Fairy Maid Scrap Attack design doc

Table of Contents

1 General info about the game and mechanics.

The game is heavily based on the SCRAP ATTACK mission in Blue Revolver: Double Action.

https://youtu.be/ue16e_Nosdg

Basically: Game where you have to quickly clear a path through 'scrap' that explode into bullets when destroyed. Some scrap causes chain reactions when destroyed and have to be carefully managed to not spawn overwhelming amount of bullets. Some scrap causes chain reactions when not destroyed in time.

2 Design

This is all mostly based on speculation/opinions/assumptions/intuition. Feel free to challenge the assumptions etc and make new suggestions. Everything will have to be verified by playtesting to see if it is fun. Still need reasonable mental models of what makes games like this fun in order to save time, since playtesting is slow and hard to interpret without a model.

2.1 Dynamic difficulty

Since most of the bullets are spawned by player actions (destroying scrap), the player can choose their own difficulty by being careful with destorying scraps. To not make the dodging trivial, there must be pressure on the player to destroy a lot of scrap at once to create interesting dodging situations. Scrap attack in Blue Revolver have two main "drives" that push the player towards destroying lots of scrap: time pressure to clear a path before the next wave starts (survival) and using the piercing shot to cash in on combo (score).

2.2 Risk and anti-frustration

Since the player is encourage to take on risk for score, it is entirely possible that they'll frequently misjudge and kill themselves. We can soften the consequences of bad decision and make the game less frustrating through a number of means.

  • Bombs (or other resources) the player can spend to get out of trouble
  • Artificial slowdown when many bullets are on screen
  • Cap to the number of bullets that can be spawned (may also be important for scoring system)

2.3 Infinitely looping, Single-screen arcade design

For single screen shmups like Centipede, Space Invaders, etc, the goal is not to beat some distant final boss. Since there is no distant goal to motivate the player, the gameplay needs to be engaging already from stage 1. Thus, we want to avoid an aggressively scaling scoring system where the bulk of the score comes from later stages, as it makes early-game gameplay meaningless. If anything, the ideal for single stage arcade games is for the early stages to matter the most, since the difficulty is low enough that the player can truly play to maximize score.

Some players may still be motivated to clear as many stages as possible. It is thus beneficial to tie gaining lives to scoring, since it will encourage more players to engage with the scoring system.

2.4 Scoring system ideas

2.4.1 Suggestion 1

  • destroy as much as possible with a single piercing shot (some multiplier depending on how much you destroyed quickly?)
  • dodge the bullets you created
  • bombing turns bullets into more piercing shot ammo, so taking risks and creating a ton of bullets and realizing you have to bomb can be extra rewarding.
  • scoring gives lives, and dying refills bombs.

Point 1 encourages risk, which is good for making the early game interesting. Bombing for score/ammo when many bullets are on-screen makes the game less punishing since what'd otherwise be a mistake (generating many bullets) can be rewarding without being risky. On the other hand, if this is important for scoring, it may make the late-game more rewarding than the early game if more bullets are spawned, so need to be careful with that.

It is better to give lives than bombs, since it'll make the player often be in a situation where they don't have bombs and have to try their hardest to dodge. May not need any way to gain bombs except for dying.

Can have generous death bomb window - or not. Slight preference towards not since I'd like the player to bomb willingly for score rather than save bombs for survival.

2.4.2 How to increase importance of early game

Could give bonuses for full-clearing a wave before it automatically advances to next wave. If you get more bonus the faster you are, earlier stages having a larger time limit would intrinsically award more bonus score for early waves.

  • This also encourages fast clearing (riskier) in the early game.

Do not ramp up number of bullets as the game progresses, instead make the game harder in other ways.

  • Make it harder to avoid chain reactions
  • Lower time limit
  • Add blocks with more HP or that cannot be destroyed at all. (Otherwise, it is always easy to clear a path quickly by using piercing shots)

Make late game unsustainable so that the game in practice has a finite-length.

  • Bombing to cancel bullets don't give enough score back to give players enough extends back.
  • Cap the amount of piercing ammo you can carry
  • Cap the amount of extra lives you can carry? (I don't like this. Being able to accumulate life in the early game makes the early game more important.)

2.5 Story ideas

Fairy maid tasked to clean up waste from Patchouli's magical experiments

Why are you forced to clean up waves with strict time limits? Idea: Sakuya knives you if you're too slow.

  • Could be made into an explicit mechanic instead of just fluff: you have to end waves 'manually' by moving to the top of the screen or possibly pressing a button to advance to the next wave, and if you're too slow, knives start spawning. Could let more skilled players 'milk' a stage for longer if they are comfortable dodging more knives.

2.5.1 Other characters?

Story would have to be mostly told through gameplay so hard no dialog and they'd need art assets. Could just be other fairy maids with different hair color (so a simple palette swap). Nature of characters fully conveyed through art since no dialog. So title screen art. Character selection screen art (do we even want such a thing? multiple shot types?). Possibly in-game sprite but harder if simple palette swaps. Design around clear architypes to convey characteristics without dialog?

2.6 Additions compared to Blue Revolver "Scrap Attack" mission

Needed to justify why people should player our game instead of Blue Revolver.

  • An obvious advantage we have is that we are not tied to Blue Revolver's scoring system: we can make scoring system that acknowledges the wave-based nature of the game.
  • Additional types of scrap.
    • Indestructable scrap that must be navigated around.
    • Books (like EoSD stage 4) that spawn bullets without the player doing anything?
    • Scrap that drops ammo (could have more HP to make them a time investment, or hide them among blocks you don't want to destroy or can't destroy - want a way to make them not trivial to grab.)
  • Good art / production values? <- This is hard… ;_;
  • Boss waves? A single short spellcard every 5 waves or so for variety? Don't want long bosses as it breaks flow. Should be noticeably shorter than most Touhou spell cards, maybe just a few seconds of dodging. (The boss can just be a recolored EoSD stage 4 book or something.)
  • Actual end of content (like a final boss)? Might attract players that are more motivated by seeing an ending. We can still loop the game forever. Softens up the single-screen aspect a bit for perhaps wider appeal.

2.7 Art

Need something like:

  • 1 infinitely scrolling background.
  • 1 shmup player sprite (lightly animated)
  • A few different types of scrap. Could be recolored with shaders for even more variety.
  • Title screen.
  • Game over screen.
  • High score screen
  • Tutorial screen
  • Characters select screen? (Not sure if needed)

Many of these should not need fully original art. The different screens can just use rearranged elements from the title screen for most of the screen space.

2.7.1 Art style

Obviously should enhance the mood we want the game to convey and not interfere with gameplay.

Not clear what this means in practice - there is probably a lot of freedom here and something I'm happy to let graphics maker(s) decide.

  • Do we want a seperate artist for title screen + game over screen, etc art?

2.8 Sound design / music

Settle on after art style to enhance mood we want to convey. We need like 1 BGM track so it's less time sensitive than the art. SFX can probably be stock from opengameart etc. Could use voice synth like SynthV if we want voiced lines. I have used it in the past and it's easy.

More BGM might be desireable if we want to do funky things like boss waves. Those songs wouldn't need to be long: the main point is to provide contrast with the normal music to indicate that something abnormal is going on.

Author: Dåli (create2019.itch.io)

Created: 2025-04-10 Thu 23:34

Validate