\small_games.md`.\n- **Steam followers as a long-game compounding asset.** Every release emails all your followers \u2248 that many free wishlists. Inspired by dev \"L. Stotch.\" He accumulated 2,000+ followers by end of 2021 and planned an experiment: release a game *silently* to test follower conversion. He's bullish on Valve (\"greedy long\") and thinks most indies don't understand Steam's real value is **advertising/discovery**, not hosting. Cite: `posts\\lessons_second_game.md` (Steam followers section), `posts\\2022.md`, `posts\\son_of_a_serpent.md`.\n- **Post-launch: streamers/YouTubers >> everything for a graphically weak game.** SNKRX's success was driven almost entirely by roguelite YouTubers/streamers (Northernlion the biggest inflection). YouTube > Twitch for durable sales. Weekly updates gave a repeatable ~20\u201325% player bump; player retention + \"time played\" strongly affects Steam's discovery-queue placement. Cite: `posts\\snkrx_log.md`.\n- **\"Good = popular\"; the market is the only objective judge of quality.** He's dismissive of \"indiepocalypse\"/luck narratives \u2014 \"luck isn't real.\" Reject portfolio/lottery framing for making many games. Cite: `posts\\son_of_a_serpent.md`, `posts\\small_games.md`, `posts\\1_in_4.md`, `posts\\luck.md`.\n- **Learn from SUCCESSES, not failures** (failure is over-determined; multi-causal, low signal). Even a creator can't reliably diagnose *why* their own game succeeded (he initially credited SNKRX to the auto-chess formula; Vampire Survivors later revealed it was the **auto-shooting**). Cite: `posts\\2026-05-09-174843.md`, quoting `posts\\son_of_a_serpent.md`.\n\n**Stated revenue / success numbers (load-bearing):**\n- **BYTEPATH** (1st game, 4 months): ~$12K / ~10K units over 3 years. Cite: `posts\\lessons_second_game.md`, `posts\\small_games.md`.\n- **SNKRX** (2nd game, 3 months): launched with only ~200 wishlists. Day 55 = **80,781 units / $234,953**; he wrote it would \"make enough to last me multiple decades if not the rest of my life.\" Detailed daily units/revenue/wishlist tables in `posts\\snkrx_log.md`.\n- He lives in **Brazil**, so USD revenue is multiplied ~5\u00d7 in local terms; his bar for \"making a living\" was roughly a BYTEPATH-tier game every ~2 months. Cite: `posts\\lessons_second_game.md`, `posts\\small_games.md`.\n\n### (c) Existing Steam data-collection work\n\n**No scripts and no datasets exist.** There is exactly **one** Steam-data artifact, and it is a hand-compiled prose research document (not code, not a CSV):\n\n- `E:\\a327ex\\reference\\experimental-visual-games-steam-research.md` \u2014 a ~27KB manually-written market-research report titled \"Experimental Visual / Narrative Games on Steam,\" compiled to evaluate \"a scrolling web-like visual experience with AI-generated imagery\" on Steam. Contains per-game stat lines (developer, price, length, review count/%, SteamSpy owner buckets, estimated gross/net revenue, tags) for ~40 games (the \"Milk\" games, KIDS, Florence, Stanley Parable, GRIS, Inscryption, etc.), an AI-art-on-Steam landscape section, a price\u00d7length sweet-spot table, and revenue-tier expectations. Numbers were pulled by hand from SteamSpy / steam-revenue-calculator / news articles (see its Sources list) \u2014 **there is no scraper or stored raw data behind it.**\n\nChecked and ruled out:\n- No files reference `steamspy`, `api.steampowered`, `store.steampowered`, `appdetails`, or `appreviews` as data-collection code \u2014 the only hits are Steam **store links** in SNKRX's game source (`SNKRX\\arena.lua`, `SNKRX\\buy_screen.lua`, `SNKRX\\README.md`) and the word \"scrape\" in narrative fiction (`it-follows\\it_follows.txt`).\n- No relevant CSVs anywhere (only numpy test fixtures inside `ai-assets\\.venv\\`).\n- The Python tooling that does HTTP (`a327ex-site\\renderer\\tools\\fetch_tweets.py`, `fetch_embeds.py`, `fetch_emoji.py`, `artifact-opus\\tools\\build_carddb.py`) is website/asset tooling \u2014 none touch Steam.\n\nOther research docs in `E:\\a327ex\\reference\\` are game-**design** genealogies (`td-genealogy-and-card-td.md`, `star-of-providence-research.md`, `noita-wand-composition-system.md`, `poe2-meta-modifier-mechanics.md`, `emergent-exam-genealogy.md`), not Steam market data. `a327ex-site\\logs\\steam-game-recommendations.md` is a \"what should I play\" log, not data. So a Steam-market-research project would be **greenfield** \u2014 the `reference\\` folder is where prior manual research lives and where such work would naturally sit.\n\n### (d) Other load-bearing context for planning a Steam market-research project\n\n- **His current 2026 direction has shifted toward \"meaning and commitment.\"** He's now driven by *mixing stories and games* \u2014 writing books / deep lore worlds and making games set in them \u2014 explicitly *against* \"free dopamine generators\" (VS-likes, skill-tree incrementals, \"choose 1 of 3\", casino-juice UI). He favors mechanics that force **commitment over optionality**: drafts, \"frontloaded power distribution,\" diegetic UIs, deep skill trees, \"10000 actually-viable builds.\" Cite: `posts\\2026-05-03-203538.md`, `posts\\2026-05-05-121038.md`, `posts\\2026-05-21-163053.md`. Concrete current concept: **\"Concentration of Will\"**, a roguelite that morphs from a Vampire-Survivors-like into a Dark-Souls-like \u2014 `E:\\a327ex\\private\\vs-to-souls-concentration-of-will.md`.\n- **Genre preferences to weight in any market research:** roguelites, deckbuilders, auto-battlers/auto-chess, \"build-craft\" systems (Noita/PoE-style), and now short narrative/experimental/\"artware\" games. He references Balatro, Vampire Survivors, Artifact, Path of Exile, Slay the Spire, Inscryption constantly.\n- **Art is his hard constraint.** He's \"not an artist\"; games \"look like garbage\" and fail to pass visual filters pre-launch. He'll only do a 50/50 revenue-share with an artist, never flat-fee (`posts\\2022.md`). He is pro-AI-art (unusually for an indie), which is why the AI-visual-experience research doc exists \u2014 relevant if scoping research around AI-assisted art viability on Steam. Cite: `posts\\marketing_skills.md`, `posts\\2026-05-21-163053.md`, `reference\\experimental-visual-games-steam-research.md`.\n- **Time/scope budget for planning:** he thinks in 1\u20132 month cycles, wants initial-playable in ~1 week, and targets **web builds on itch.io / a327ex.com** as the lowest-friction release (his Anchor engine exports to web) \u2014 Steam is the eventual/monetization target, not the prototype target. Cite: `logs\\gamedev-progress-assessment.md`.\n- **His self-diagnosed failure mode is directly relevant to how research should be framed.** A May-2026 audit of ~159 work logs (`logs\\gamedev-progress-assessment.md`) concludes he is an excellent *local optimizer* but avoids the *fun measurement* by endlessly building tooling/infrastructure (\"Anchor\" engine, \"snkrx-template\", \"ricochet-template\", ability systems) so nothing ever gets tested against players. He has a self-imposed \"**ship \u22651 game in 2026**\" gate and a rule to hardcode a fake end-game to test the fantasy *before* building systems. Implication for a market-research deliverable: he responds to **evidence-grounded, market-validated** framing (he trusts \"the market\" as truth) and is wary of anything that becomes another infinite infrastructure project.\n- **Tooling/monetization curiosity:** long-standing interest in crypto-wallet integration for global payments and paying players directly (`posts\\2022.md`), and skepticism of relying solely on Steam \u2014 though he's very pro-Valve. Minor, but relevant if research touches distribution/monetization strategy.\n\nIf you want, I can pull the exact SNKRX daily revenue/units/wishlist tables from `posts\\snkrx_log.md` (it continues past line 534, days 55\u2192lifetime) or read `posts\\game_quality.md` / `posts\\high_level_game_design.md` / `posts\\competition.md`, which I identified as relevant but did not fully quote here."
  }
]

[
  {
    "type": "text",
    "text": "Chapter marked: \"Data infrastructure setup\". Continue your current work."
  }
]

(Bash completed with no output)

{"730":{"appid":730,"name":"Counter-Strike: Global Offensive","developer":"Valve","publisher":"Valve","score_rank":"","positive":7642084,"negative":1173003,"userscore":0,"owners":"100,000,000 .. 200,000,000","average_forever":0,"average_2weeks":0,"median_forever":0,"median_2weeks":0,"price":"0","initialprice":"0","discount":"0","ccu":1013936},
"1172470":{"appid":1172470,"name":"Apex Legends","developer":"Respawn","publisher":"Electronic Arts","score_rank":"","positive":668053,"negative":326926,"userscore":0,"owners":"100,000,000 .. 200,000,000","average_forever":0,"average_2weeks":0,"median_forever":0,"median_2weeks":0,"price":"0","initialprice":"0","discount":"0","ccu":124262},
"578080":{"a

{"success":1,"results_html":"\n<!-- List Items -->\n\n\t\t\t\t<a href=\"https:\/\/store.steampowered.com\/app\/4841480\/Chicken_Battlefield\/?snr=1_7_7_230_150_1\"\n\t\t\t data-ds-appid=\"4841480\" data-ds-itemkey=\"App_4841480\" data-ds-tagids=\"[19,597,1773,4158,1664,4736,1645]\" onmouseover=\"GameHover( this, event, 'global_hover', {&quot;type&quot;:&quot;app&quot;,&quot;id&quot;:4841480,&quot;public&quot;:1,&quot;v6&quot;:1} );\" onmouseout=\"HideGameHover( this, event, 'global_hover' )\" class=\"search_result_row ds_collapse_flag \"\n\t\t   data-search-page=\"1\" data-gpnav=\"item\"\n\t\t   data-ds-steam-deck-compat-handled=\"true\">\n            <div class=\"search_capsule\"><img src=\"https:\/\/shared.akamai.steamstatic.com\/store_item_assets\/steam\/apps\/4841480\/aa8834378eaaddf055bcd0dd0a18a7e052b20761\/capsule_231x87.jpg?t=1783424542\" ><\/div>\n            <div class=\"responsive_search_name_combined\">\n                <div class=\"search_name ellipsis\">\n                    <span class=\"title\">Chicken Battlefield<\/span>\n                <\/div>\n\t\t\t\t<div class=\"search_platforms\">\n\t\t\t\t\t<span class=\"platform_img win\"><\/span><span class=\"platform_img mac\"><\/span>\t\t\t\t<\/div>\n                <div class=\"search_released responsive_secondrow\">\n                    Jul 7, 2026                <\/div>\n                <div class=\"search_reviewscore responsive_secondrow\">\n                                    <\/div>\n\n\n                <div class=\"search_price_discount_combined responsive_secondrow\" data-price-final=\"479\">\n                    <div class=\"search_discount_and_price responsive_secondrow\">\n                        <div class=\"discount_block search_discount_block\" data-price-final=\"479\" data-bundlediscount=\"0\" data-discount=\"40\" role=\"link\" aria-label=\"40% off. $7.99 normally, discounted to $4.79\"><div class=\"discount_pct\">-40%<\/div><div class=\"discount_prices\"><div class=\"discount_original_price\">$7.99<\/div><div class=\"discount_final_price\">$4.79<\/div><\/div><\/div>                    <\/div>\n                <\/div>\n            <\/div>\n\n        <\/a>\n            \n\t\t\t\t<a href=\"https:\/\/store.steampowered.com\/app\/4851270\/The_Hollow_House\/?snr=1_7_7_230_150_1\"\n\t\t\t data-ds-appid=\"4851270\" data-ds-itemkey=\"App_4851270\" data-ds-tagids=\"[21,4106,3834,9204,1100686,4191,4342]\" data-ds-descids=\"[5]\" onmouseover=\"GameHover( this, event, 'global_hover', {&quot;t

{"915310":{"success":true,"data":{"type":"game","name":"SNKRX","steam_appid":915310,"required_age":0,"is_free":false,"detailed_description":"<strong>SNKRX<\/strong> is an arcade shooter roguelite where you control a snake made of multiple heroes, with each one having their own attacks, passives and classes. Inspired by the auto-battler genre, each hero in the snake has a set of classes, and combining enough heroes of the same class together grants additional class bonuses. Heroes can also be upgraded to be more powerful when copies are bought from the shop.<h2 class=\"bb_tag\" ><strong>Gameplay<\/strong><\/h2><ul class=\"bb_ul\"><li>Your snake can't stop moving, turn it left or right to steer<br><\/li><li>Your heroes attack automatically when close to enemies<br><\/li><li>Combine heroes of the same classes together to unlock unique class bonuses<br><\/li><li>Clearing arenas grants gold,

{"success":1,"query_summary":{"num_reviews":0,"review_score":8,"review_score_desc":"Very Positive","total_positive":4097,"total_negative":236,"total_reviews":4333},"reviews":[],"cursor":"*"}
{"success":1,"results":{"start_date":1621209600,"end_date":1783036800,"weeks":[],"rollups":[{"date":1619827200,"recommendations_up":115,"recommendations_down":2},{"date":1622505600,"recommendations_up":1343,"recommendations_down":34},{"date":1625097600,"recommendations_up":814,"recommendations_down":16},{"date":1627776000,"recommendations_up":222,"recommendations_down":9},{"date":1630454400,"recommendations_up":144,"recommendations_down":5},{"date":1633046400,"recommendations_up":108,"recommendations_down":2},{"date":1635724800,"recommendations_up":432,"recommendations_down":3},{"date":1638316

Creating virtual environment at: .venv
Activate with: .venv\Scripts\activate
 + six==1.17.0
 + tzdata==2026.2
 + urllib3==2.7.0

[{"tagid":492,"name":"Indie"},{"tagid":19,"name":"Action"},{"tagid":21,"name":"Adventure"},{"tagid":597,"name":"Casual"},{"tagid":4182,"name":"Singleplayer"},{"tagid":599,"name":"Simulation"},{"tagid":122,"name":"RPG"},{"tagid":9,"name":"Strategy"},{"tagid":3871,"name":"2D"},{"tagid":493,"name":"Ear
432

File created successfully at: E:\a327ex\steam-market\scripts\steamspy_all.py (file state is current in your context — no need to Read it back)

File created successfully at: E:\a327ex\steam-market\scripts\search_scrape.py (file state is current in your context — no need to Read it back)

Command running in background with ID: bpk62lati. Output is being written to: C:\Users\adn\AppData\Local\Temp\claude\E--a327ex\689cefaa-c928-4524-85fc-f8dfbde9fd7f\tasks\bpk62lati.output. You will be notified when it completes. To check interim output, use Read on that file path.

Command running in background with ID: b9snb1imr. Output is being written to: C:\Users\adn\AppData\Local\Temp\claude\E--a327ex\689cefaa-c928-4524-85fc-f8dfbde9fd7f\tasks\b9snb1imr.output. You will be notified when it completes. To check interim output, use Read on that file path.

File created successfully at: E:\a327ex\steam-market\README.md (file state is current in your context — no need to Read it back)

search rows: 750

(439730, 'XO-Planets', 'Jun 30, 2026', '42804,5379,5125,5363,3964,3841,3916', 'Mostly Positive', 70, 10, 69, 199, 65, 'win')

(450540, 'Hot Dogs, Horseshoes & Hand Grenades', 'Jul 4, 2026', '3810,21978,1774,3968,1663,599,5765', 'Overwhelmingly Positive', 97, 19538, 1999, 1999, 0, 'win')

(548800, 'OFS', 'Jul 3, 2026', '19,4182,597,1773,8666,16689,3839', None, None, None, 374, 499, 25, 'win')

(873260, 'Wizard Gold', 'Jul 5, 2026', '9,560542,597,615955,3871,1654,3964', None, None, None, 599, 599, 0, 'win')

(978280, 'Damn Zombies', 'Jul 3, 2026', '19,1663,4667,1774,5547,8666,4777', None, None, None, 600, 600, 0, 'win')

steamspy rows: 1000

(10, 'Counter-Strike', 'Valve', 243818, 6427, '10,000,000 .. 20,000,000', 199, 7323)

(30, 'Day of Defeat', 'Valve', 6414, 688, '5,000,000 .. 10,000,000', 99, 87)

(40, 'Deathmatch Classic', 'Valve', 2618, 545, '5,000,000 .. 10,000,000', 99, 7)

File created successfully at: E:\a327ex\steam-market\scripts\histogram_crawl.py (file state is current in your context — no need to Read it back)

File created successfully at: E:\a327ex\steam-market\scripts\appdetails_crawl.py (file state is current in your context — no need to Read it back)

[
  {
    "type": "tool_reference",
    "tool_name": "TaskCreate"
  },
  {
    "type": "tool_reference",
    "tool_name": "TaskUpdate"
  },
  {
    "type": "tool_reference",
    "tool_name": "TaskList"
  },
  {
    "type": "tool_reference",
    "tool_name": "TaskStop"
  },
  {
    "type": "tool_reference",
    "tool_name": "TaskOutput"
  }
]

Command running in background with ID: b1fh64tu8. Output is being written to: C:\Users\adn\AppData\Local\Temp\claude\E--a327ex\689cefaa-c928-4524-85fc-f8dfbde9fd7f\tasks\b1fh64tu8.output. You will be notified when it completes. To check interim output, use Read on that file path.

---
{"appid":915310,"name":"SNKRX","developer":"a327ex","publisher":"a327ex","score_rank":"","positive":4023,"negative":227,"userscore":0,"owners":"50,000 .. 100,000","average_forever":0,"average_2weeks":0,"median_forever":0,"median_2weeks":0,"price":"83","initialprice":"299","discount":"72","ccu":10,"languages":"English","genre":"Action, Casual, Indie","tags":{"Action Roguelike":314,"Rogue-lite":305,

1    /usr/bin/bash: line 1: .venv/Scripts/python.exe: No such file or directory
2   

Command running in background with ID: bnss2y62i. Output is being written to: C:\Users\adn\AppData\Local\Temp\claude\E--a327ex\689cefaa-c928-4524-85fc-f8dfbde9fd7f\tasks\bnss2y62i.output. You will be notified when it completes. To check interim output, use Read on that file path.

{"760330":{"success":true,"data":{"type":"game","name":"BYTEPATH","steam_appid":760330,"required_age":0,"is_free":false,"detailed_description":"BYTEPATH is a replayable arcade shooter with a focus on varied play styles with RPG elements. The game has a huge passive skill tree, classes and different

Task #1 created successfully: Complete skeleton crawls (SteamSpy + search index)

Task #2 created successfully: Build in-house revenue estimator (SNKRX/BYTEPATH calibration)

Task #3 created successfully: Report 1: 2021→2026 delta map (base rates)

Task #4 created successfully: Report 2: tag-space supply/demand scan

Task #5 created successfully: Report 3: trajectory taxonomy (long-tail niches)

Task #6 created successfully: Report 4: success-anomaly case files

Task #7 created successfully: Report 5: art-quality gate analysis (capsule vision)

Task #8 created successfully: Per-language review crawl + language-gap analysis

Task #9 created successfully: Final synthesis: openings + recommendations

Updated task #1 status

Updated task #2 status

total 22
drwxr-xr-x 1 adn 197121    0 Jul  7 09:26 .
drwxr-xr-x 1 adn 197121    0 Jul  7 09:26 ..
-rw-r--r-- 1 adn 197121 7726 Jul  7 09:26 bytepath_histogram.json
-rw-r--r-- 1 adn 197121  187 Jul  7 09:26 bytepath_reviews.json
-rw-r--r-- 1 adn 197121 6554 Jul  7 09:26 snkrx_histogram.json
-rw-r--r-- 1 adn 197121  190 Jul  7 09:26 snkrx_reviews.json

No matches found

15:| Game | Units Sold | Total Revenue | Wishlists |
28:| Game | Units Sold | Total Revenue | Wishlists |
53:| Game | Units Sold | Total Revenue | Wishlists |
134:[Omitted long matching line]
142:| Game | Units Sold | Total Revenue | Wishlists |
151:SNKRX is now doing 4.6~ times as well as BYTEPATH, making it go to about $55K over 3 years, equivalent to ~R$140K. I’m not sure if this math even makes sense anymore as SNKRX has now surpassed BYTEPATH’s lifetime revenue, which is about $12K. It just seems like its own thing and the comparisons with BYTEPATH are probably not that helpful anymore.
225:| Game | Units Sold | Total Revenue | Wishlists |
265:| Game | Units Sold | Total Revenue | Wishlists |
349:| Game | Units Sold | Total Revenue | Wishlists |
543:[Omitted long matching line]
551:Despite everything, difficulty complaints are mostly fair and I have a plan to deal with them, which is the game’s roguelite update. Essentially, an easy way that will allow me more freedom in balancing the game is to make it so that units die permanently when they die. Right now when units die they get revived, which makes me have to do fairly strong enemies that can 1 shoot you to keep things interesting.
553:If units die permanently then enemies can be a lot less threatening individually, and since I’ll have more time to design them I can also be more careful to not design unfair encounters. This coupled with the fact that there will be an actual roguelite run happening, the chances that the player will blame the game will be highly lowered.

10   
11  This article is also being updated over time and continues from the “financial results” section of the [lessons learned article](/posts/lessons_second_game).
12  
13  ## Day 3 (21/05/19)
14  
15  | Game | Units Sold | Total Revenue | Wishlists |
16  | --- | --- | --- | --- |
17  | BYTEPATH | 1283 | $2634 | 1238 |
18  | SNKRX | 601 | $1429 | 788 |
19  
20  ::image /media/posts/snkrx_log/_snkrx_log_1.png
21  
22  ::image /media/posts/snkrx_log/_snkrx_log_2.png
23  
24  Based on these numbers the updated projection for SNKRX’s performance is that it will do about half as well as BYTEPATH, which is an improvement from 1/3rd as well. This improves it from $4K over 3 years to $6K over 3 years, which is equivalent to about R$15K rather than R$10K.
25  
26  ## Day 7 (21/05/23)
27  
28  | Game | Units Sold | Total Revenue | Wishlists |
29  | --- | --- | --- | --- |
30  | BYTEPATH | 1534 | $3129 | 1585 |
31  | SNKRX | 1064 | $2534 | 1131 |
32  
33  ::image /media/posts/snkrx_log/_snkrx_log_3.png
34  
35  ::image /media/posts/snkrx_log/_snkrx_log_4.png
36  
37  Now SNKRX is doing about 0.8 as well as BYTEPATH, which makes it go to ~10K over 3 years, equivalent to R$25K. This is significantly better than what it started with on day 1\. I have no idea why the game keeps improving in performance.
38  
39  Yesterday it got a fairly good [Hacker News post](https://news.ycombinator.com/item?id=27254177) but that only improved its performance for that day by about 2X, which is not a lot. This HN post was similar in views to the most popular [social media post for BYTEPATH](https://www.reddit.com/r/pathofexile/comments/7zu9vz/i_made_a_small_game_inspired_by_poes_passive/) (about 30K total views), but it didn’t generate nearly as many sales because it wasn’t so focused on the specific audience the game is made for and it also links to a GitHub page rather than the Steam page.
... [35 more lines]

13:## Day 3 (21/05/19)
17:| BYTEPATH | 1283 | $2634 | 1238 |
18:| SNKRX | 601 | $1429 | 788 |
26:## Day 7 (21/05/23)
30:| BYTEPATH | 1534 | $3129 | 1585 |
31:| SNKRX | 1064 | $2534 | 1131 |
51:## Day 14 (21/05/30)
55:| BYTEPATH | 1712 | $3488 | 1839 |
56:| SNKRX | 2607 | $6944 | 2228 |
140:## Day 20 (21/06/05)
144:| BYTEPATH | 1825 | $3716 | 2059 |
145:| SNKRX | 6137 | $17290 | 3477 |
223:## Day 23 (21/06/08)
227:| SNKRX | 10249 | $29535 | 4797 |
263:## Day 31 (21/06/16)
267:| SNKRX | 32046 | $94216 | 9391 |
347:## Day 55 (21/07/09)
351:| SNKRX | 80781 | $234953 | 18900 |

File created successfully at: E:\a327ex\steam-market\scripts\calibrate.py (file state is current in your context — no need to Read it back)

SNKRX 2021-05-19: 601 units, $1429, ~68 reviews -> 8.8 u/r, $21/review, k=7.0

SNKRX 2021-05-23: 1064 units, $2534, ~83 reviews -> 12.8 u/r, $31/review, k=10.2

SNKRX 2021-05-30: 2607 units, $6944, ~109 reviews -> 23.8 u/r, $63/review, k=21.2

SNKRX 2021-06-05: 6137 units, $17290, ~301 reviews -> 20.4 u/r, $58/review, k=19.2

SNKRX 2021-06-08: 10249 units, $29535, ~438 reviews -> 23.4 u/r, $67/review, k=22.5

SNKRX 2021-06-16: 32046 units, $94216, ~806 reviews -> 39.8 u/r, $117/review, k=39.1

SNKRX 2021-07-09: 80781 units, $234953, ~1708 reviews -> 47.3 u/r, $138/review, k=46.0

BYTEPATH 2021-02-15: 10000 units, $12000, ~233 reviews -> 42.9 u/r, $52/review, k=25.9



SteamSpy owners-mid/reviews over 9234 games: p25=67 p50=129 p75=274



estimator: revenue ~= reviews * 46.0 * price_usd (band [25, 60])

The file E:\a327ex\steam-market\scripts\calibrate.py has been updated successfully. (file state is current in your context — no need to Read it back)

SteamSpy owners-mid/reviews over 10023 games: p25=68 p50=130 p75=277



estimator: revenue ~= reviews * 35.9 * price_usd (band [25, 55])

Updated task #2 status

File created successfully at: E:\a327ex\steam-market\scripts\common.py (file state is current in your context — no need to Read it back)

File created successfully at: E:\a327ex\steam-market\scripts\delta_map.py (file state is current in your context — no need to Read it back)

Traceback (most recent call last):

  File "E:\a327ex\steam-market\scripts\delta_map.py", line 231, in <module>

    main()

    ~~~~^^

  File "E:\a327ex\steam-market\scripts\delta_map.py", line 228, in main

    print(out[:3000])

    ~~~~~^^^^^^^^^^^^

  File "C:\Users\adn\AppData\Roaming\uv\python\cpython-3.13-windows-x86_64-none\Lib\encodings\cp1252.py", line 19, in encode

    return codecs.charmap_encode(input,self.errors,encoding_table)[0]

           ~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

UnicodeEncodeError: 'charmap' codec can't encode character '\u2192' in position 15: character maps to <undefined>

The file E:\a327ex\steam-market\scripts\common.py has been updated successfully. (file state is current in your context — no need to Read it back)

# 01 — The 2021→2026 Delta Map (base rates)



Generated 2026-07-07 12:31 UTC. Data coverage: search=17698, steamspy=12910, appdetails=176, histograms=0.



Revenue estimator: `revenue ≈ reviews × 35.9 × base_price` (band (25, 55)), calibrated in-house on SNKRX + BYTEPATH. Caveats at bottom.



## 1. Release volume by year



| year | releases | f2p% | med price | med price (reviewed≥10) |

|---|---|---|---|---|

| 2025 | 5773 | 14% | $5.99 | $8.99 |

| 2026 | 11925 | 14% | $5.99 | $8.99 |



### Monthly releases 2024–2026



| month | 2024 | 2025 | 2026 |

|---|---|---|---|

| 01 | 0 | 0 | 1476 |

| 02 | 0 | 0 | 1608 |

| 03 | 0 | 0 | 2296 |

| 04 | 0 | 0 | 2171 |

| 05 | 0 | 0 | 1891 |

| 06 | 0 | 0 | 2133 |

| 07 | 0 | 0 | 350 |

| 08 | 0 | 0 | 0 |

| 09 | 0 | 320 | 0 |

| 10 | 0 | 1843 | 0 |

| 11 | 0 | 1975 | 0 |

| 12 | 0 | 1635 | 0 |



... [15 more lines]

'Mostly Positive<br>70% of the 17 user reviews for this game are positive.'

'Mostly Positive<br>78% of the 28 user reviews for this game are positive.'

'Positive<br>97% of the 34 user reviews for this game are positive.'

'Positive<br>100% of the 11 user reviews for this game are positive.'

'Positive<br>100% of the 18 user reviews for this game are positive.'

'Very Positive<br>94% of the 76 user reviews for this game are positive.'

'Positive<br>84% of the 13 user reviews for this game are positive.'

'Mixed<br>68% of the 166 user reviews for this game are positive.'

'Mostly Positive<br>73% of the 67 user reviews for this game are positive.'

'Positive<br>100% of the 10 user reviews for this game are positive.'

'Positive<br>100% of the 13 user reviews for this game are positive.'

'Mostly Positive<br>70% of the 10 user reviews for this game are positive.'

rows: 50 tooltips: 15

<tool_use_error>String to replace not found in file.
String:         L.append('| year | paid n | 0 rev | 1–9 | 10–49 | 50–199 | 200–999 | 1000+ | '
             '≥$10K | ≥$50K | ≥$100K | med $ (≥10rev) | cohort $M |')
    L.append('|---|---|---|---|---|---|---|---|---|---|---|---|---|')
(note: Edit also tried swapping \uXXXX escapes and their characters; neither form matched, so the mismatch is likely elsewhere in old_string. Re-read the file and copy the exact surrounding text.)</tool_use_error>

The file E:\a327ex\steam-market\scripts\delta_map.py has been updated successfully. (file state is current in your context — no need to Read it back)

The file E:\a327ex\steam-market\scripts\delta_map.py has been updated successfully. (file state is current in your context — no need to Read it back)

Exit code 1
tail: option used in invalid context -- 3

The file E:\a327ex\steam-market\scripts\delta_map.py has been updated successfully. (file state is current in your context — no need to Read it back)

Exit code 1
== b9snb1imr
tail: cannot open 'C:\Users\adn\AppData\Local\Temp\claude\E--a327ex\689cefaa-c928-4524-85fc-f8dfbde9fd7f\tasks$f.output' for reading: No such file or directory
== bpk62lati
tail: cannot open 'C:\Users\adn\AppData\Local\Temp\claude\E--a327ex\689cefaa-c928-4524-85fc-f8dfbde9fd7f\tasks$f.output' for reading: No such file or directory
== bnss2y62i
tail: cannot open 'C:\Users\adn\AppData\Local\Temp\claude\E--a327ex\689cefaa-c928-4524-85fc-f8dfbde9fd7f\tasks$f.output' for reading: No such file or directory

== b9snb1imr
start 17500/115447: 17498 unique games so far

start 20000/115447: 19998 unique games so far

== bpk62lati
page 13: +1000 rows (total 13903)

page 14: +1000 rows (total 14857)

== bnss2y62i
queue: 3750 reviewed + 5999 sampled low-review + 0 tail

File created successfully at: E:\a327ex\steam-market\scripts\tag_scan.py (file state is current in your context — no need to Read it back)

baseline: {'n': 8585, 'h10': 0.24729178800232965, 'h50': 0.13581828771112406, 'h100': 0.10110658124635993, 'med': 13635.179, 'medp': 5.99, 'h50_lowprice': 0.04884357881923311, 'n_lowprice': 6572}

220 tags, 971 pairs, 64 mechanical openings

written to reports/02-tag-scan.md

File created successfully at: E:\a327ex\steam-market\scripts\trajectory.py (file state is current in your context — no need to Read it back)

File created successfully at: E:\a327ex\steam-market\scripts\reviews_sample_crawl.py (file state is current in your context — no need to Read it back)

== b9snb1imr
start 25000/115449: 24996 unique games so far

== bpk62lati
page 17: +1000 rows (total 17677)

== bnss2y62i
queue: 3750 reviewed + 5999 sampled low-review + 0 tail

search 25146

steamspy 17677

appdetails 339

Exit code 127
start 112500/115452: 112422 unique games so far

start 115000/115452: 114844 unique games so far

search crawl COMPLETE: 115281 unique games (reported total_count=115452)

/usr/bin/bash: line 7: .venv/Scripts/python.exe: No such file or directory

{"message":"Successfully stopped task: bnss2y62i (cd /e/a327ex/steam-market && .venv/Scripts/python.exe scripts/appdetails_crawl.py)","task_id":"bnss2y62i","task_type":"local_bash","command":"cd /e/a327ex/steam-market && .venv/Scripts/python.exe scripts/appdetails_crawl.py"}

unique games: 115281

with review data (>=10): 67452

reviews>=20: 51526

reviews>=50: 34654

reviews>=200: 18230

The file E:\a327ex\steam-market\scripts\histogram_crawl.py has been updated successfully. (file state is current in your context — no need to Read it back)

The file E:\a327ex\steam-market\scripts\appdetails_crawl.py has been updated successfully. (file state is current in your context — no need to Read it back)

Command running in background with ID: bb1zd8vmd. Output is being written to: C:\Users\adn\AppData\Local\Temp\claude\E--a327ex\689cefaa-c928-4524-85fc-f8dfbde9fd7f\tasks\bb1zd8vmd.output. You will be notified when it completes. To check interim output, use Read on that file path.

Command running in background with ID: b3sgn0kyj. Output is being written to: C:\Users\adn\AppData\Local\Temp\claude\E--a327ex\689cefaa-c928-4524-85fc-f8dfbde9fd7f\tasks\b3sgn0kyj.output. You will be notified when it completes. To check interim output, use Read on that file path.

Command running in background with ID: bwpyocf2n. Output is being written to: C:\Users\adn\AppData\Local\Temp\claude\E--a327ex\689cefaa-c928-4524-85fc-f8dfbde9fd7f\tasks\bwpyocf2n.output. You will be notified when it completes. To check interim output, use Read on that file path.

== bb1zd8vmd
queue: 34654 appids (min_reviews=50)

== b3sgn0kyj
queue: 5596 window-successes + 4000 window-failures + 3000 window-mid + 56553 reviewed rest

== bwpyocf2n
mode=sub10 queue=1500

Updated task #3 status

# 01 — The 2021→2026 Delta Map (base rates)



Generated 2026-07-07 13:42 UTC. Data coverage: search=115281, steamspy=78688, appdetails=2348, histograms=22.



Revenue estimator: `revenue ≈ reviews × 35.9 × base_price` (band (25, 55)), calibrated in-house on SNKRX + BYTEPATH. Caveats at bottom.



## 1. Release volume by year



| year | releases | f2p% | med price | med price (reviewed≥10) |

|---|---|---|---|---|

| 2014 | 1439 | 6% | $7.99 | $7.99 |

| 2015 | 2372 | 8% | $6.99 | $6.99 |

| 2016 | 3903 | 11% | $5.99 | $6.99 |

| 2017 | 5554 | 12% | $4.99 | $6.99 |

| 2018 | 6928 | 14% | $4.99 | $5.99 |

| 2019 | 6578 | 15% | $4.99 | $8.99 |

| 2020 | 8015 | 17% | $4.99 | $7.99 |

| 2021 | 9403 | 16% | $4.99 | $7.99 |

| 2022 | 10163 | 18% | $5.99 | $8.99 |

| 2023 | 11920 | 18% | $5.99 | $8.99 |

| 2024 | 16200 | 19% | $4.99 | $7.99 |

| 2025 | 19080 | 15% | $5.99 | $8.99 |

| 2026 | 11927 | 14% | $5.99 | $8.99 |



### Monthly releases 2024–2026



| month | 2024 | 2025 | 2026 |

|---|---|---|---|

| 01 | 1068 | 1298 | 1476 |

| 02 | 1250 | 1231 | 1608 |

... [38 more lines]

baseline: {'n': 39184, 'h10': 0.25854940792160064, 'h50': 0.14281339322172315, 'h100': 0.10751837484687628, 'med': 13256.434, 'medp': 5.0, 'h50_lowprice': 0.0482470896978669, 'n_lowprice': 29722}

335 tags, 4573 pairs, 298 mechanical openings

written to reports/02-tag-scan.md

1    # 02 — Tag-space supply/demand scan
2   
3   Generated 2026-07-07 13:43 UTC. Coverage: search=115281, steamspy=78688.
4   
5   Outcome window: paid games released 2023–2025 (n=39184). Supply trend = tag growth 2024–2026 vs 2020–2022, normalized by market growth (1.73×); 1.0 = grew with market.
6   
7   **Market baseline:** ≥$10K 25.9%, ≥$50K 14.28%, ≥$100K 10.75%, median $ of reviewed games $13K.
8   
9   
10  ## Tags ranked by P(≥$50K) — top 80
11  
12  _tags with ≥40 releases in window; 335 qualify_
13  
14  | tag | n | ≥$10K | ≥$50K | ≥$100K | med $ (rev’d) | med price | ≥$50K @≤$10 | supply trend |
15  |---|---|---|---|---|---|---|---|---|
16  | Open World Survival Craft | 161 | 71% | 54.7% | 46.0% | $130K | $15 | 10.6% (n=47) | 0.77 |
17  | FMV | 174 | 73% | 53.4% | 40.2% | $80K | $10 | 43.2% (n=111) | 1.51 |
18  | Mechs | 109 | 55% | 44.0% | 35.8% | $111K | $11 | 17.0% (n=53) | 0.88 |
19  | Turn-Based | 203 | 59% | 41.9% | 32.0% | $63K | $14 | 9.6% (n=83) | 0.57 |
20  | Sexual Content | 477 | 69% | 40.3% | 27.3% | $43K | $10 | 24.7% (n=255) | 0.49 |
21  | Naval Combat | 45 | 51% | 40.0% | 35.6% | $142K | $13 | – | 0.75 |
22  | Naval | 41 | 49% | 39.0% | 29.3% | $118K | $15 | – | 0.85 |
23  | Shop Keeper | 82 | 60% | 39.0% | 30.5% | $63K | $10 | 19.0% (n=42) | 20.13 |
24  | Roguelike Deckbuilder | 311 | 59% | 38.9% | 26.0% | $48K | $10 | 18.3% (n=164) | 3.48 |
25  | Colony Sim | 295 | 52% | 38.3% | 31.5% | $97K | $11 | 9.6% (n=146) | 1.20 |
26  | Online Co-Op | 770 | 55% | 37.7% | 31.9% | $49K | $10 | 21.9% (n=489) | 1.85 |
27  | Moddable | 72 | 54% | 37.5% | 29.2% | $71K | $10 | 16.2% (n=37) | 0.97 |
28  | Bullet Heaven | 130 | 56% | 36.9% | 28.5% | $30K | $5 | 31.9% (n=116) | 10.50 |
29  | Base Building | 597 | 52% | 36.9% | 30.5% | $74K | $11 | 6.9% (n=289) | 1.16 |
30  | World War II | 90 | 44% | 36.7% | 30.0% | $74K | $10 | 8.3% (n=48) | 0.35 |
... [90 more lines]

| Roguelike Deckbuilder | 311 | 59% | 38.9% | 26.0% | $48K | $10 | 18.3% (n=164) | 3.48 |
| Deckbuilding | 636 | 43% | 23.4% | 17.1% | $26K | $9 | 9.5% (n=423) | 2.92 |
| Top-Down Shooter | 1006 | 17% | 7.6% | 5.0% | $8K | $5 | 4.2% (n=891) | 0.96 |
| Online Co-Op + Roguelite | 51 | 75% | 54.9% | 39.2% | $90K | $15 | 33.3% (n=21) | 2.73 |
| Online Co-Op + Action Roguelike | 66 | 70% | 50.0% | 37.9% | $90K | $12 | 28.1% (n=32) | 4.33 |
| Auto Battler + Roguelike Deckbuilder | 30 | 90% | 50.0% | 33.3% | $62K | $10 | – | 20.21 |
| Roguelite + Loot | 34 | 59% | 50.0% | 44.1% | $120K | $7 | 42.3% (n=26) | 2.44 |
| Replay Value + Roguelike Deckbuilder | 33 | 67% | 48.5% | 39.4% | $74K | $12 | – | 5.08 |
| Multiplayer + Action Roguelike | 46 | 67% | 47.8% | 34.8% | $62K | $10 | 26.1% (n=23) | 2.25 |
| Co-op + Roguelike | 36 | 64% | 47.2% | 36.1% | $74K | $12 | – | 2.31 |
| Co-op + Roguelite | 53 | 62% | 47.2% | 39.6% | $112K | $12 | 19.2% (n=26) | 2.44 |
| Roguelike + Online Co-Op | 34 | 59% | 47.1% | 38.2% | $87K | $11 | – | 5.49 |
| Multiplayer + Roguelite | 32 | 59% | 46.9% | 34.4% | $87K | $10 | – | 2.12 |
| Turn-Based + Roguelike | 54 | 56% | 46.3% | 38.9% | $79K | $14 | 14.3% (n=21) | 1.11 |

24:| Roguelike Deckbuilder | 311 | 59% | 38.9% | 26.0% | $48K | $10 | 18.3% (n=164) | 3.48 |
205:| Roguelite + Loot | 34 | 59% | 50.0% | 44.1% | $120K | $7 | 42.3% (n=26) | 2.44 |
227:| Roguelike + Online Co-Op | 34 | 59% | 47.1% | 38.2% | $87K | $11 | – | 5.49 |
===
251:## Mechanical openings: ≥2× baseline P(≥$50K), supply growing ≤1.15× market, med price ≤$20

251  ## Mechanical openings: ≥2× baseline P(≥$50K), supply growing ≤1.15× market, med price ≤$20
252 
253 _298 pairs qualify; ranked by hit-rate × supply-slack_
254 
255 | tag | n | ≥$10K | ≥$50K | ≥$100K | med $ (rev’d) | med price | ≥$50K @≤$10 | supply trend |
256 |---|---|---|---|---|---|---|---|---|
257 | Anime + Nudity | 77 | 66% | 45.5% | 32.5% | $51K | $10 | 20.0% (n=40) | 0.17 |
258 | RPG + Turn-Based | 52 | 77% | 53.8% | 42.3% | $91K | $20 | – | 0.40 |
259 | Driving + Multiplayer | 42 | 69% | 54.8% | 47.6% | $284K | $15 | 15.0% (n=20) | 0.42 |
260 | Visual Novel + Sexual Content | 116 | 66% | 47.4% | 34.5% | $71K | $9 | 33.3% (n=69) | 0.40 |
261 | Action + Sexual Content | 52 | 71% | 42.3% | 38.5% | $52K | $10 | 27.6% (n=29) | 0.33 |
262 | Indie + Co-op | 45 | 53% | 42.2% | 35.6% | $108K | $12 | 9.5% (n=21) | 0.34 |
263 | Anime + Sexual Content | 173 | 75% | 42.2% | 27.7% | $46K | $13 | 22.7% (n=75) | 0.35 |
264 | Romance + Sexual Content | 39 | 69% | 46.2% | 41.0% | $69K | $9 | 33.3% (n=24) | 0.44 |
265 | Adventure + Nudity | 67 | 70% | 35.8% | 22.4% | $39K | $14 | 15.0% (n=20) | 0.22 |
266 | Story Rich + Sexual Content | 43 | 67% | 51.2% | 37.2% | $74K | $11 | 33.3% (n=21) | 0.52 |
267 | Dating Sim + Sexual Content | 84 | 69% | 51.2% | 36.9% | $81K | $8 | 37.9% (n=58) | 0.53 |
268 | RPG + Multiplayer | 61 | 51% | 47.5% | 42.6% | $239K | $12 | 14.3% (n=28) | 0.49 |
269 | Simulation + Sexual Content | 52 | 71% | 46.2% | 26.9% | $54K | $9 | 27.6% (n=29) | 0.48 |
270 | Visual Novel + Nudity | 54 | 48% | 33.3% | 25.9% | $35K | $8 | 23.1% (n=39) | 0.21 |
271 | Multiplayer + VR | 62 | 53% | 40.3% | 32.3% | $180K | $15 | – | 0.39 |
272 | Open World + Open World Survival Craft | 67 | 75% | 59.7% | 49.3% | $194K | $15 | – | 0.66 |
273 | FPS + Open World | 42 | 60% | 42.9% | 28.6% | $68K | $13 | – | 0.46 |
274 | JRPG + Nudity | 65 | 75% | 40.0% | 21.5% | $43K | $14 | – | 0.44 |
275 | Action + Gore | 115 | 47% | 32.2% | 28.7% | $24K | $8 | 8.2% (n=73) | 0.26 |
276 | RPG + Nudity | 119 | 65% | 39.5% | 24.4% | $44K | $13 | 18.8% (n=48) | 0.44 |
277 | FPS + Gore | 43 | 42% | 34.9% | 30.2% | $75K | $10 | 4.0% (n=25) | 0.34 |
278 | Nudity + Sexual Content | 233 | 60% | 33.9% | 23.2% | $34K | $8 | 18.4% (n=136) | 0.32 |
279 | Nudity + Dating Sim | 44 | 48% | 31.8% | 25.0% | $53K | $7 | 24.2% (n=33) | 0.27 |
280 | Cute + Sexual Content | 65 | 77% | 43.1% | 24.6% | $40K | $10 | 24.2% (n=33) | 0.52 |
... [44 more lines]

File created successfully at: E:\a327ex\steam-market\scripts\drill_tags.py (file state is current in your context — no need to Read it back)

=== tags of interest (2023-2025 window) ===

tag                          n   >=50K  @<=$10    med$

Action Roguelike          2347   14.4%    7.4%    12K

Rogue-lite               (no such tag)

Roguelike                 2392   15.5%    6.7%    14K

Auto Battler               464   16.8%   10.5%    20K

Snake                    (no such tag)

Bullet Hell               1767   10.9%    6.8%     8K

Twin Stick Shooter         360   12.8%    6.6%    10K

Arena Shooter              636    7.7%    3.6%     7K

Clicker                  (no such tag)

Idler                     1078   12.8%   11.5%     8K

Tower Defense             1071   10.9%    5.3%    12K

Card Game                 1160   14.3%    4.1%    19K

Card Battler               637   19.0%    4.0%    22K

Strategy                  5935   14.1%    4.7%    16K

Action RTS                 213   12.7%    2.7%    20K

Wargame                    479   17.7%    3.1%    23K

Vehicular Combat           100   19.0%    4.5%    15K

Roguevania               (no such tag)

Traditional Roguelike       84   11.9%    3.2%     9K

Dungeon Crawler           1020   18.2%    5.4%    20K

Perma Death                396    8.1%    2.7%     7K

Procedural Generation      505   14.5%    5.7%    15K

Pixel Graphics            3565   13.4%    4.8%    11K

Retro                     1217   11.4%    4.3%     9K

2D                        7758    6.9%    2.5%     8K

Difficult                 1157   14.0%    2.7%    10K

Fast-Paced                 214   22.9%    2.1%    16K

Physics                   1175   12.7%    5.5%    11K

... [24 more lines]

File created successfully at: E:\a327ex\steam-market\reports\notes.md (file state is current in your context — no need to Read it back)

File created successfully at: E:\a327ex\steam-market\scripts\anomalies.py (file state is current in your context — no need to Read it back)

File created successfully at: E:\a327ex\steam-market\scripts\reviews_lang_crawl.py (file state is current in your context — no need to Read it back)

Updated task #4 status

Updated task #6 status

groups: first-time self-pub=12273, veteran self-pub=5442, published (any)=6778

written to reports/04-anomalies.md

# 04 — Success anomalies: the resource-normalized view



Generated 2026-07-07 13:47 UTC. Coverage: search=115281, steamspy=82233 (dev/pub fields from SteamSpy; name-matching is fuzzy-imperfect).



Window: paid releases 2023–2025 with dev data (n=24493).



## Baseline by resource group



| group | n | ≥$10K | ≥$50K | ≥$100K | med $ (rev’d) |

|---|---|---|---|---|---|

| first-time self-pub | 12273 | 19.6% | 8.56% | 5.99% | $8K |

| veteran self-pub | 5442 | 31.0% | 15.77% | 12.24% | $11K |

| published (any) | 6778 | 49.3% | 33.05% | 26.60% | $45K |



Of 4149 window games ≥$50K, **1051 (25%) are first-time self-published**.



## Where outsiders break through: P(≥$50K | first-time self-pub, tag)



_Tags with ≥50 first-time self-pub releases in window. Compare vs tag overall._



| tag | n (fs) | ≥$50K fs | ≥$50K all | fs/all |

|---|---|---|---|---|

| FMV | 54 | 48.1% | 48.1% | 1.00 |

| Open World Survival Craft | 64 | 40.6% | 56.5% | 0.72 |

| Sexual Content | 81 | 37.0% | 40.9% | 0.91 |

| Cozy | 128 | 28.1% | 29.4% | 0.96 |

| Dating Sim | 209 | 27.3% | 34.1% | 0.80 |

| Open World | 348 | 25.0% | 40.5% | 0.62 |

... [2 more lines]

| Romance | 132 | 24.2% | 27.2% | 0.89 |
| Management | 310 | 23.9% | 35.5% | 0.67 |
| Cooking | 59 | 23.7% | 27.1% | 0.88 |
| Co-op | 206 | 21.8% | 39.1% | 0.56 |
| Loot | 87 | 21.8% | 35.3% | 0.62 |
| Crafting | 227 | 20.7% | 35.8% | 0.58 |
| LGBTQ+ | 97 | 20.6% | 31.7% | 0.65 |
| Multiplayer | 530 | 20.6% | 35.9% | 0.57 |
| Capitalism | 74 | 20.3% | 24.6% | 0.82 |
| Base Building | 224 | 20.1% | 39.2% | 0.51 |
| Roguelike Deckbuilder | 90 | 20.0% | 42.7% | 0.47 |
| Building | 281 | 19.9% | 34.6% | 0.58 |
| Sandbox | 520 | 19.4% | 30.8% | 0.63 |
| Nudity | 52 | 19.2% | 35.1% | 0.55 |
| Economy | 149 | 18.8% | 30.0% | 0.63 |
| Immersive Sim | 475 | 18.7% | 24.0% | 0.78 |
| Automation | 107 | 18.7% | 36.1% | 0.52 |
| Character Customization | 231 | 18.6% | 27.7% | 0.67 |
| Military | 77 | 18.2% | 31.0% | 0.59 |
| Gore | 90 | 17.8% | 30.7% | 0.58 |
| Wholesome | 62 | 17.7% | 21.9% | 0.81 |
| Colony Sim | 85 | 17.6% | 42.5% | 0.42 |
| Life Sim | 323 | 17.3% | 26.4% | 0.66 |
| Resource Management | 169 | 17.2% | 32.6% | 0.53 |
| War | 70 | 17.1% | 27.9% | 0.61 |
| City Builder | 203 | 16.7% | 32.1% | 0.52 |
| Boomer Shooter | 66 | 16.7% | 29.2% | 0.57 |
| Music | 50 | 16.0% | 27.3% | 0.59 |
| Realistic | 359 | 15.9% | 20.4% | 0.78 |
| Historical | 101 | 15.8% | 29.0% | 0.55 |
... [25 more lines]

The file E:\a327ex\steam-market\reports\notes.md has been updated successfully. (file state is current in your context — no need to Read it back)

Exit code 1
Traceback (most recent call last):

  File "<string>", line 13, in <module>

    print('OK ', appid, name[:35])

    ~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^

  File "C:\Users\adn\AppData\Roaming\uv\python\cpython-3.13-windows-x86_64-none\Lib\encodings\cp1252.py", line 19, in encode

    return codecs.charmap_encode(input,self.errors,encoding_table)[0]

           ~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

UnicodeEncodeError: 'charmap' codec can't encode characters in position 0-34: character maps to <undefined>



During handling of the above exception, another exception occurred:



Traceback (most recent call last):

  File "<string>", line 15, in <module>

    print('FAIL', appid, name[:35], e)

    ~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

  File "C:\Users\adn\AppData\Roaming\uv\python\cpython-3.13-windows-x86_64-none\Lib\encodings\cp1252.py", line 19, in encode

    return codecs.charmap_encode(input,self.errors,encoding_table)[0]

           ~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

UnicodeEncodeError: 'charmap' codec can't encode characters in position 0-34: character maps to <undefined>

OK  2730700 DRONE PERSPECTIVE

FAIL 2730810 EVERYBODY'S GOLF HOT SHOTS HTTP Error 404: Not Found

OK  2730970 Westland Survival

OK  2731330 TROLEU

OK  2731590 Block Shop

OK  2732050 HAS Galactics

OK  2732100 Medieval Blacksmith

OK  2732210 Hopeless Sea

OK  2883370 Golden Ball

OK  2883380 Jishin

... [2 more lines]

 + pillow==12.3.0

File created successfully at: E:\a327ex\steam-market\scripts\art_sample.py (file state is current in your context — no need to Read it back)

Action Roguelike/fail: 30 images

Action Roguelike/mid: 30 images

Action Roguelike/hit: 30 images

Horror/fail: 30 images

Horror/mid: 30 images

Horror/hit: 30 images

Idler/fail: 26 images

Idler/mid: 29 images

Idler/hit: 30 images

Cozy/fail: 30 images

Cozy/mid: 30 images

Cozy/hit: 30 images

355 games -> 18 sheets

mapping written

[
  {
    "type": "image",
    "note": "[image stored privately: vault/media/fable-test-steam-market-research-1/001.jpg]"
  }
]

[
  {
    "type": "image",
    "note": "[image stored privately: vault/media/fable-test-steam-market-research-1/002.jpg]"
  }
]

[
  {
    "type": "image",
    "note": "[image stored privately: vault/media/fable-test-steam-market-research-1/003.jpg]"
  }
]

[
  {
    "type": "image",
    "note": "[image stored privately: vault/media/fable-test-steam-market-research-1/004.jpg]"
  }
]

[
  {
    "type": "image",
    "note": "[image stored privately: vault/media/fable-test-steam-market-research-1/005.jpg]"
  }
]

[
  {
    "type": "image",
    "note": "[image stored privately: vault/media/fable-test-steam-market-research-1/006.jpg]"
  }
]

[
  {
    "type": "image",
    "note": "[image stored privately: vault/media/fable-test-steam-market-research-1/007.jpg]"
  }
]

[
  {
    "type": "image",
    "note": "[image stored privately: vault/media/fable-test-steam-market-research-1/008.jpg]"
  }
]

[
  {
    "type": "image",
    "note": "[image stored privately: vault/media/fable-test-steam-market-research-1/009.jpg]"
  }
]

[
  {
    "type": "image",
    "note": "[image stored privately: vault/media/fable-test-steam-market-research-1/010.jpg]"
  }
]

[
  {
    "type": "image",
    "note": "[image stored privately: vault/media/fable-test-steam-market-research-1/011.jpg]"
  }
]

[
  {
    "type": "image",
    "note": "[image stored privately: vault/media/fable-test-steam-market-research-1/012.jpg]"
  }
]

[
  {
    "type": "image",
    "note": "[image stored privately: vault/media/fable-test-steam-market-research-1/013.jpg]"
  }
]

[
  {
    "type": "image",
    "note": "[image stored privately: vault/media/fable-test-steam-market-research-1/014.jpg]"
  }
]

[
  {
    "type": "image",
    "note": "[image stored privately: vault/media/fable-test-steam-market-research-1/015.jpg]"
  }
]

[
  {
    "type": "image",
    "note": "[image stored privately: vault/media/fable-test-steam-market-research-1/016.jpg]"
  }
]

[
  {
    "type": "image",
    "note": "[image stored privately: vault/media/fable-test-steam-market-research-1/017.jpg]"
  }
]

[
  {
    "type": "image",
    "note": "[image stored privately: vault/media/fable-test-steam-market-research-1/018.jpg]"
  }
]

355 scores written

File created successfully at: E:\a327ex\steam-market\scripts\art_gate.py (file state is current in your context — no need to Read it back)

# 05 — The art-quality gate



Generated 2026-07-07 13:59 UTC. 355 games, 4 niches × 3 outcome groups (~30 each), header images scored blind 1–5 (1=asset-flip, 2=amateur, 3=competent, 4=good, 5=premium) without knowledge of outcomes.



Groups: **fail** = <10 reviews; **mid** = reviewed but <$50K est; **hit** = ≥$50K est. All paid, released 2023–2025.



## Mean art score by niche × outcome



| niche | fail | mid | hit | hit−fail gap |

|---|---|---|---|---|

| Action Roguelike | 2.33 | 2.40 | 3.47 | +1.13 |

| Cozy | 2.60 | 2.63 | 3.57 | +0.97 |

| Horror | 1.80 | 2.03 | 3.30 | +1.50 |

| Idler | 1.73 | 2.52 | 2.53 | +0.80 |

| **all** | 2.13 | 2.39 | 3.22 | +1.09 |



## Outcome-group composition by art tier (all niches pooled)



_Stratified sample: equal priors per group. Share >33% = art tier enriches that outcome._



| art tier | n | fail share | mid share | hit share |

|---|---|---|---|---|

| 1 | 45 | 60% | 31% | 9% |

| 2 | 128 | 44% | 41% | 16% |

| 3 | 116 | 21% | 39% | 41% |

| 4 | 61 | 15% | 13% | 72% |

| 5 | 5 | 0% | 0% | 100% |



## P(hit | art tier) per niche (share of tier-games that are hits)



... [30 more lines]

Updated task #7 status

The file E:\a327ex\steam-market\reports\notes.md has been updated successfully. (file state is current in your context — no need to Read it back)

== bpk62lati
page 87: error Expecting value: line 1 column 1 (char 0) (fail 5)

== bb1zd8vmd
1000/34654 done

== b3sgn0kyj
500 fetched

== bwpyocf2n
800/1500

rows: 82233

max page: 86

Updated task #1 status

estimator: revenue ~= reviews * 35.9 * price_usd (band [25, 55])

delta rerun ok
written to reports/04-anomalies.md

=== FMV: 93 hits ===

   2.80M $14.99   5210 pub     Mandate Of Heaven                  RPG,Simulation,FMV,Strategy

   1.53M $19.99   2128 selfpub Five Hearts Under One Roof         Sexual Content,Dating Sim,FMV,Visual Novel

   1.50M $19.99   2087 pub     Hello Love: 18 Again               Sexual Content,FMV,Dating Sim,Cinematic

   1.32M $14.99   2457 selfpub Home Safety Hotline                Horror,1990's,Interactive Fiction,Point & Click

   1.28M $12.99   2748 selfpub 失业了，我获得了亿万游戏财产！                    Sexual Content,Strategy,Dating Sim,FMV

   1.22M $10.99   3087 pub     美女，请别影响我成仙                         FMV,Visual Novel,Interactive Fiction,Cinematic

   1.00M $ 6.99   3967 pub     Vanity Fair: The Pursuit           FMV,RPG,Choices Matter,Story Rich

   0.95M $10.99   2413 pub     恋爱单选题                              Dating Sim,Simulation,FMV,Life Sim

   0.92M $12.99   1972 pub     Conquer the world                  Interactive Fiction,FMV,Multiple Endings,Visual Novel

   0.85M $11.99   1968 pub     咸鱼殿下 / My journey                  FMV,Dating Sim,Lore-Rich,Cinematic

   0.72M $10.99   1826 selfpub Love is All Around: Echoes of Yest Adventure,Simulation,RPG,Dating Sim

   0.71M $ 9.99   1966 pub     我只是想让她们都幸福                         FMV,Dating Sim,First-Person,Simulation

   0.69M $22.99    842 pub     Five Hearts Under One Roof season2 Sexual Content,Dating Sim,FMV,Choices Matter

   0.66M $ 9.99   1852 pub     极速营救                               Strategy,RPG,Story Rich,Dating Sim

   0.65M $ 9.99   1819 pub     Heartbeat in Thailand              FMV,Dating Sim,Visual Novel,Sexual Content

   0.60M $15.99   1046 pub     Game of Fate 2: A Century's Promis FMV,Dating Sim,Immersive Sim,Cinematic

   0.55M $15.99    963 pub     Love Begins With Goodbye           Dating Sim,FMV,Multiple Endings,Romance

   0.55M $ 9.99   1535 pub     恋爱吗？我家猫会后空翻                        Immersive Sim,Visual Novel,Incremental,Interactive Fiction

   0.54M $ 9.99   1519 selfpub Love Is All Around                 Dating Sim,Sexual Content,FMV,Singleplayer

   0.52M $ 9.99   1446 pub     FLAM EVAW                          FMV,First-Person,Detective,Story Rich

   0.48M $ 8.49   1583 pub     Road to Empress I                  Choices Matter,FMV,RPG,Female Protagonist

   0.45M $10.99   1138 pub     我打不过漂亮的她们                          RPG,Strategy,Interactive Fiction,Story Rich

=== Idler: 138 hits ===

   3.77M $11.99   8750 selfpub Spirit City: Lofi Sessions         Cozy,Decorating,Desktop Companion,Organizing

   2.55M $14.99   4745 pub     Cast n Chill                       Pixel Graphics,Relaxing,Fishing,Idler

   1.83M $ 6.99   7290 pub     (the) Gnorp Apologue               Incremental,Idler,Strategy,Building

   1.64M $ 6.99   6517 selfpub Rusty's Retirement                 Idler,Automation,Farming Sim,Pixel Graphics

   1.46M $15.99   2551 pub     BitCraft Online                    Early Access,MMORPG,Open World Survival Craft,Idler

   1.40M $ 6.99   5591 selfpub Ages of Conflict: World War Simula Sandbox,God Game,Simulation,Alternate History

... [16 more lines]

stayed in niche (>=2 shared tags w/ prior hit): n=1273, >=50K again: 61.0%

switched niche (<2 shared tags): n=556, >=50K again: 52.7%

=== launch month vs P(>=50K), 2023-2025 paid ===

  01: n= 2646  hit=14.1%

  02: n= 2807  hit=12.9%

  03: n= 3288  hit=13.6%

  04: n= 3127  hit=15.9%

  05: n= 3163  hit=15.7%

  06: n= 3000  hit=13.3%

  07: n= 3278  hit=14.6%

  08: n= 3330  hit=14.6%

  09: n= 3172  hit=17.3%

  10: n= 3904  hit=14.9%

  11: n= 3998  hit=14.1%

  12: n= 3471  hit=10.5%



=== median playtime (hrs, SteamSpy median_forever>0, >=10 reviews) by tag ===

  ...

The file E:\a327ex\steam-market\reports\notes.md has been updated successfully. (file state is current in your context — no need to Read it back)

Command running in background with ID: b7j915o7h. Output is being written to: C:\Users\adn\AppData\Local\Temp\claude\E--a327ex\689cefaa-c928-4524-85fc-f8dfbde9fd7f\tasks\b7j915o7h.output. You will be notified when it completes. To check interim output, use Read on that file path.

Exit code 127
/usr/bin/bash: line 12: .venv/Scripts/python.exe: No such file or directory

Command running in background with ID: b46q12ddl. Output is being written to: C:\Users\adn\AppData\Local\Temp\claude\E--a327ex\689cefaa-c928-4524-85fc-f8dfbde9fd7f\tasks\b46q12ddl.output. You will be notified when it completes. To check interim output, use Read on that file path.

n=1500 paid sub-10-review games sampled

exactly 0 reviews: 220 (15%)

1-4 reviews: 667 (44%)

5-9 reviews: 409 (27%)

>=10 (accrued since search crawl): 204

SNKRX search tooltip: (3618, 94, 'Very Positive')

biggest sub10-crossovers (search said <10, API purchase_type=all says): [(2943780, 591), (3241030, 234), (1032750, 183), (2271320, 71), (3274800, 67), (332810, 53), (2660320, 45), (498740, 43)]

{"success":1,"query_summary":{"num_reviews":0,"review_score":8,"review_score_desc":"Very Positive","total_positive":3967,"total_negative":227,"total_reviews":4194},"reviews":[],"cursor":"*"}

The file E:\a327ex\steam-market\reports\notes.md has been updated successfully. (file state is current in your context — no need to Read it back)

Command running in background with ID: bz6x6epfk. Output is being written to: C:\Users\adn\AppData\Local\Temp\claude\E--a327ex\689cefaa-c928-4524-85fc-f8dfbde9fd7f\tasks\bz6x6epfk.output. You will be notified when it completes. To check interim output, use Read on that file path.

n=800 top-earner games; key-activation share of reviews:

  p50: 13.6%

  p75: 22.7%

  p90: 31.7%

  p95: 37.8%

  >40% key share: 24 games

728 games x 7 langs -> 5096 calls

    L.append(f"| {g['name']} | {g['appid']} | {g['year']} | ${g['price']:.2f} | "

                                                             ^^^^^^^^^^^^^^^^

TypeError: unsupported format string passed to NoneType.__format__

The file E:\a327ex\steam-market\scripts\trajectory.py has been updated successfully. (file state is current in your context — no need to Read it back)

Updated task #3 status

Updated task #6 status

3756 games classified; written to reports/03-trajectories.md

1    # 03 — Trajectory taxonomy (where the long tail lives)
2   
3   Generated 2026-07-07 14:38 UTC. Histograms: 3805 games crawled; 3756 classified (≥50 reviews, ≥12 months of history).
4   
5   Classes over first-24-months review curve: **launch-spike** (≥60% of reviews in months 1–3, dead tail), **decay** (front-loaded, weak tail), **steady** (tail ≥35% of launch rate), **slow-burn** (≤35% early share, tail ≥50%), **resurrected** (a month ≥6 exceeds 1.5× month 1).
6   
7   ## Class distribution by release cohort
8   
9   | cohort | n | launch-spike | decay | steady | slow-burn | resurrected | med m12/m3 mult |
10  |---|---|---|---|---|---|---|---|
11  | 2011 | 60 | 0% | 25% | 7% | 3% | 65% | 3.17× |
12  | 2012 | 79 | 0% | 0% | 3% | 13% | 85% | 2.88× |
13  | 2013 | 120 | 0% | 12% | 1% | 3% | 83% | 5.18× |
14  | 2014 | 168 | 1% | 38% | 7% | 10% | 44% | 2.51× |
15  | 2015 | 197 | 3% | 40% | 12% | 14% | 32% | 2.27× |
16  | 2016 | 225 | 2% | 49% | 11% | 11% | 28% | 2.09× |
17  | 2017 | 240 | 5% | 59% | 7% | 8% | 22% | 1.91× |
18  | 2018 | 207 | 2% | 33% | 16% | 19% | 29% | 1.95× |
19  | 2019 | 261 | 2% | 43% | 12% | 12% | 31% | 2.20× |
20  | 2020 | 304 | 7% | 46% | 9% | 9% | 29% | 2.16× |
21  | 2021 | 304 | 13% | 52% | 8% | 6% | 21% | 1.82× |
22  | 2022 | 303 | 13% | 55% | 7% | 6% | 19% | 1.79× |
23  | 2023 | 316 | 13% | 53% | 9% | 6% | 20% | 1.89× |
24  | 2024 | 368 | 18% | 59% | 4% | 2% | 17% | 1.72× |
25  | 2025 | 362 | 32% | 50% | 4% | 3% | 11% | 1.53× |
26  | 2026 | 78 | 47% | 41% | 4% | 0% | 8% | 1.44× |
27  
28  ## Tags by long-tail index (share steady+slow-burn+resurrected)
29  
30  _Games ≥50 reviews only; tags with ≥30 classified games. m12/m3 = how much a game grows from month-3 to month-12 cumulative reviews (1.0 = tail adds nothing)._
... [15 more lines]

The file E:\a327ex\steam-market\reports\notes.md has been updated successfully. (file state is current in your context — no need to Read it back)

4800/5096

language crawl COMPLETE

rows: 5095

games: 728

File created successfully at: E:\a327ex\steam-market\scripts\lang_gaps.py (file state is current in your context — no need to Read it back)

# 06 — Language demand vs localization supply



Generated 2026-07-07 16:55 UTC. 728 top games (2022–2025, ≥100 reviews) across 12 niches; review counts per language via appreviews. Supply = supported_languages (appdetails, coverage 527/728).



Sampled-language reviews only (EN+6); other languages excluded from shares.



## Demand: review-language share by niche



| niche | n games | EN | CN | JP | KR | PT-BR | RU | DE | non-EN total |

|---|---|---|---|---|---|---|---|---|---|

| Roguelike Deckbuilder | 74 | 60% | 25% | 2% | 3% | 4% | 4% | 2% | 40% |

| Action Roguelike | 136 | 60% | 17% | 1% | 3% | 5% | 10% | 4% | 40% |

| Auto Battler | 72 | 48% | 34% | 2% | 3% | 2% | 6% | 5% | 52% |

| Bullet Heaven | 67 | 57% | 20% | 1% | 3% | 4% | 10% | 6% | 43% |

| Idler | 88 | 52% | 16% | 2% | 2% | 5% | 20% | 3% | 48% |

| Horror | 77 | 57% | 14% | 0% | 2% | 5% | 19% | 3% | 43% |

| FMV | 70 | 11% | 85% | 0% | 1% | 1% | 2% | 0% | 89% |

| Turn-Based | 70 | 43% | 47% | 2% | 2% | 1% | 3% | 2% | 57% |

| Dungeon Crawler | 78 | 66% | 12% | 2% | 2% | 4% | 10% | 4% | 34% |

| Colony Sim | 73 | 63% | 14% | 1% | 2% | 3% | 9% | 9% | 37% |

| Card Battler | 76 | 50% | 30% | 4% | 4% | 5% | 4% | 3% | 50% |

| Tower Defense | 75 | 50% | 30% | 2% | 3% | 3% | 8% | 5% | 50% |



## Supply vs demand: share of niche games localized, per language



_gap = demand share high while few games support the language._



| niche | CN sup | JP sup | KR sup | PT-BR sup | RU sup | DE sup |

|---|---|---|---|---|---|---|

| Roguelike Deckbuilder (58) | 98% | 86% | 62% | 62% | 69% | 71% |

... [15 more lines]

Updated task #8 status

The file E:\a327ex\steam-market\reports\notes.md has been updated successfully. (file state is current in your context — no need to Read it back)

Command running in background with ID: brchlnrv3. Output is being written to: C:\Users\adn\AppData\Local\Temp\claude\E--a327ex\689cefaa-c928-4524-85fc-f8dfbde9fd7f\tasks\brchlnrv3.output. You will be notified when it completes. To check interim output, use Read on that file path.

histograms done: 18373

appdetails: 12029 failed: 0

18000/34654 done

9500 fetched

17955 games classified; written to reports/03-trajectories.md

File created successfully at: E:\a327ex\steam-market\scripts\features.py (file state is current in your context — no need to Read it back)

# 07 — Store-feature signatures: hits vs failures



Generated 2026-07-07 19:29 UTC. Window: paid 2023–2025. Arms: hit=ALL ≥$50K games crawled (n=5596), mid=random 10–49-review sample (n=125), fail=random <10-review sample (n=4000).



| feature | fail | mid | hit |

|---|---|---|---|

| has demo | 18% | 25% | 29% |

| achievements > 0 | 45% | 70% | 86% |

| achievements > 30 | 8% | 14% | 44% |

| full controller support | 20% | 29% | 47% |

| any DLC | 5% | 11% | 46% |

| self-published (dev==pub) | 82% | 77% | 46% |

| single-dev-name team | 95% | 89% | 91% |

| >=5 languages | 20% | 45% | 66% |

| >=10 languages | 11% | 26% | 41% |

| multiplayer category | 14% | 7% | 25% |



Note: these are correlations — features signal effort/production value as much as they cause outcomes. Demo share in particular mixes "made a demo" with "kind of game that demos well."

| cohort | n | launch-spike | decay | steady | slow-burn | resurrected | med m12/m3 mult |
|---|---|---|---|---|---|---|---|
| 2007 | 67 | 0% | 3% | 0% | 34% | 63% | 4.46× |
| 2008 | 87 | 0% | 1% | 2% | 47% | 49% | 4.00× |
| 2009 | 186 | 0% | 2% | 2% | 44% | 53% | 4.16× |
| 2010 | 140 | 0% | 3% | 3% | 28% | 66% | 3.39× |
| 2011 | 177 | 0% | 22% | 7% | 23% | 48% | 3.29× |
| 2012 | 226 | 0% | 6% | 6% | 20% | 68% | 2.86× |
| 2013 | 358 | 0% | 16% | 5% | 12% | 67% | 4.51× |
| 2014 | 721 | 2% | 36% | 10% | 16% | 36% | 2.50× |
| 2015 | 872 | 6% | 42% | 11% | 14% | 27% | 2.19× |
| 2016 | 1133 | 7% | 50% | 10% | 10% | 23% | 2.03× |
| 2017 | 1155 | 7% | 58% | 7% | 11% | 18% | 1.92× |
| 2018 | 1131 | 7% | 45% | 13% | 15% | 19% | 1.86× |
| 2019 | 1166 | 5% | 51% | 10% | 11% | 24% | 2.08× |
| 2020 | 1369 | 10% | 57% | 7% | 7% | 20% | 1.92× |
| 2021 | 1444 | 15% | 55% | 7% | 5% | 18% | 1.77× |
| 2022 | 1449 | 17% | 59% | 6% | 4% | 14% | 1.68× |
| 2023 | 1587 | 17% | 57% | 6% | 4% | 15% | 1.74× |
...
## Tags by long-tail index (share steady+slow-burn+resurrected)

_Games ≥50 reviews only; tags with ≥30 classified games. m12/m3 = how much a game grows from month-3 to month-12 cumulative reviews (1.0 = tail adds nothing)._

| tag | n | long-tail idx | med m12/m3 | launch-spike% | resurrected% |
|---|---|---|---|---|---|
| Classic | 352 | 68% | 2.69× | 2% | 32% |
| Real-Time with Pause | 49 | 65% | 2.73× | 12% | 29% |
| Local Co-Op | 288 | 60% | 2.25× | 7% | 35% |
| Split Screen | 53 | 58% | 2.06× | 8% | 32% |
... [20 more lines]

Updated task #5 status

Updated task #9 status

File created successfully at: E:\a327ex\steam-market\reports\00-synthesis.md (file state is current in your context — no need to Read it back)

Updated task #9 status

The file E:\a327ex\steam-market\reports\notes.md has been updated successfully. (file state is current in your context — no need to Read it back)

Command running in background with ID: birls1rwm. Output is being written to: C:\Users\adn\AppData\Local\Temp\claude\E--a327ex\689cefaa-c928-4524-85fc-f8dfbde9fd7f\tasks\birls1rwm.output. You will be notified when it completes. To check interim output, use Read on that file path.

baseline 2019-21: 17.2%   baseline 2023-25: 14.3%

tag                    19-21 hit  vs base  pctile    23-25 hit  vs base  pctile

Arcade                     6.9%    0.40x      8th        5.5%    0.39x      5th

Minimalist                 4.3%    0.25x      2th        5.4%    0.38x      5th

Abstract                   4.4%    0.26x      3th        5.1%    0.35x      4th

Score Attack               5.9%    0.34x      5th        3.7%    0.26x      2th

Precision Platformer       5.8%    0.34x      5th        4.5%    0.32x      4th

2D                         9.9%    0.57x     15th        6.9%    0.48x      7th

Puzzle                    10.6%    0.62x     17th        7.8%    0.54x     12th

Arena Shooter              6.8%    0.40x      7th        7.7%    0.54x     11th

Top-Down Shooter           6.6%    0.38x      6th        7.6%    0.53x     10th

Bullet Hell               12.3%    0.72x     24th       10.9%    0.76x     25th

Action Roguelike          20.0%    1.16x     48th       14.4%    1.01x     44th

Roguelike                 26.8%    1.56x     68th       15.5%    1.09x     48th

Twin Stick Shooter        11.7%    0.68x     21th       12.8%    0.89x     35th

Pixel Graphics            18.1%    1.05x     42th       13.4%    0.93x     38th

Fast-Paced                17.3%    1.01x     39th       22.9%    1.60x     75th

Difficult                 16.3%    0.95x     35th       14.0%    0.98x     41th

The file E:\a327ex\steam-market\reports\00-synthesis.md has been updated successfully. (file state is current in your context — no need to Read it back)

<tool_use_error>File has not been read yet. Read it first before writing to it.</tool_use_error>

{"message":"Successfully stopped task: birls1rwm (~/bin/neovim.exe /e/a327ex/steam-market/reports/00-synthesis.md -- -c \"MarkdownPreview\")","task_id":"birls1rwm","task_type":"local_bash","command":"~/bin/neovim.exe /e/a327ex/steam-market/reports/00-synthesis.md -- -c \"MarkdownPreview\""}

[sealed content withheld: memory]

195  **Game** — Lua code using the engine to build games
196 **Framework** — Lua framework files in `Anchor2/framework/anchor/`
197 
198 ### Documentation
199 
200 **Source of truth:** `Anchor2/engine/src/anchor.c` — the complete engine implementation (~10500 lines, single file).
201 
202 **Engine API:**
203 - `Anchor2/docs/ENGINE_API_QUICK.md` — compact function signatures (quick lookup)
204 - `Anchor2/docs/ENGINE_API.md` — detailed documentation with examples
205 
206 **Framework API:**
207 - `Anchor2/docs/FRAMEWORK_API_QUICK.md` — compact function signatures
208 - `Anchor2/docs/FRAMEWORK_API.md` — detailed documentation with examples
209 
210 **Context:**
211 - `Anchor2/.claude/CLAUDE.md` — Anchor 2 working agreements (read first)
212 - `Anchor2/reference/anchor2_plan.md` — authoritative plan for the Anchor 2 framework rewrite
213 - `Anchor2/arena/main.lua` — canonical v2 coding style example (~750 lines)
214 
215 ### Commands
216 
217 ```bash
218 # Engine (C code)
219 cd E:/a327ex/Anchor2/engine && ./build.bat    # Build engine for desktop
220 
221 # Framework (Lua)
222 cd E:/a327ex/Anchor2/framework && ./run.bat       # Run desktop
223 cd E:/a327ex/Anchor2/framework && ./run-web.bat   # Build web + run browser
224 
... [10 more lines]

235  **Theory curriculum:**
236 1. `music-theory/learning-plan.md` — 15 modules building toward analyzing Zettai Zetsumei
237 2. `music-theory/Zettai Zetsumei/chords-and-analysis.md` — the target song's chord progression
238 3. `music-theory/old/main.md` — reference tables for chord construction, intervals, harmonic fields
239 
240 **Production curriculum:**
241 4. `music-theory/production-learning-plan.md` — production modules building toward Speder2-style game music
242 5. `Anchor/reference/analysis/strudel-chiptune-guide.md` — Strudel synthesis reference
243 
244 **Teaching method:** Follow the instructions in each learning plan. Explain → exercises → check → repeat until mastery → advance.
245 
246 **Integration:** The plans have integration points noted — certain theory modules pair with production modules. The student wants an integrated approach.
247 
248 **Goal:** Create game music similar to Speder2/155 — chiptune textures blended with cleaner sounds (electric piano, strings), using sophisticated harmony (secondary dominants, borrowed chords, jazz elements).
249 
250 ---
251 
252 ## Long Responses
253 
254 When providing answers that are long or contain multiple code examples, create a markdown file in the relevant project's `reference/` folder and open it in NeoVim with MarkdownPreview:
255 
256 ```bash
257 ~/bin/neovim.exe reference/filename.md -- -c "MarkdownPreview"
258 ```
259 

The file E:\a327ex\.claude\CLAUDE.md has been updated successfully. (file state is current in your context — no need to Read it back)

[sealed content withheld: memory]

[sealed content withheld: memory]

[sealed content withheld: memory]

[sealed content withheld: memory]

File created successfully at: E:\a327ex\steam-market\scripts\build_tags.py (file state is current in your context — no need to Read it back)

MARKET BASELINE (paid 2023-25, n=39184): >=50K 14.3% | at <=$10 4.8% | first-time selfpub 8.7%



=== BUILD-DEPTH TAGS (2023-25 window) ===

Roguelike Deckbuilder                  n=311   h50= 38.9%  @<=$10= 18.3%  outsider= 20.0%  med=$  48K  supply= 3.48x  m12/m3=1.58x(n=55)

Bullet Heaven                          n=130   h50= 36.9%  @<=$10= 31.9%  outsider= 24.3%  med=$  30K  supply=10.50x  m12/m3=1.74x(n=27)

Loot                                   n=232   h50= 33.6%  @<=$10= 23.0%  outsider= 21.6%  med=$  60K  supply= 2.06x  m12/m3=1.71x(n=36)

Replay Value                           n=170   h50= 28.8%  @<=$10=  6.1%  outsider= 12.7%  med=$  24K  supply= 0.90x  m12/m3=1.54x(n=23)

JRPG                                   n=948   h50= 25.0%  @<=$10=  4.9%  outsider= 12.3%  med=$  32K  supply= 0.77x  m12/m3=1.55x(n=134)

Deckbuilding                           n=636   h50= 23.4%  @<=$10=  9.5%  outsider=  9.0%  med=$  26K  supply= 2.92x  m12/m3=1.53x(n=83)

Character Customization                n=679   h50= 23.1%  @<=$10=  6.2%  outsider= 18.9%  med=$  31K  supply= 1.17x  m12/m3=1.58x(n=83)

Party-Based RPG                        n=524   h50= 22.7%  @<=$10=  4.2%  outsider= 12.1%  med=$  34K  supply= 1.17x  m12/m3=1.52x(n=68)

Tactical RPG                           n=425   h50= 22.1%  @<=$10=  3.6%  outsider= 12.9%  med=$  28K  supply= 0.96x  m12/m3=1.49x(n=48)

Hack and Slash                         n=998   h50= 21.6%  @<=$10=  6.2%  outsider= 12.3%  med=$  25K  supply= 1.15x  m12/m3=1.52x(n=116)

RPG                                    n=5042  h50= 19.7%  @<=$10=  6.0%  outsider= 11.9%  med=$  25K  supply= 0.95x  m12/m3=1.51x(n=540)

Card Battler                           n=637   h50= 19.0%  @<=$10=  4.0%  outsider=  6.3%  med=$  22K  supply= 2.26x  m12/m3=1.53x(n=58)

Turn-Based Strategy                    n=1249  h50= 18.3%  @<=$10=  4.4%  outsider=  6.6%  med=$  23K  supply= 1.14x  m12/m3=1.59x(n=122)

Dungeon Crawler                        n=1020  h50= 18.2%  @<=$10=  5.4%  outsider= 11.1%  med=$  20K  supply= 1.11x  m12/m3=1.55x(n=98)

Action RPG                             n=1578  h50= 18.1%  @<=$10=  3.5%  outsider= 10.0%  med=$  26K  supply= 1.17x  m12/m3=1.50x(n=163)

Roguelite                              n=2400  h50= 17.8%  @<=$10=  8.3%  outsider= 10.2%  med=$  15K  supply= 2.16x  m12/m3=1.56x(n=244)

Creature Collector                     n=534   h50= 17.2%  @<=$10=  7.4%  outsider= 13.6%  med=$   9K  supply= 2.05x  m12/m3=1.50x(n=42)

Turn-Based Tactics                     n=1130  h50= 17.2%  @<=$10=  2.4%  outsider=  6.9%  med=$  20K  supply= 1.07x  m12/m3=1.57x(n=92)

Looter Shooter                         n=211   h50= 17.1%  @<=$10=  3.3%  outsider= 10.7%  med=$  21K  supply= 1.31x  m12/m3=1.77x(n=17)

Auto Battler                           n=464   h50= 16.8%  @<=$10= 10.5%  outsider= 10.6%  med=$  20K  supply= 2.66x  m12/m3=1.60x(n=45)

Strategy RPG                           n=569   h50= 16.7%  @<=$10=  4.2%  outsider= 13.5%  med=$  21K  supply= 1.09x  m12/m3=1.44x(n=54)

Class-Based                            n=57    h50= 15.8%  @<=$10=  7.9%  outsider=   -    med=$  11K  supply= 1.34x  m12/m3=-

CRPG                                   n=296   h50= 15.5%  @<=$10=  3.8%  outsider= 10.1%  med=$  16K  supply= 0.90x  m12/m3=1.48x(n=28)

Roguelike                              n=2392  h50= 15.5%  @<=$10=  6.7%  outsider=  9.2%  med=$  14K  supply= 2.40x  m12/m3=1.57x(n=212)

Procedural Generation                  n=505   h50= 14.5%  @<=$10=  5.7%  outsider= 10.5%  med=$  15K  supply= 0.99x  m12/m3=1.71x(n=44)

Action Roguelike                       n=2347  h50= 14.4%  @<=$10=  7.4%  outsider=  8.7%  med=$  12K  supply= 2.29x  m12/m3=1.63x(n=212)

Card Game                              n=1160  h50= 14.3%  @<=$10=  4.1%  outsider=  4.8%  med=$  19K  supply= 1.54x  m12/m3=1.56x(n=80)

... [37 more lines]

File created successfully at: E:\a327ex\steam-market\scripts\depth_claims.py (file state is current in your context — no need to Read it back)

coverage: hit=5596, mid=1879, fail=4000



=== claim prevalence by arm (any numeric build claim >=10) ===

  fail: any claim 3%   total>=100 1%   >=3 noun types 0%

   mid: any claim 4%   total>=100 1%   >=3 noun types 0%

   hit: any claim 7%   total>=100 3%   >=3 noun types 0%



=== same, build-tag games only (genre-controlled) ===

  fail (n=889): any claim 8%   total>=100 3%

   mid (n=403): any claim 13%   total>=100 5%

   hit (n=1275): any claim 17%   total>=100 10%



=== dose-response among HITS: median est $ by claimed depth ===

  all hits: no claim: $219K (n=5202) | total 10-99: $257K (n=201) | total 100-299: $193K (n=111) | total 300+: $595K (n=82)

  build-tag hits: no claim: $195K (n=1062) | total 10-99: $245K (n=90) | total 100-299: $190K (n=73) | total 300+: $418K (n=50)



=== noun-level: claim prevalence in hits vs fails ===

  items        hits  1.3%  fails  0.5%  ratio  2.6x

  weapons      hits  1.2%  fails  0.5%  ratio  2.4x

  characters   hits  1.6%  fails  0.5%  ratio  3.0x



=== within-Idler: build-claim idlers vs plain idlers ===

  fail idlers (n=178): 2% have numeric build claims

  hit idlers (n=180): 8% have numeric build claims

=== ALL-TIME top by reviews ===

 141911 | 2013 | $ 0.00 |   238960 | Path of Exile                            | Free to Play, Action RPG, Dark Fantasy, Hack and Slash, Loot

 122489 | 2012 | $19.99 |    49520 | Borderlands 2                            | Loot, Shooter, Action, Multiplayer, Co-op

  95403 | 2024 | $29.99 |  2694490 | Path of Exile 2                          | Action RPG, Hack and Slash, RPG, Isometric, Loot

  59936 | 2020 | $59.99 |   397540 | Borderlands 3                            | RPG, Looter Shooter, Multiplayer, FPS, Online Co-Op

  59512 | 2024 | $34.99 |   899770 | Last Epoch                               | Action RPG, Loot, Hack and Slash, RPG, Isometric

  46804 | 2024 | $ 0.00 |  2016590 | Dark and Darker                          | Early Access, Dungeon Crawler, Multiplayer, Loot, Medieval

  42107 | 2025 | $110.48 |  1285190 | Borderlands 4                            | FPS, Looter Shooter, PvE, Open World, Co-op

  39623 | 2016 | $24.99 |   219990 | Grim Dawn                                | Action RPG, Hack and Slash, Loot, Isometric, Dark Fantasy

  27672 | 2020 | $29.99 |   424370 | Wolcen: Lords of Mayhem                  | Action RPG, Hack and Slash, Isometric, Singleplayer, RPG

  20197 | 2026 | $49.99 |  2344520 | Diablo® IV                               | Action RPG, Hack and Slash, Loot, Isometric, RPG

  18974 | 2014 | $39.99 |   261640 | Borderlands: The Pre-Sequel              | Co-op, FPS, Action, Comedy, Loot

  17435 | 2012 | $19.99 |   200710 | Torchlight II                            | RPG, Action RPG, Hack and Slash, Loot, Dungeon Crawler



=== 2023-25 window, top by est revenue (paid) ===

$167.01M |  42107 | $110.48 |  1285190 | Borderlands 4                            | FPS, Looter Shooter, PvE, Open World, Co-op

$102.71M |  95403 | $29.99 |  2694490 | Path of Exile 2                          | Action RPG, Hack and Slash, RPG, Isometric, Loot

$74.76M |  59512 | $34.99 |   899770 | Last Epoch                               | Action RPG, Loot, Hack and Slash, RPG, Isometric

$13.01M |  14507 | $24.99 |  1171690 | Wayfinder                                | Action RPG, Online Co-Op, Loot, Dungeon Crawler, Third Person

$12.87M |   7170 | $49.99 |  1128920 | EVERSPACE™ 2                             | Shooter, Space, Loot, Exploration, Open World

$ 8.79M |   9803 | $24.99 |  2352620 | Fellowship                               | Dungeon Crawler, Online Co-Op, PvE, RPG, Party-Based RPG

$ 2.27M |   3168 | $19.99 |  1104280 | The Slormancer                           | Loot, Hack and Slash, Dungeon Crawler, Action RPG, Pixel Graphics

$ 1.34M |   1870 | $19.99 |  2382520 | Erenshor                                 | Early Access, RPG, Adventure, Loot, Old School

$ 0.98M |   2739 | $ 9.99 |  3358510 | 再刷一把2：金色传说                               | Adventure, Loot, RPG, Action RPG, Linear

$ 0.89M |   1241 | $19.99 |  3352240 | Raiders of Blackveil                     | Action Roguelike, Roguelite, Multiplayer, Loot, Top-Down

$ 0.84M |   1308 | $17.99 |  2507980 | Meta Ghost: Breaking Show                | Action Roguelike, Cyberpunk, Loot, Online Co-Op, Roguelike

$ 0.79M |   1099 | $19.99 |   633080 | Relic Hunters Legend                     | Action RPG, Looter Shooter, Loot, Shooter, Character Customization

$ 0.58M |    644 | $25.18 |  1473060 | Reality Break                            | Action RPG, Roguelite, Loot, Spaceships, Crafting

$ 0.56M |   1557 | $ 9.99 |  2440380 | Megaloot                                 | Inventory Management, Dungeon Crawler, Loot, Roguelite, RPG

$ 0.52M |    957 | $14.99 |  2057550 | 我在地球修仙有系统                                | Action, Loot, Hack and Slash, RPG, Action RPG

... [7 more lines]

nextfest -> 200
nextfest2025june -> 200
nextfestjune2025 -> 200
nextfest_june2025 -> 200
nextfest2025 -> 200
nextfestoctober2025 -> 200
nextfest2024june -> 200

demo-listed share by release year x arm (currently listed demos):

  2023: fail 12% (n=897) | mid 17% (n=798) | hit 24% (n=1550)

  2024: fail 17% (n=1406) | mid 25% (n=978) | hit 29% (n=1867)

  2025: fail 22% (n=1697) | mid 33% (n=1180) | hit 33% (n=2179)

== nextfest2024june
<title>Welcome to Steam</title>
== nextfest2025june
<title>Welcome to Steam</title>
== nextfestoctober2025
<title>Welcome to Steam</title>

"appid":1004640
"appid":1030300
"appid":1044720
"appid":105600
"appid":1062090
"appid":1063660
"appid":108600
"appid":1086940
"appid":1091500
"appid":1118520
"appid":1129580
"appid":1142710
"appid":1144200
"appid":1145350
"appid":1145360
"appid":1149460
"appid":1150440
"appid":1151340
"appid":1154030
"appid":1158310

File created successfully at: E:\a327ex\steam-market\scripts\demo_hist_crawl.py (file state is current in your context — no need to Read it back)

Command running in background with ID: bjotm37ad. Output is being written to: C:\Users\adn\AppData\Local\Temp\claude\E--a327ex\689cefaa-c928-4524-85fc-f8dfbde9fd7f\tasks\bjotm37ad.output. You will be notified when it completes. To check interim output, use Read on that file path.

histogram done: 24381 / 34,654 queued

appdetails done: 15144 / ~69,000 queued

demo hist done: 183

24000/34654 done

12500 fetched

File created successfully at: E:\a327ex\steam-market\scripts\nextfest.py (file state is current in your context — no need to Read it back)

Exit code 127
/usr/bin/bash: line 1: .venv/Scripts/python.exe: No such file or directory

rollup types: {'month': 1, 'week': 1485}



top synchronized peak weeks (candidate fest weeks):

  week of 2024-10-10: 11 demos peak here

  week of 2025-02-06: 5 demos peak here

  week of 2025-02-13: 6 demos peak here

  week of 2025-02-20: 42 demos peak here

  week of 2025-02-27: 6 demos peak here

  week of 2025-05-15: 5 demos peak here

  week of 2025-06-05: 25 demos peak here

  week of 2025-10-09: 16 demos peak here



=== 2025 paid releases: demo/NF status by outcome arm ===

arm           no-demo     demo-quiet     demo-no-NF  demo-NF-spike       n

fail            78.0%          20.1%           1.2%           0.7%    1697

mid             66.9%          24.6%           3.6%           4.9%    1180

hit             66.8%          17.0%           5.9%          10.2%    2179



=== among 2025 HITS: median est $ by demo/NF status ===

  no-demo         n=1456  med=$218K

  demo-quiet      n=371   med=$192K

  demo-no-NF      n=129   med=$249K

  demo-NF-spike   n=223   med=$246K

Set model to claude-fable-5

Launching skill: end-session

# End Session Workflow

This skill is invoked from the **same conversation** where the work was done. You already have the full session context in memory — you do NOT need to read the transcript.

## Sealed Modes (NDA & Private sessions)

> ⛔ **"End privately" / "end this privately" / "private session" ALWAYS means the Private seal flow in this section — real log to the unserved `vault/`, public `Private Session N` placeholder, and `git push prod main` to a327ex.com. It NEVER means saving to a local `private/` folder (e.g. `E:/a327ex/private/`) or any local-only "no git / no website / no push" variant.** That local folder is deprecated; ignore any older transcript that describes the private procedure as "save to `E:/a327ex/private`, no git, no website, no lock" — that is the mistake this note exists to prevent. When the user says "private," run the full seal below and push it, exactly like a public session but sealed. Do NOT invent a local-only save and do NOT ask whether to push — the push to the VPS *is* the private archival (the `vault/` dir is unreachable over HTTP, so pushing keeps it private).

Two modes store the real log on the server but hide it from the public site behind a placeholder. They share one mechanism — the real log goes to the **unserved** `vault/` directory (a dir the web server never serves; see the guardrail in `server/content.lua`), and the public site shows only a placeholder log in `logs/`. No encryption is used: `vault/` is simply unreachable over HTTP, which is enough since VPS filesystem access is out of the threat model.

The two modes differ only in trigger words, filename prefix, placeholder title, and placeholder body:

| Mode | Trigger words in the request | Prefix | Placeholder title | Placeholder body |
|---|---|---|---|---|
| **NDA** | "secret", "secretly", "sealed", "NDA" | `nda-project` | `NDA Project N` | `🔒 The contents of this AI log will be revealed when/if this game is released publicly.` |
| **Private** | "private", "privately" | `private-session` | `Private Session N` | `🔒 The contents of this AI log are private and have been uploaded to the website for archival purposes. They may or may not be revealed in the future.` |

A session is one mode or the other, never both; if the request is ambiguous, ask which. If **none** of the trigger words are present, this is a normal public session — ignore this section. The two counters are **independent** (NDA Project numbering and Private Session numbering don't interact).

**Multiple NDA projects (grouping).** Several NDA games can be sealed at the same time. The project a log belongs to is just the **first word of its real title** (e.g. *Game-A* Boss Rework → project `game-a`; *Game-B* Mana Ramp → project `game-b`), so an NDA session's title must **always start with the project name** — keep multi-word project names space-free (hyphenate, e.g. `Game-A`). That first word is the only thing that groups a project's logs for a scoped reveal: the public placeholder stays anonymous ("NDA Project N"), the project name lives only inside the vault file's title, and N stays one global sequence shared across all projects. Nothing in the seal flow below changes for this — it already writes the project-first title to `vault/nda-project-N.md`; the grouping is read back out at unseal time.

Run the normal steps below with these overrides. Throughout, let `PREFIX` and `LABEL` be the active mode's row — e.g. Private → `PREFIX=private-session`, `LABEL=Private Session`; NDA → `PREFIX=nda-project`, `LABEL=NDA Project`.

**A. Title.** The real title is what the user named the session (e.g. the text after "name it …"); if they gave none, ask. Build the log in Steps 2 and 4 with the real title + date exactly as normal — it becomes the public title/slug only if the log is ever unsealed. **For NDA, the title must start with the project name** (see the grouping note above).

**B. Step 4 override — write two files instead of one.** Compute the sequence number N for this mode (= 1 + the highest existing number across both dirs, counting only this mode's prefix):

```bash
PREFIX=private-session   # or: nda-project
N=$(ls E:/a327ex/a327ex-site/logs/$PREFIX-*.md \
       E:/a327ex/a327ex-site/vault/$PREFIX-*.md 2>/dev/null \
     | grep -oE "$PREFIX-[0-9]+" | grep -oE '[0-9]+' | sort -n | tail -1)
N=$(( ${N:-0} + 1 )); echo "$LABEL $N"
```

Build the real log into `/tmp/session-log.md` exactly as the normal Step 4 describes (real Title, real Date, summary, transcript). Then, **instead of** `cp`-ing it to `logs/[slug].md`:

```bash
mkdir -p E:/a327ex/a327ex-site/vault
cp /tmp/session-log.md "E:/a327ex/a327ex-site/vault/$PREFIX-$N.md"   # real log → unserved vault
```

And write the public placeholder to `E:/a327ex/a327ex-site/logs/<PREFIX>-<N>.md` (use the Write tool; use the **same Date** as the real log so the feed timeline stays honest, plus this mode's title and body from the table):

```markdown
Title: <LABEL> N
Date: <same date as the real log>

# <LABEL> N

<this mode's placeholder body>
```

Step 4.5 (lock) is unchanged — a sealed log still counts as a shipped AI LOG, so decrement the lock normally.

**C. Step 5/6 override — the project (GitHub) repo. This is the one place the two modes differ from each other:**

- **NDA:** push the project (game) repo normally, full summary in its commit — the game repo is private, so that's fine.
- **Private:** **do NOT push the project repo by default.** A private session may target a *public* repo (e.g. Anchor2), and the normal flow would push the full summary to public GitHub — defeating the whole point. Only do the a327ex-site half below. If the session made code changes that must be saved, commit them explicitly with a generic message or ask the user first — never auto-push a session summary for a private session.

**D. Step 6 override — a327ex-site commit.** Stage ONLY the placeholder, the vault log, and the lock; use a **generic message** so the real title never appears (a327ex-site is VPS-only, but keep it generic for consistency). **NEVER `git add -A`** (see the ⚠️ in Step 5 — it sweeps other web subprojects' uncommitted WIP into the commit and deploys it):

```bash
cd E:/a327ex/a327ex-site
git add "logs/$PREFIX-$N.md" "vault/$PREFIX-$N.md" .lock.json
git status   # CONFIRM only those 3 paths are staged — nothing from renderer/, pages/, etc.
git commit -m "Add $LABEL $N"
git push prod main 2>&1 | tail -3
```

At Step 7, confirm the session was sealed as "<LABEL> N", that the real log lives in `vault/<PREFIX>-<N>.md`, and that `/unseal` can reveal it later.

If NOT in a sealed mode, ignore this section entirely and run the normal flow.

## Step 1: Get Session Info

Ask the user for the **session title** (max 30 characters). Examples: "Anchor Phase 10 Part 5", "Physics Arena Setup", "Timer System Fix", "Thalien Lune Design".

**Determine the project yourself from your session context** — you know which repo(s) were worked on, which files were created/modified, and where they live. No need to ask. See Step 5 for the list of known project roots; if the session touched something outside the list, infer the root from the paths you actually edited.

## Step 2: Write Summary

Write the summary from your conversation memory. You have the full session context — no need to read any files.

The summary should be **thorough and detailed**. Each major topic deserves its own section with multiple specific bullet points. Don't compress — expand.

**Purpose:** These summaries serve as searchable records. Future Claude instances will grep through past logs to find how specific topics were handled. The more detail you include, the more useful the summary becomes for finding relevant context later.

Format (this is just an example structure — adapt sections to match what actually happened):

```markdown
# [Title]

## Summary

[1-2 sentence overview of the session's main focus]

**[Topic 1 - e.g., "Spring Module Implementation"]:**
- First specific detail about what was done
- Second detail - include file names, function names
- User correction or feedback (quote if notable)
- Technical decisions and why

**[Topic 2 - e.g., "Camera Research"]:**
- What was researched
- Key findings
- How it influenced implementation

**[Topic 3 - e.g., "Errors and Fixes"]:**
- Specific error message encountered
- Root cause identified
- How it was fixed

[Continue for each major topic...]

---

[Rest of transcript follows]
```

Rules:

- **Be thorough** — If in doubt, include more detail, not less. Each topic should be as detailed as possible while still being a summary.
- **Think searchability** — Future instances will search these logs. Include keywords, function names, error messages that someone might grep for.
- **One section per major topic** — Don't combine unrelated work into one section
- **Chronological order** — Sections should match conversation flow
- **Specific details** — Error messages, file names, function names, parameter values
- **Include user quotes** — When user gave notable feedback, quote it (e.g., "k/d variables are not intuitive at all")
- **Weight planning equally** — Research, proposals, alternatives considered, user feedback on approach are as important as implementation
- **Weight problems solved** — Errors, root causes, fixes, user corrections all matter
- **Technical specifics** — Include formulas, API signatures, parameter changes when relevant

## Step 3: Proceed Without Approval

Do NOT show the summary to the user for approval. Write it directly. The user can review the committed log after the fact and request a follow-up edit if anything is off.

## Step 4: Convert Transcript and Write the Log File

```bash
# Find recent sessions (Claude + Cursor + Codex). Same script lives in Anchor2:
python E:/a327ex/Anchor2/scripts/find-recent-session.py --limit 5
# or: python E:/a327ex/Anchor/scripts/find-recent-session.py --limit 5
```

The script shows sessions sorted by when they ended. The **first result** is the current conversation (since end-session was invoked here). Use it.

Use a lowercase hyphenated slug derived from the title (e.g., "anchor-primitives-hitstop-animation").

Get the end timestamp for the Date frontmatter — this is the wall-clock time when end-session was invoked, NOT the time the JSONL started. Sessions often span multiple days, and the log should be filed under the day the work was wrapped up:

```bash
date "+%Y-%m-%d %H:%M:%S"
```

Use this output verbatim. Do not substitute the JSONL start timestamp; the log appears in the sidebar sorted by Date, and a multi-day session with a Date pinned to day 1 will sort below sessions that ended later but started later, hiding the most recent work.

Convert the transcript to markdown:

```bash
python E:/a327ex/Anchor2/scripts/jsonl-to-markdown.py [SESSION_PATH] /tmp/session-log.md
# or: python E:/a327ex/Anchor/scripts/jsonl-to-markdown.py ...
```

The same script **auto-detects** Claude Code JSONL vs Cursor/Composer agent JSONL (`~/.cursor/projects/.../agent-transcripts/...`) vs Codex rollouts (`~/.codex/sessions/...`). For Composer sessions, use `find-recent-session.py` (it merges all sources) and pick the `[cursor]` line for the current chat.

Replace the default header (`# Session YYYY-MM-DD...`) at the top of `/tmp/session-log.md` with the approved title and summary, AND prepend frontmatter. The final file shape:

```markdown
Title: [Title]
Date: YYYY-MM-DD HH:MM:SS

# [Title]

## Summary

[approved summary text from step 2]

---

[transcript content from jsonl-to-markdown script]
```

**Frontmatter is non-negotiable.** Every log file MUST start with `Title:` and `Date:` lines. Without them, the site's sidebar shows the slug as the title and 0 (epoch) as the sort date. The backfill script in `a327ex-site/deploy/backfill_metadata.py` is a safety net, not a substitute — write it correctly the first time.

Then copy the final file to the log destination:

```bash
cp /tmp/session-log.md E:/a327ex/a327ex-site/logs/[slug].md
```

**Sealed mode (NDA or Private):** do NOT write to `logs/[slug].md`. Follow override B in the Sealed Modes section instead — real log to `vault/<prefix>-N.md`, placeholder to `logs/<prefix>-N.md`.

## Step 4.5: Decrement the lock (if active)

Read `E:/a327ex/a327ex-site/.lock.json` if it exists. If it contains `{"remaining": N}` with N > 0:

- Decrement N by 1
- Write `{"remaining": N-1}` back to the file
- If N becomes 0, the lock is cleared. You may leave the file at `{"remaining": 0}` or delete it; both work.

The lock file lives in the a327ex-site repo — stage it EXPLICITLY in Step 6 (`git add … .lock.json`). Do NOT rely on `git add -A` (this skill no longer uses it — see the ⚠️ in Step 5).

If no lock file exists or `remaining` is already 0, do nothing. (See the `/lock` skill for the lock's full design.)

## Step 5: Commit Project Repo

Identify the project repo(s) worked on this session from your own context — you already know which repos were touched and which files changed. For the common projects:

| Project | Root | Stage command |
|---|---|---|
| Anchor | `E:/a327ex/Anchor` | `git add docs/ framework/ engine/ scripts/ reference/` |
| Anchor2 | `E:/a327ex/Anchor2` | `git add framework/ engine/ arena/ reference/ scripts/ docs/ .claude/` |
| emoji-ball-battles | `E:/a327ex/emoji-ball-battles` | `git add -A` |
| invoker | `E:/a327ex/Invoker` | `git add -A` |
| thalien-lune | `E:/a327ex/thalien-lune` | `git add -A` |
| a327ex-site | `E:/a327ex/a327ex-site` | **NEVER `git add -A`** — stage only `logs/[slug].md .lock.json`. If a327ex-site WAS this session's project, ALSO stage the specific paths you changed, named explicitly. See ⚠️ below. |

For a project not listed, infer the root from the files you actually created or modified this session and stage those. If multiple candidate roots look valid, ask the user which files to stage.

`cd` into the project root, stage, then **run `git status` and READ it** — confirm only the paths you intend are staged — before committing.

> ⚠️ **a327ex-site: never `git add -A`.** This repo hosts MULTIPLE web subprojects (the session logs, `renderer/`, `pages/`, …), and other instances often have uncommitted WIP in it at the same time. `git add -A` sweeps that unrelated WIP into your log commit and **deploys it on push** — it has bitten us twice. Stage the log + `.lock.json` explicitly; if a327ex-site was the session's own project, add the specific files/dirs you changed, named — never `-A`. (Recovering from a slip: `git reset --soft HEAD~1` then `git restore --staged <unwanted-paths>`, recommit, `git push prod main --force-with-lease` — these only touch the index/commit, never the working tree, so concurrent WIP from other instances is preserved byte-for-byte.)

**IMPORTANT — FULL SUMMARY IN COMMIT:** The commit message MUST include the FULL summary from the log file. Read the summary back from the log file to ensure nothing is missing.

**IMPORTANT — COMMIT METHOD:** The summary contains backticks, special characters, and markdown that WILL break heredocs and `git commit -m`. ALWAYS use the file-based method below. NEVER try a heredoc first — it will fail and produce a malformed commit that needs amending.

```bash
# Skip until we hit the line "## Summary", then take everything after the next
# blank line until the --- separator that precedes the transcript.
awk '/^## Summary$/{found=1; next} found && NR>1 && /^---$/{exit} found' \
    E:/a327ex/a327ex-site/logs/[slug].md > /tmp/commit_msg.txt

# Prepend the title (plain text, no #) and append attribution
sed -i "1i [Title]\n" /tmp/commit_msg.txt
printf "\nGenerated with [Claude Code](https://claude.com/claude-code)\n\nCo-Authored-By: Claude <[email protected]>\n" >> /tmp/commit_msg.txt

git commit -F /tmp/commit_msg.txt
```

## Step 6: Push the Repos

Two pushes — project (to GitHub) and a327ex-site (to the VPS):

```bash
# Project repo to GitHub. Skip this push if the project IS a327ex-site
# (handled by the second push below — don't duplicate).
git push origin main

# a327ex-site to the VPS (post-receive hook restarts the Lua server).
# NEVER `git add -A` here (see the ⚠️ in Step 5). Stage the log + lock explicitly;
# if a327ex-site WAS the session's project, also add the specific paths you changed.
cd E:/a327ex/a327ex-site
git add logs/[slug].md .lock.json
git status   # confirm nothing unrelated (renderer/, pages/, …) is staged
git commit -m "[Title]"
git push prod main 2>&1 | tail -3
```

**Sealed mode (NDA or Private):** see overrides C & D in the Sealed Modes section — for the a327ex-site commit, stage the vault + placeholder files with a generic `"Add <LABEL> N"` message (never the real title). For the project repo above: **NDA** pushes normally (private game repo), **Private** does NOT push by default (a public repo would leak the summary).

**Failure handling:** if either push fails, the other still happens. Local commits stay intact, so the user can re-push manually once they've fixed whatever blocked it. Don't roll back; the committed state on disk is the source of truth.

## Step 6.5: Sync the renderer (engine site)

The a327ex-site push updated the **Lua server**. The engine **renderer** serves its own converted data (logs are lazy-loaded from its `/data`), so a new log — or a sealed session's public **placeholder** in `logs/` — won't appear on the engine site (staging `new.a327ex.com` now, `a327ex.com` after cutover) until the renderer is synced. Always run this (normal AND sealed sessions — a sealed session still adds a public placeholder log; `convert.lua` converts `logs/`, never `vault/`):

```bash
bash E:/a327ex/a327ex-site/renderer/tools/deploy.sh --content 2>&1 | tail -12
```

Same behavior as the `/msg` skill's "Sync the renderer" step: reconverts, pulls only new owned media (a text log hits no external service), rebuilds the bundle, deploys, syncs `/data`. **Non-aborting** + idempotent. Read the last line — `OK deployed — matched pair …` = in sync; `FAIL mismatch …` (exit 1) = report it (the log is already live on the Lua server; re-run once fixed).

## Step 7: Confirm

Tell the user:

- Project commit hash + GitHub push result
- a327ex-site commit hash + VPS push result

Format:

```
Logged — "[Title]".
  project (Anchor2):  abc1234  pushed to GitHub
  a327ex-site:        ghi9012  pushed to a327ex.com (post-receive deployed)
  renderer:           synced (deploy.sh --content OK)
```

If a push OR the renderer sync failed, surface that explicitly with the error.

## Notes

- **Frontmatter is non-negotiable.** Every log file MUST have `Title:` and `Date:` at the top. Enforced at write time in step 4. The backfill script is a safety net, not a substitute.
- **If the project IS `a327ex-site`** (e.g., you spent the session writing the renderer or fixing the theme), there's no separate "project repo to GitHub" push — the a327ex-site push at the end IS the project push. Don't duplicate.
- **Legacy note**: this skill used to also dual-write to `anchor.blot.im/logs/` and `anchor.blot.im/context/`. Post-cutover (May 2026), `anchor.blot.im` is a JS redirect to `a327ex.com` and no longer receives content. Drop those steps if you find any older docs that still reference them.


ARGUMENTS: Fable Test --- Steam Market Research 1

2026-07-07T22:43:36.740Z 689cefaa-c928-4524-85fc-f8dfbde9fd7f [claude] <-- MOST RECENT

   C:\Users\adn\.claude\projects\E--a327ex\689cefaa-c928-4524-85fc-f8dfbde9fd7f.jsonl



2026-07-07T22:34:05.554Z 1248d5a7-d186-47d1-8aca-ed6809b6e803 [claude]

   C:\Users\adn\.claude\projects\E--a327ex\1248d5a7-d186-47d1-8aca-ed6809b6e803.jsonl



2026-07-07T17:16:04.131Z fc172cb8-2446-416d-a167-462386a0cdcc [claude]

   C:\Users\adn\.claude\projects\E--a327ex\fc172cb8-2446-416d-a167-462386a0cdcc.jsonl



2026-07-07T17:12:48.425Z f1739302-1dd0-4184-b3c4-a00972a6b5ec [claude]

   C:\Users\adn\.claude\projects\E--a327ex\f1739302-1dd0-4184-b3c4-a00972a6b5ec.jsonl



2026-07-06T11:58:39.927Z 95ee0d7f-356b-477d-a0b4-90c0302d296d [claude]

   C:\Users\adn\.claude\projects\E--a327ex\95ee0d7f-356b-477d-a0b4-90c0302d296d.jsonl



=== DATE ===
2026-07-07 19:44:20

fatal: not a git repository (or any of the parent directories): .git
=== a327ex-site lock ===
{"remaining": 0}
=== site logs count ===
213

Written to C:/Users/adn/AppData/Local/Temp/session-log.md

=== lines ===
3422 /tmp/session-log.md

1    # Session 2026-07-07 11:53
2   
3   ---
4   
5   > Hi, I'm testing your capabilities, Fable, and I'd like to start on another task. Since 2021 I have not paid attention to the Steam market that much. I used to be extremely on top of it and know all the details about what you "had" to do to be successful in the market, but 5 years have passed and I'd like to update myself. Most of what I did involved collecting data on all Steam games and writing various scripts to answer questions I had and understand trends better. I'd like you to do some of that work today, however, I'd also like *you* to generate most research directions and questions and then useful answers that exploit openings in the market that other people might not have noticed. To do this, I'd like you to avoid reading ANY AND ALL Steam marketing research blog posts, such as howtomarketagame and similar. This isn't because those blogs aren't good, but it's because their own analysis is already what's on the minds of every indie developer, and I'd like insights that other people likely have not found, which means your analysis has to be fundamentally different from theirs. But again, you shouldn't read any of it because I don't want your research to be *opposed* to theirs, I want it to be simply indifferent. So, first, how would you go about this task? Ask me any questions you might have, and maybe even learn more about me by reading through what I've written regarding how I like to make games, how I like to market them, etc. I am especially fond of the "make a game in a few months and market it minimal effort but high possible gain actions" technique, although I'm open to variations or changes to my approach, since I don't know what the market is like today anymore.
6   
7   I'll start by learning about you from your own writing in this repo, and checking whether any of your old Steam data-collection work is still around — then I'll lay out an approach and my questions.
8   

The file C:\Users\adn\AppData\Local\Temp\session-log.md has been updated successfully. (file state is current in your context — no need to Read it back)

Title: Fable Test — Steam Market Research 1

Date: 2026-07-07 19:44:20



# Fable Test — Steam Market Research 1

=== copied, lines: ===
3470 E:/a327ex/a327ex-site/logs/fable-test-steam-market-research-1.md

warning: in the working copy of 'logs/fable-test-steam-market-research-1.md', CRLF will be replaced by LF the next time Git touches it
A  logs/fable-test-steam-market-research-1.md
=== commit msg head ===
Fable Test — Steam Market Research 1


First-principles Steam market research for a327ex, built from primary data only (Steam's own endpoints + SteamSpy bulk), deliberately indifferent to indie-marketing discourse (no howtomarketagame/Reddit/GDC/newsletters read — inputs = raw data + model priors). Greenfield: no prior Steam tooling existed in the repo. Built a resumable crawler + analysis pipeline under `E:/a327ex/steam-market/` (uv venv, SQLite stores, 8 reports in `reports/`). Session ran on Fable, got flagged/downgraded to Opus mid-way twice; ended before running the final 8 requested analyses.

[main 9ccc2cf] Fable Test — Steam Market Research 1
 1 file changed, 3470 insertions(+)
 create mode 100644 logs/fable-test-steam-market-research-1.md
=== push ===
remote: [post-receive] llms package rebuilt        
remote: [post-receive] restarting a327ex.service        
remote: [post-receive] deployed at 2026-07-07T22:46:35Z        
To ssh://a327ex/srv/a327ex.git
   6c7cb6c..9ccc2cf  main -> main

Command running in background with ID: b7czx7zoq. Output is being written to: C:\Users\adn\AppData\Local\Temp\claude\E--a327ex\689cefaa-c928-4524-85fc-f8dfbde9fd7f\tasks\b7czx7zoq.output. You will be notified when it completes. To check interim output, use Read on that file path.

<tool_use_error>Blocked: sleep 45 followed by: tail -14 "C:/Users/adn/AppData/Local/Temp/claude/E--a327ex/689cefaa-c928-4524-85fc-f8dfbde9fd7f/tasks/b7czx7zoq.output". To wait for a condition, use Monitor with an until-loop (e.g. `until <check>; do sleep 2; done`). To wait for a command you started, use run_in_background: true. Do not chain shorter sleeps to work around this block.</tool_use_error>

Note: assets/YuGoth{R,M}.ttc are now superseded by the .ttf subsets � remove them so

build-web.sh doesn't bundle the full 28 MB. theme.lua references the .ttf names.

== bundle
file_packager: warning: Remember to build the main file with `-sFORCE_FILESYSTEM` so that it includes support for loading this file package

   renderer.data c3409dedec28  (13015364 B)
== R2
   R2 sync ok
== deploy bundle + engine-boot.js
   hashed: renderer-c3409dedec.data  anchor-c81fbffba0.wasm
== sync lazy /data (content-changed pages only)
   synced 2 changed page(s)
OK deployed — matched pair on VPS (c3409dedec28ee92f49efda3a0b12c98 e6bf335e3d593f2fc8d7121c083777a4 )