GoWind Game Radar is the "development quantitative brain" for small and mid-sized game studios: using objective community big data, streaming viewership, rankings, and competitor review data, it computes the next development direction, hit concept, and market gap in the overseas game market — addressing the pain of "strong execution, but idea selection by gut feeling".
25 data sources merge into hourly "game candlesticks"; six R&D signal types are auto-scanned and pushed; the niche-survival scan and same-tier survival math answer the two questions small studios care about most — "where do we stand" and "can we survive there"; the greenlight sandbox turns signals into "which games to build, how much to invest in each, and when to start".
The same product paradigm as GoWind Quant and GoWind Trade Radar — "collect → data foundation → signals → backtest → decision → feedback loop" — applied to game greenlight decisions.
- 25 sources, 24 credential-free: Steam (concurrents/charts/reviews/upcoming), App Store top charts, Twitch, Reddit, YouTube, itch.io and Game Jams, Discord, Roblox, TapTap, Bilibili, Google Trends, and more — collecting out of the box
- Six signal types covering every greenlight judgment: genre momentum, cross-platform arbitrage, competitor pain points, trend forecast, competitor downswing, UA poison
- Backtest first, use later: parameter-grid backtesting for momentum/downswing/arbitrage strategies, with hit rate and payoff multiples verified up front
- A small-studio lens: the niche-survival scan answers "where do we stand"; same-tier survival math answers "can we survive"; the sandbox's single-bet mode answers "should we go all-in"
- Deep player-sentiment mining: negative-review clustering yields pain points and micro-innovation hooks; delight-point deep-dives extract competitors' core experience loops and delight maps — pain points tell you what players curse; delight points tell you why players can't stop
- Sharper with use: signal annotation → accuracy stats → entity trustworthiness ranking, with the feedback loop built in
| Feature | Description |
|---|
| 25 sources covered | Steam (concurrent players/official most-played/news/reviews/CN bestsellers/groups/upcoming/pricing & discounts), Apple App Store free + grossing charts (CN/US), Twitch, Reddit, YouTube, itch.io rankings + Game Jams, Discord community size, Roblox concurrents, Epic free games, TapTap, Bilibili game zone, HackerNews, Google News, Wikipedia pageviews, Google Trends interest index, and more |
| Page-managed game library | You decide which games to monitor — each entity carries cross-platform footprints (Steam AppID/Apple track id/TapTap/wiki page); saved settings take effect immediately |
| Source operations | Per-source health status, last success time, and freshness badges, with automatic alerts on consecutive failures |
| Feature | Description |
|---|
| Hourly "game candlesticks" | Raw data validated/normalized/hourly-aggregated into 13 metric time series: concurrents/rankings/mentions/rating/community size/search interest/current price & discounts, and more |
| Long-term retention | Aggregated time series are retained long term, so backtests and reviews always have history to draw on |
| Signal | Question it answers |
|---|
| Genre momentum | What's starting to be discussed |
| Cross-platform arbitrage | It's hot on PC but nobody's making the mobile version yet |
| Competitor pain point | What are players complaining about |
| Trend forecast | Where will this genre be in 6-12 months |
| Competitor downswing | Who is slipping — where's our window |
| UA poison | Who is burning money to climb the charts — false prosperity, stay away |
Signals are deduplicated on disk and pushed to Feishu / DingTalk (with signing) / Slack / Telegram / email / webhook, with page-managed channels routed by type and severity.
| Feature | Description |
|---|
| Backtest tuning | Momentum/downswing/arbitrage strategies × 9-point parameter grid heatmaps, hit rate + payoff P50/P90 — strategies are validated on history before use |
| Pain-point reports | Negative-review clustering (pain index/share/typical quotes/micro-innovation hooks) |
| Delight-point deep-dive (AI) | One-click extraction of competitors' core experience loops, delight maps, player sentiment, and lessons for your own greenlight |
| Fake-ad concept tests | Validate genre momentum before committing: register concept creatives → AI creative brief → backfill impressions/clicks per channel → auto CTR comparison |
| AI analysis buttons | Signal interpretation/pain-point analysis/backtest review/watch brief, one click inside the page |
| Feature | Description |
|---|
| Opportunity scorecard | Pain point × downswing × arbitrage joint scoring (0-100 entry-window score) for cross-project scheduling |
| Greenlight sandbox | Enter total budget and planning horizon; get "which games, how much each, when to start" — portfolio/single-bet modes, UA poison auto-excluded |
| Niche-survival scan | A 0-100 survival score per genre root: demand heat × supply gap × head-brand breakability × small-team-won evidence |
| Same-tier survival math | Budget → break-even units → break-even review count × same-tier benchmark distribution → CLEAR/TIGHT/BELOW verdict |
| Decision weekly | Scorecard top/UA competition/genre landscape/pain-point moves in one report, with a dedicated "small-studio greenlight lens", headed by an AI editor's note (publishing-director persona): who to【launch now】and who to【avoid】this week |
| Competitor release calendar | Competitor launch timelines — supply floods at a glance |
| Feature | Description |
|---|
| Radar-context chat | Streaming chat that can attach the current page's data to the conversation |
| Natural-language queries | Ask for business data in plain language, guarded by a table allowlist, read-only transactions, timeouts, and row limits |
| AI morning report | Daily auto-summary of yesterday's metrics/signals/pain points, synthesized by LLM and pushed |
| Knowledge-base RAG | Document chunking and vector retrieval; chat answers cite the knowledge base automatically |
| Provider management | OpenAI-compatible protocol — DeepSeek/OpenAI/Qwen/GLM all work; page-managed with encrypted credentials |
| Feature | Description |
|---|
| Signal annotation | One-click useful/false-alarm labels, per-type accuracy stats and false-alarm attribution |
| Entity trustworthiness ranking | How accurate a genre's historical signals have been, visible at a glance |
| Data export | Signal streams/metric time series exported as CSV/Excel |
- Small/mid-studio greenlight reviews: survival score × survival math × sandbox — turning gut-feeling greenlights into debatable, data-backed decisions
- Indie teams choosing genres: emerging terms and survival scans find the gaps "too small for the big studios, but with rising demand"
- Publishers evaluating candidates: use same-tier survival math to assess a partner team's survival odds
- UA teams: UA poison identifies money-burning chart climbers — don't pour ad budgets into a red ocean
