Ravuya's Articles!

Freezer 2 post-mortem

Views: 6310

I really like the Game Development Magazine's post-mortem articles, and I figure I've had a long enough time now that Freezer 2 has been released to tell everyone about what I thought I did properly and what I thought I did wrong.

What went right

  1. Simple plans -- It's not like Freezer 2 is a gameplay Renaissance -- it is a game where you kill people and stuff their bodies in freezers. While this plan made me nervous about explaining the game to people I had just met (I usually just told them I "didn't know" what it would be) it works pretty well and is easy to explain. Of course, I had to make up a hokey story line for it -- I think this is my first game where aliens invade.

  2. Good tools -- While I ended up building most of the technology myself, the tools that I had really improved the experience of building games. Freezer 2 was unreasonably complex and difficult to build compared to CSRPG2, but it also has the possibility for much more complex and emotive behaviours... well, that's if everyone wasn't just shooting my precious scripted NPCs instead of talking to them. Building the Propane Injector tools really helped me build the game quickly and cleanly without the mass of masochistic cruft that I had kicking around in CSRPG2, and actually made the Freezer 2 development process more fun and easier to prototype.

  3. Scripts -- Embedded in Freezer 2 is SHilScript. I only use this for the cinema scenes, but there were various revisions of Freezer 2 in which I used scripting for puzzle missions (that NPCs would give you), nurses and scripted custom items. Unfortunately, due to the short gameplay and narrow scope of Freezer 2 after I realized I just had to get the fucker done, I never really got to take advantage of all this scripting. Glow and CSRPG3 will probably take more advantage of my stage-hand system.

  4. Guns -- It is a lot of fun to shoot the weapons in Freezer 2. Hell, I even had a land-mine gun at some point that shot explosive projectiles, which sat there until an NPC blundered onto them and died horribly.

  5. NPC AI -- While the cops aren't the brightest at trying to take you down, it is also a lot of fun to drag corpses past them to see how antagonized they can get. Alternatively, you can just shoot one of the bastards and watch him run around, screaming and bleeding. The things they say are also funny -- well, they would be if one of you sick bastards bothered to talk to them before shooting them in the face.

What went wrong

  1. Art -- Again, I am miserable at art. I tried to get some other sprites made, but that fell through due to the other person's obligations. It's all right, though -- I think it "works" well enough here. If I had better art perhaps I would've gotten more downloads.

  2. Focus -- I was massively unfocused with the game (as you can see from the large number of prototype versions I built which ended up being completely different from the final game). This vastly increased my development time, and the fact that I still wasn't happy with the final version made me advertise it less -- which is probably why it's significantly less popular than CSRPG2. At one point Freezer 2 was an action/RPG, then it was a puzzler, then it was survival horror, and at some points in development it had multiple endings (helping the aliens, killing everyone, leaving town). And that was just the design focus. I had a ton of other crap that I inserted and removed during development -- enough to have made three different games. I also ended up refactoring the final game's map to 1/4 of its original size, thus reducing the effect of any "scripted quests" I may have come up with to make the game original.

  3. Bad Cop AI -- While this is most significant in the story mode, there's also a problem with it in the survival mode -- NPCs often spawn inside walls and are too goddamned retarded to figure out just what it is you are up to. I'd like to see a version in the future that tracks blood stains and uses real pathfinding to take you down. It's acceptable, but by no means tough.

  4. Depth -- If perhaps driven by my willingness to just get the fucking game out (see Focus) I ended up factoring out a lot of gameplay and content out of the game. This is my absolute worst habit in game development, and it sickened me so much that I didn't advertise Freezer 2 beyond word of mouth and a silent insertion into the GD Showcase. It's just not very deep. A fun action game (especially since I added survival mode) but nothing that you will ever remember.

  5. Complexity Management -- I learned a lot about managing a large game system when producing this game -- trying to remember the names of enums or how to make NPCs do what I want was an enormous pain. I have since adopted a more rational naming scheme and are studying better ways to build game systems. It's likely that the Propane Injector tech (which is just a duplicate of the Freezer 2 tech) will be completely rebuilt so I can gain more performance and elegance when producing a game.

Freezer 2 was an enormous learning experience for me (as all games should be), especially considering I figured it would be a "simple" game to keep my fans happy and that I would release in two months or so. Unfortunately, my own taste in gameplay often far exceeds my ability to deliver and so I get bored and head off into new realms.

Clearly the solution for the future is more rigid design for myself, and stricter development. Taking waffling paths off in CSRPG2's rigid design gave us the sea monster, but taking waffling paths off in Freezer 2 (which already had a vague design) gave us a lot of crap I never put in the final game (and am ashamed to have wasted).

See other articles!
Tips for Writing Portable Games, Bad Pen & Paper Games, Pygame Introduction

Copyright 2003-2008 ravuya. Front page web design originally by Josh Powell.