Video summary

The Only Game Idea Guide You’ll Ever Need

Main summary

Key takeaways

Educational

Main ideas / lessons (game idea process)

  • Most game ideas fail to get made: The creator estimates ~80–90% of ideas won’t be produced.
  • A “good” idea is not just original: It must be understandable, within feasibility, and marketable.
  • Core goal: Find the overlap between: 1) the game you want to make 2) the game you can make (skills/feasibility) 3) the game the market will buy/profit from
  • Ideas die for predictable reasons—this talk focuses on fixing those reasons with practical validation steps.

Module 1: Why game ideas fail (and how to recognize the failure modes)

3 main reasons ideas fail

  1. Too vague

    • Example pattern: “I’ll make a Soulslike MMO” (broad concept; unclear uniqueness).
    • If you can’t clearly explain your game to others (e.g., at events), it’s likely too vague.
    • Fix direction (implied): clarify the core loop + differentiator.

  2. Too big / too complex for your context

    • Example pattern: solo dev aiming for “Skyrim in space” (or similar scope).
    • Even if it’s attractive, it may not be feasible (time, skill gaps, technical difficulty).
    • Fix direction (implied): scale down or choose a more attainable path.

  3. Too niche

    • Example pattern: extremely narrow genre intersections where the audience is tiny (e.g., “horror racing game with horses in space”).
    • Niche can work for a deliberate art project, but commercial success on Steam becomes difficult.
    • Fix direction (implied): widen audience via understandable framing and tags.

“Original” doesn’t automatically mean “marketable”

  • Originality alone is a red flag if it implies no validated audience.
  • The speaker emphasizes “Yoink and Twist”:
    • Take a proven genre/framework
    • Improve/modify aspects so the result is still familiar but meaningfully better/different

Validation exercise: Steam short description (300 characters)

A practical check: if you can’t write a 300-character Steam short description, you probably don’t have a clear, sellable concept yet.

What it should do

  • Use familiar genre terms (e.g., “tower defense,” “city building”)
  • Include verbs (what players do)
  • Give enough frame of reference so people instantly understand what they’d buy

Module 2: Reverse-engineer “what made it work” + define game “DNA” and emotion

Reverse engineering: identify “core DNA”

  • Analyze successful games by breaking them into components shared across related successes.
  • Examples:
    • Vampire Survivors: simple core loop (auto-attacking + movement) → many clones.
    • Phasmophobia: simple core (first-person controller + flashlight/tools + hunting ghosts), expanded through:
      • single-map start
      • co-op tag direction
      • Twitch/streamer-friendly viral appeal

Use Steam tags as direction

  • Determine which tags/theme direction best fits your core mechanic.
  • Tags act like a planning constraint:
    • same core DNA
    • different Steam-tag direction → different market potential

The “core emotion” framework (4 feelings)

Games are positioned as emotional experiences. The creator insists you should avoid mixing incompatible emotions.

  1. Power (power fantasy)
  2. Cozy (escape/comfort)
  3. Tension (fear of failure, suspense, uncertainty)
  4. Wonder (exploration/discovery “wow”)

Do not combine emotions like a blender. Mixing “cozy” and “dark tension” as your main identity can ruin both unless you can reliably support the shift with writing/team resources.

Feature list should support the emotion

  • After choosing the main emotion, features should reinforce it.
  • Example: Papers, Please (tension) → moral dilemma mechanics, guard visibility, stamping choices, and money vs. morality pressure.

Exercise: plan the “first 5 minutes feel”

  • Don’t describe features—describe what the player feels immediately.
  • Consider:
    • music tone (whimsical vs agonizing)
    • opening visuals/capsule vibe consistency
    • early player status (power fantasy should start empowered, not broken)

Build a more complete “mock Steam pitch page” (ideation stage)

Keep it conceptual, not optimized.

Steps described

  • Start from your 300-character short description
  • Write a long description with:
    • 2–3 main feature “headers”
    • ideally one GIF (not required yet, but part of the concept)
  • Optional Features tab items (controller support, customization, etc.)
  • Create a capsule mock:
    • pick a font/style (cozy/dark/fantasy vibes, etc.)
    • assemble a background/foreground concept in a capsule-sized image

Purpose: ensure all signals match the emotion/vibe you’re selling.

Early pitching to others (without shipping a game)

  • Pitch = early feedback:
    • “Does this interest you?”
  • Use your long description as the explanation fallback.
  • Timing note: don’t rely on Steam visibility yet—you generally need gameplay/trailer material later.

Module 3: Align with the market (Steam tags + competition)

Tag selection workflow

  1. Decide your likely core tags based on your idea + earlier direction.
  2. Use tag tools/sites to compare and estimate performance:
    • SteamTagHelper (mentioned for finding tag overlap between 2–3 games)
    • GameInsights and VG Insights (mentioned for tag performance checks)
  3. Use a recent window:
    • recommended: last ~3 years (avoid misleading older history)
  4. Pick ~2–3 strong tags
    • not “every niche tag”
    • avoid false confidence from correlated tags

Caveats on tag metrics

  • Tag performance has correlation/causation issues:
    • some tags may show high “monetary performance” but weaker median revenue
  • Prefer conservative measures (median, bottom percentiles) to estimate risk.

Competitor analysis (“Yoink and Twist” in practice)

  • Compare against closest competitors, typically 2–3 games.
  • Too many competitors creates blind spots and spreads focus.

Process

  • Find reviews + video reactions
  • List major complaints/pain points
  • Play briefly (about an hour suggested)
  • Create a list of:
    • what they get wrong
    • what you will improve
    • what missing QoL/mechanics players wish were present

Examples

  • Stardew Valley success vs later “take”:
    • Fields of Mistria credited for addressing pain points (notably fishing) and improving dialogue.
  • Kyoto Anomaly process:
    • read negative reviews
    • compile complaints
    • implement fixes (e.g., hint text when missing an anomaly rather than blind reset)

Stand out when the genre is saturated

If the genre is mature and hard to “out-better,” use a different angle:

  • re-skin / setting shift
  • meaningful thematic/mechanical reframing

Example narrative

  • Ever March
    • started as a generic RTS lacking uniqueness
    • reworked around a necromancer/underworld portal invasion concept
    • same underlying RTS structure, but differentiated enough to reduce direct competition

Module 4: Prototype to answer “can we build it?” and “is it fun?”

Prototype goal

The “make or break” question:

  • can you prototype it quickly?
  • does it feel fun and match the intended emotion?

VIP prototyping rules (time + assets + feel)

  • Suggested timebox:
    • 1 to 5 days (fast preferred)
    • even 2–3 hours can provide early signal
  • Use assets immediately:
    • the speaker strongly disagrees with graybox-only prototyping
    • without art/sound, you don’t test emotion/player feel
  • Prototype should include:
    • minimal gameplay loop (core mechanic)
    • basic art style direction
    • early music/sound controller (opening credits music sets tone)

Example outcomes

  • Overtime Anomaly prototype:
    • used demo scene + first-person controller
    • quickly assessed tension
    • expanded into a simple mechanic (refilling a fax machine) that validated potential
  • Virtual friend prototype:
    • built in ~2 days to test interest/novelty

Prototype red flags

  • Multiplayer first
    • generally build single-player first to reduce complexity
    • multiplayer-only designs are described as “dead in the water” initially (market loss + feasibility)
  • Custom engines at prototype time
    • avoid if it slows iteration; use Unity/Godot/etc. to test quickly
  • If single-player prototype doesn’t work, multiplayer won’t save it
  • Focus on proving the minimum viable gameplay loop.

Scoping tools (“4.2 module”)

  • Use asset packs (including free assets) to validate visuals quickly.
  • Fake stuff to move fast:
    • hardcode dialogue initially
    • avoid localization
    • simplify systems (don’t build full modular upgrade trees early)
  • Early game feel (80/20 rule):
    • particles/explosions/impact feedback create perceived fun
    • even if combat is slightly glitchy, strong impact sells the feel
  • If stuck:
    • skip the hard part and move on
    • hop between tasks instead of getting trapped
  • Be ruthless with time:
    • finish fast; long prototypes create attachment
    • if it fails, you need the ability to kill it quickly

Module 5: Testing and validation (get outside signals)

Quick publish for validation

  • Upload prototype to itch.io (with an itchio page).
  • Update based on what you actually built (scope mismatches are normal).

Gather data signals

Use:

  • itchio download counts
  • user feedback (even limited positive feedback helps validate)
  • Discord/community channels (e.g., “improve my game” submissions)
  • Reddit posts

Emphasis: showing the game matters more than describing it.

Make a simple trailer early

  • Create a short (~30 seconds) trailer once core loop and assets exist.
  • Trailer should look decent because you used assets (not pure graybox).
  • Optional: short “TikTok-style” dev clips (even if scuffed/phone-recorded).

“Hook” requirement

By this stage, you should be able to explain your game in:

  • ~10 seconds (Reddit-title style)

If you can’t communicate it after building, the idea likely still has a core problem.

Timebox the decision

  • “Give it a week”
  • Don’t grind endlessly:
    • if it’s not fun or nobody engages, kill it or modify major aspects
    • return to drawing board, but keep useful DNA

Momentum bar concept (avoid over-investing in failing ideas)

  • If excitement/momentum drops while doing nothing, you’ll abandon sooner.
  • If feedback keeps it up, the idea likely has legs.
  • Successful validation bridges toward full production (speaker mentions a future video on that step).

If new ideas appear

  • Don’t overthink them:
    • write them down quickly so your brain can “flush” them (RAM analogy)
  • Put them in a folder; return to the main project.

Final condensed checklist (step-by-step)

  1. High-level idea filter (3-circle overlap)

    • Can you make it (skills/feasibility)?
    • Do you want to make it?
    • Is it sellable/profitable in the market?

  2. Diagnose why it fails

    • Too vague? (can’t explain it)
    • Too big/complex? (not feasible for your context)
    • Too niche? (audience too small)

  3. Marketability test

    • Write a 300-character Steam short description (include verbs + known genre references)

  4. Identify “core DNA”

    • Reduce to the simplest repeatable mechanic(s)

  5. Choose a single main emotion

    • From: Power / Cozy / Tension / Wonder
    • Ensure the game supports that emotion consistently (especially early)

  6. Create mock Steam assets

    • Short + long description (2–3 feature headers + GIF concept)
    • Optional feature tab
    • Capsule vibe mock (ensure capsule matches emotion)

  7. Choose 2–3 Steam tags

    • Use tag overlap and tag performance sites
    • Look at recent years (~last 3 years)
    • Use conservative metrics (median/bottom percentiles)

  8. Competitor analysis

    • Pick 2–3 closest competitors
    • Read reviews + watch gameplay
    • Compile complaints and decide what you’ll improve

  9. Prototype (1–5 days)

    • Use assets + music + art direction early (not graybox only)
    • Build the minimum viable gameplay loop
    • Test single-player first
    • Avoid slow tech risks (no custom engines early)

  10. Validate

    • Publish on itch.io
    • Share with community (Discord/Reddit)
    • Create a short trailer
    • Ensure a strong hook (can explain in ~10 seconds)

  11. Decide within a week

    • If it’s fun and people respond → proceed toward production
    • If not → revise major direction or kill and restart from earlier DNA/emotion/tag stages

Speakers / sources featured

Speaker

  • The course creator / narrator (described as “the main ideas guy of Bite Me Games”).

Named individuals (as sources/examples)

  • Thomas (example team member behind “Ever March” and its early generic RTS version)
  • Chris Sukowski (told the ice cream vs steak analogy)
  • Chris (referenced as the person behind that analogy)

Games referenced (as examples / case studies)

  • Vampire Survivors, Vampire Survivors clones (“survivor-likes”), Phasmophobia, Unpacking, Celeste, Stardew Valley, Fields of Mistria, Papers, Please, Elder Ring, Kyoto Anomaly, Overtime Anomaly, Ever March, Age of Rot, Unicycle Pizza Time, Forge Industry, Dungeon Inc., Overwatch?, Exit 8, Shinkansen Zero, Nine Kings, Worshippers of Cthulhu, “alchemy game” (general reference)
  • Hypotheticals/genre references: “Skyrim in space” (hypothetical), “Age of Mythology/Age of Empires” (genre references)

Websites/tools referenced

  • steamtaghelper.com
  • GameInsights (spelled “gameytic” in subtitles) and VG Insights
  • itch.io
  • GPT (ChatGPT) (referenced as ideation/prototyping aid)

Brands/platforms referenced

  • Steam, YouTube, TikTok, Reddit, Discord, Gamescom
  • Unity (engine option), Godot (implied by “GDau” in subtitles)

Original video