GML What do you prioritize at the start of a new project? (rambling)

[fxokz]

Me personally, I always have the same method for every platformer game, make a simple movement/collision system follow it up with some badly drawn art, make a few enemies, make ways to kill them and progress, eventually redesigning levels and adding sound effects from bfxr.net and then making more levels bosses, optimizing code etc etc

Ive almost feel as if ive limited myself by being self conditioned to following these steps in order. How can I spice things up or just have something very specific to work on for a few days, how do huge companies make games in YEARS???? what takes them so long? is it the core of the game or glitches/bug fixing or is it simply making things feel perfectly polished?

(I know this topic is all over the place but I just need to get it off my chest and hear from others)

 

[jf_knight]

I always start on the menus first. I feel like working on the user interface always sets the mood for what genre the game is planed to be. After a few rough sketches of activated/idle buttons and sliders, I open up Photoshop and get to work.
Then, I transition into the coding aspect of tying functions to buttons, and so on..

 

[ajan-ko]

For me, it's more mechanical first and see how it feels using placeholder.

If I feel the mechanic won't work, I will move on to another projects.

If it's feels good, then I will move into next step alpha testing it to couple of people.

Then I will move into next step, creating assets, sprite, or music.

I've tried to create asset first then make the game without alpha testing, and it's so bad, I think I've wasted so much resource for game that doesn't work.

For the menu, I think I can copy paste them from another projects.

 

[YellowAfterlife]

I always start on the menus first. I feel like working on the user interface always sets the mood for what genre the game is planed to be. After a few rough sketches of activated/idle buttons and sliders, I open up Photoshop and get to work.
Then, I transition into the coding aspect of tying functions to buttons, and so on..

I see people saying that their flow is like that, and, like... what if you do all menus, artwork, and stuff, and then the core gameplay loop turns out to be not as fun as expected?

 

[SamuraiSnowman]

For me, its looking at what type of game I want to make (platformer,rpg,shooter,etc) and then trying to figure out ways to be completely original and different. I guess thats style. Figure out a way to make it stand out and add your own flair to the genre then go from there. And hey! adding homemade horrible art is my favorite part of the process!! ::

 

[Guest User]

movement and collisions. then i just add to that with whatever i feel like doing that day.

that's probably why i make little progress. but i don't mind.

 

[coryru]

I start off with a core concept first (platformer,shooter,etc). Then I write everything out on paper and plan as much of the mechanics there (maybe some paper level designs) after that I do a mock up if photo shop so I have an idea of the graphic style and ui. Then I plan out controls. Then I make a bare bones prototype of the main game play. If it's not fun scrap it keep a couple of the ideas for later. If it's fun I go on to making the main game play feel as tight as possible. Than back to paper to plan out the entire game.

 

[Yal]

• Core gameplay first, with just a 'press enter to go to the next room, init all global variables in the create event' title screen
• Spend like 50% or more of the dev time just playing around with the physics in the debug room
  
• If core gameplay isn't fun, create a new project and forget this one existed
• Make immediately visible stuff first (filled with placeholders and kludges) and code subsystems later when they're actually needed: I find that coding subsystems from the ground up and it taking days until you can first test play a game is a super-efficient way at getting bored at a project and wasting work at something you'll abandon.
• Once the project is underway, write down all fun ideas I get, then implement the ideas that seem like the least work, recursively. Probably the reason i've finished so few big games

how do huge companies make games in YEARS???? what takes them so long?

Project management Essentially, the amount of communication you need between team members increases exponentially with team size, and at some point coordinating work takes more time than the actual work.
Also, AAA games are trying very hard to be realistic, so a lot of work goes into polishing stuff for no real content-amount gain.
Finally, AAA games uses way too many setpiece moments, which is content that has to be custom-made for every case and cannot be recycled for other places in the game. Games like Fallout 4 has everything made using prefab modules that are made to allow for maximal reuse, and they're actively trying to reduce the amount of setpieces by saving them for last and have the level designers try doing something with ordinary tiles first.

Strongly recommending this GDC talk, it's very interesting even if you only do 2D indie stuff:

 

[Ladi_Pix3l]

• Paper sketches
• Concept art
• Core mechanics
• Menus
• Sprites and backgrounds
• Test
• Worst game ever made

How I do everything

 

[Khao]

Mechanics first, no matter what. Honestly, the way I feel about starting with anything else is... What's the point? What makes a game is the actual gameplay mechanics. Your art won't show you how good your game is going to be. Your extremely complex 100000-word storyline won't show you how good your game is going to be. Your menus definitely won't show you how good your game is going to be. Your game mechanics, however? They will make or break your project from the very start.

Not to mention, if you have done a ton of work elsewhere and you make a dramatic change to your gameplay because of something that didn't work, or a feature you invented out of nowhere, or something you discovered by accident... You might actually end up un-doing all your work. Your old animations might not work with the way your character moves, or certain elements might be hard to see in-game. Heck, maybe your pen & paper levels won't feel good to traverse with a set of mechanics you created after the level was designed.

If you have your mechanics down, you won't ever run into these problems, because things will be built with these mechanics in mind from the very start.

But I gotta admit, I put a lot of thought into animation at the same time I'm making my character move, and though I'll have the actual movement first, I'll add a fully animated character right away, even if its only a silhouette. I feel like it's a good indicator on how the game is actually going to feel. I don't mean I actually polish it to death and make it look good, I just make it move like a character. Buuuuuuuuuut that's just me, really. I don't personally feel comfortable working with boxes.

 

[fxokz]

Then I write everything out on paper and plan as much of the mechanics there (maybe some paper level designs) after that I do a mock up if photo shop so I have an idea of the graphic style and ui. Then I plan out controls.

I can never do this without getting bored right away and scrapping the whole thing

Your art won't show you how good your game is going to be.

Please............ take a look at Playdeads LIMBO and INSIDE, the games have the most simple mechanics, move, jump and interact. its the art style that sells those games..

 

[sitebender]

This might not be a priority, but it is the process I go through.

PREP PHASE
- Game idea as a jumping off point: Mega Man with swords, Dark Souls but with football, Sonic but with a plumber
- Fake screenshot is a must. You need to sell people on your idea before you ever spend time making a game. Most games call this concept art, but I like to show the game itself.
FIRST PHASE
- Importing stuff from previous projects if I need to
- Temporary art or squares
- Programming phase 1: General mechanics (2nd longest part)
- Invent new mechanic to twist that Mega Man formula
- Temporary menu
- Play & Polish (the mechanics in a few test maps)
SECOND PHASE
- Level Design (the longest part and perhaps the most grueling for me)
- Programming phase 2: New mechanics to make levels feel unique (levels have different blocks)
- Program phase 3: Realize there needs to be a mechanic to make the game quicker or enhance gameplay
- Play & Polish (over and over again)
THIRD PHASE
- Cut stuff (always cut stuff or you'll be here forever, cut stuff before you do art for it)
- Art (the third longest part if I'm doing the art)
- Music (I never do this)
FOURTH PHASE
- HUD / GUI
- Final menu if it drastically changes from the temporary
- Title screen
- Sound
- Play & Polish (over and over again)

MORE DETAILS
Overall, the level design is the most time consuming. You need ideas to keep it interesting, details to make it look interesting. A four minute level can take hours and I've become super efficient doing it. I even have systems to assign details so I don't have to.

LEVEL DESIGN
- While I'm here I might as well cover the level design steps. Each cycle is done multiple times until it feels good.
PREP before you start levels!
- You will forget and need to access your own examples easily. (example of springboard, signpost, lava trap, key + door)
- Have an easy was to access later content like warps and codes.
- Have cheat codes so you don't waste time such as "no enemies" "ability all" "invulnerability"
- How many levels or maps? Plot the environments.
- What is your ultimate length of the game or will it just be over when its over? I target 30 - 90 minutes for me to get through because with other players it will take longer.
- What abilities will you unlock and the order (this may change as time goes on)
Cycle 1: The Outline (rough draft) (takes 30 - 90 minutes for each)
- Plot before you map (what will be this environment, what will you get here, where will you go)
- Plot a course A, B, C, D, E but you will go A to C which unlocks B to get you to E which will send you to B.
- Make the playable outline with basic geometry
- Make key points like (cut scene here, boss there, need key here, find thing here, end here, start there, cool thing here)
- Play & Polish (it would flow better if I put A next to C)
Cycle 2: The Skeleton (first draft) ( takes a few hours for each for each)
- Add enemies if you have them or hazards.
- Program new enemies and or mechanics if the need for a new one arises
- Make advanced geometry to compliment enemy placement.
- What makes this level unique? (environment, mechanics, special things)
- Play & Polish (it would play better if I moved or tweaked this)
- Have a target time and adapt to hit target time such as adding to a level or cutting stuff. ALWAYS remember you can cut or move to a different later, longer level.
Cycle 3: The Pretty (second draft) (takes hours especially if I make the art)
- Art time! Make art! (I try to break this up so I'm not doing art back-to-back as that wears me down just as much as doing levels back-to-back wears me down)
- Add backgrounds
- Add things between the background and solid objects
- Add background objects (plants, tables, chairs)
- Add foreground objects (I have automated systems for this)
- Play & Polish (it would look better if I did this)
Cycle 4: The Details (final draft) (if there's a boss, takes hours to days for each)
- Add other details like lighting
- Add interactions (like signs, NPCs or whatever to elaborate things)
- Add bosses (you do it now as these can take a significant percentage of time and they could be their own development phase)
- Play & Polish (it would be cool if it had this)

There it is. Those are my priorities. I manage to hit game jam deadlines on time. Everything feels streamlined at this point since I've done it so many times for better or worse.

 

[Yal]

Please............ take a look at Playdeads LIMBO and INSIDE, the games have the most simple mechanics, move, jump and interact. its the art style that sells those games..

More like the emotional experience, I'd say... they're super-hard platformers just so you get to feel how harsh those worlds are yourself instead of just reading about it (like in e.g. the Hunger Games triology). Relevant interesting game design video:

 

[sitebender]

I can never do this without getting bored right away and scrapping the whole thing



Please............ take a look at Playdeads LIMBO and INSIDE, the games have the most simple mechanics, move, jump and interact. its the art style that sells those games..

You're right. Its the art style that sells any game with little to no marketing over even the quality of art or mechanics.

If you get bored and scrap the whole thing, you might want to aim for quicker projects like game jams to see if you can stick with something.

 

[fxokz]

You're right. Its the art style that sells any game with little to no marketing over even the quality of art or mechanics.

If you get bored and scrap the whole thing, you might want to aim for quicker projects like game jams to see if you can stick with something.

yeah... that's actually a very good idea.. because I cant really dedicate myself to anything that will require some plans ill just prepare myself for the upcoming jam.

 

[madfast]

#1 Story, Environment, Actors (who?, why?, who cares?) - My current project started with as little as 3 sentences in notepad, and 2 beers with my partners. The main character doesn't have a name, doesn't need one, he/she/it has a purpose and that's where the story starts, or ends....
#2 Capabilities / Mechanics - This has to include cost, development time, impact on art/media generation, and consider your target audience for the story. What is it similar to? Can you do the same story with a simpler dynamic? If so, then make it easy on yourself. Finally in #2 must be skill and ability to acheive the mechanics.
#3 Artistic Style - What is it going to look like? Can you acheive #2 and #1 in this style effectively? If no, change #2, if still no, change #1. I'm half creative and half engineer, so can I get some decent anime style characters on math and partial drawing skill, maybe, and if it doesn't work out (this is part #2 and part #3) then I'll cut it and make more simple characters I can manage.

Agree with many of the comments here, I think starting with a creative idea with no editing, but by the time you start doing complex commitments you better be cutting content/edit like sitebender said.

There are many aspects of mechanics that will work in any game, a button... There are things you take on when committing too much, say, I want a fully destructable environment. Does that mean you have created more work? Can you achieve this without creating more work? Maybe it's doable in your game, maybe not. Set a deadline and make it there with some features, some steps ahead. Cut when it's not feasible for you and when you find yourself spending 3 weeks trying to figure out how to make that ONE mechanic work. Those mechanics that turn into black holes for you, cut them early, very early.