Wednesday, January 27, 2016

Making a Playable Demo map

As you may have red between the lines, “we” (not much help at the moment :( ) are making a playable demo. Finishing T22 in the current setup will take as long as traveling to the closest neighbour stellar system on a child’s tricycle, so that's obviously not an option. What we can do about that? Make a playable demo, hope you like it, then launch a Kickstarter campaign, fund/gather artists, cross fingers, and kick-ass.

But to let people open their wallets, you'll need to impress first. So, a playable demo. Anyhow, I wasn't planning to write about future plans and strategies. Let's focus on the fun part instead; making game-maps! 
I love painting maps like this in MS Paint. Too bad MS Paint is broken since Win7.


Of course we made some maps before, for the demo movies. But those were usually small, isolated parts, and weren't built for real gameplay. This time, I have a complete (but not too big to avoid endless development) map; a real piece of the actual game. Or at least, we're busy making that.

Easier said than done, of course. And I'm not only talking about writing code or doing the audio-visuals. Talking about the actual game-design and mapping itself. Since this is the first "real" map -meant to be played-, a lot of practical issues come around the corner. Stuff you may not directly think about, when coding, drawing or modelling your next game.

When making a game, or a small demo in this case, there should be some goals, asides from just finishing the darn thing. If you were about to give this demo a try, what should it do to impress? Exactly, a horror game. Scare you. Or better said, since T22 is not really about jump-scares, make you feel very uncomfortable, yet anxious to know what will happen next (after the demo). That should be our main goal. But there are a couple of sub-design-goals as well, on a more detailed layer.


Size does Matter
For one thing, the demo shouldn't be too long, nor too short. 30 .. 45 minutes would be acceptable, 1 hour or even a little bit more, would be perfect. In general I hate to pay 60 bucks for a 6 hour game, but we shouldn't forget this is a demo. There are a few ways to accomplish this “duration” aspect. First option is to make the environment so goddamn big that it simply takes THAT long to get across. Free-roaming games like GTA, Zelda or Fallout rely on large maps. But also linear "corridor" shooters like Half life or Call of Duty need plenty of space to offer some lengthy gameplay. Or very difficult battles that keep you pinned on a certain position. It's not a big surprise though, that, as game are getting easier and the visual content more beautiful (= more work), the length is getting shorter. Most of these shooters are 5, maybe 10 hours at most.

4 hours, 40 hours... Quality also counts. You don't finish a game that takes 100 hours, but has a moldy cheese gameplay. And otherwise I can highly recommend you to play this:
If you're all about lengthy games, this is the shit

Another (better) way to extend the quantity, is to make it replayable. Take Tetris or a Wrestling game. Usually only one stage (maybe a few variants), yet very replayable. The wish of any game developer. Make a minimum amount of content, and get the max out of it by making the game very addictive. A good linear shooter which takes only 9 hours, but is worth a replay, will still give you ~18 hours of fun in the end.

A third option is to make the game Very hard, and/or do a lot of backtracking, so you can basically recycle the environment. Sounds very green. A game like Resident Evil (the old ones) didn't have a massive map, if you simply count the amount of rooms or walkable square meters. Yet it did take some time, as the puzzles were pretty difficult, and the game involved a lot of backtracking. “Stinky balls, forgot the screwdriver in my limited 6-slot inventory to solve that puzzle! Back to the storage trunk, and watch out for zombies. Better take a safer route, as we only have 3 bullets left.” That kind of work.


Beauty, Smart, or Big?
What would be a best-fit for Tower22? Option1, just throwing lots of rooms and corridors on the table, will be very hard. In multiple ways. As you know, unfortunately, pushing out high-quality content with the speed of a mini-gun, isn't our talent. More maps = more work, simple as that. Also, it will be a creative challenge. Since Tower22 takes place inside a, yes, Tower, you can't rely on stretched jungles or rocky mountains. Although a skyscraper has the advantage of being tall, I must say.

But even so, more rooms doesn't make the game more interesting automatically. What could you expect in a tower? Uerhhh... rooms? Corridors? Toilets? Balconies? Elevators? Stairs? Basement rooms? You see, the available options are limited. Exploring 4 apartment rooms will be interesting. Having to explore 400 apartment rooms will become annoying. The trick for us is to make each corridor or room a bit different. Different shapes, different furniture, another wallpaper. But imagine
if you had to make skyscraper with 400 unique rooms. Arh!!
Can't blame them since their game-world is huge, but it's pretty amazing that games like Fallout4 or Crysis3 can get away with recycling the same wall-textures, furniture assets and junk-props over, and over again. You'll see the same sofa's in every house (and there a lot!). Then again, As an action-game, truly unique environments are less of a requirement compared to Tower22.


Well, I can assure you Tower22 isn't only about corridors and rooms. The environment gets more bizarre as you delve deeper (or climb higher, as you wish). And who cares about realism? Still, clearly we have some boundaries, as we have to respect the tower-theme. Making the same type of rooms over and over again, AND making them scary(!) as well, is kinda impossible. Better have a smaller floorplan, and keep some doors locked. Then put some extra love and horror-spices in the available rooms.

However... the game can't be too small either! This is not a linear corridor shooter! Yeah, corridors sure, but forget the words linear and shooter. The type of gameplay involves some monster-chasing, exploration, and puzzling. If the environment is very small and/or linear, the exploration and chasing part won't be very exciting either. The map requires to be a labyrinth, with enough secret places and spaces to hide. How to find the balance?



Braincrackers
As I'm making the first "dummy" (prototype - to be replaced / pimped) maps, I really wonder how much gameplay they will offer. And if they complement the "labyrinth / exploration / scary" components. Maps aren't just random (scary) ideas, they need to forfill those essential parts! And it's very hard to verify that. If you make a World War II shooter, there is plenty of "how-it-should-be-done" reference. Watch “Saving Private Ryan”, play “Hidden & Dangerous”, read “With the old Breed”. But a game like this... Amnesia a little bit maybe, though it's a different environment to begin with (and I never really played that game. Too scary :p).

The labyrinth, hiding and searching part is somewhat doable. Pick a piece of paper or start MS Paint, and make some floorplans. Ensure there are multiple-routes, hard-to-find locations, and think where and how you can block certain paths. Need a key first, need a rope to reach that platform, need a crank to close the bridge, etc. I found this also kinda difficult though. Tower22 is meant to be scary, and therefore somewhat serious. Where games like Zelda or Monkey Island can implement fantasy or ridiculous puzzles ("slap the pirate with a wet towel to get him out of the way"), I'm a bit stuck on real-life scenarios. DoorX needs a key. DoorY too. DoorZ a swipe card... beh, boring. And predictable.

Remember Option3 to extend game-length? Making the game hard? Well, that's another issue. Tower22 should be difficult. But when it comes to combatants... there are none! Or very little at least. You aren't safe in that building, but you won't be throwing Molotov cocktails every 10 meters either. For most of the time, puzzles are supposed to "stop" you. As well as to entertain you. Nothing
wrong with a good puzzle, but if you're only backtracking to find stupid keys the whole time... Yep, the puzzles need to be more creative than just that. Switching my head from rational logic, into weird ingenious braincrackers that can be implemented into a horror skyscraper, is very difficult. In fact, I would love to have somebody helping me on this (so if you have your head filled with ridiculous logic...).
Why do people ALWAYS lose their cranks in games? Why cranks? Honey, can you open the front-bridge? I lost my crank at work.

But other than putting ideas on paper, you need a poor unprepared Beta Tester to check if the puzzles aren't too easy/hard/boring, and if those monster chases work out. Of course. Only problem is... You need a completed map for that first. And if the Beta tester says it sucks, you can redo the whole thing. Fortunately, puzzles or core mechanics like shooting guns or running for a monster, don't require a 100% finished environment. That's why I'm making prototype maps first! Not that it still doesn't take a crapload of work to make those, but at least dropping ugly unfinished maps will hurt less than perfectly looking ones.

Making beautiful maps first, and test them on gameplay later, isn't a wise thing to do anyway. Not only a potential waste of time in case the content doesn’t make it to the final product; you will also get biased. If you put a lot of craft into a map, texture or model, it will stick to you as glue. As you hate to drop your work, modifying maps will mean that you still try to implement what has been done already, even if it really just doesn't fit in the game. Be prepared to throw away "good ideas" like used diapers.



Cabin fever
That introduces yet another problem. The Tower22 main goal is to make you wet your pants. If it does a real good job doing just that, weaker gameplay elements like dumb puzzles, shorter game-length, or stiff controls are forgivable. Man, Silent Hill isn't exactly a *fun* game either. You can't see shit most of the time, and slapping numb bodies with a wooden stick feels really awkward too. But, the game scares you, and the plot is disturbing. That's what you paid for right? Not for boobs, laughs, or slick machete action. You don't buy tickets for Star Wars to see a romantic comedy either. But, how to test if something is scary?

First problem is that, as a developer spending many hours and knowing every triangle and scripted bit, it’s very hard to get spooked by your own product. C'mon, you already know what will happen. And knowing the technical part (and shortcomings), you look at it from a very different perspective. If you and I see a dead body, we're probably like "gross! ". While those CSI dudes are more like "hmmm, hit multiple times by a blunt weapon on the cranium".

But moreover, small maps, big maps, little monsters, huge monsters, that stuff doesn't directly make a game scary. Fear is something... something unexplainable. I mentioned T22 shouldn't rely much on jump-scares. So, buckets of blood and critters jumping through windows aren't going to help us. The key lies within well-done audio-visuals. Jumpy ambient sounds, a bone-chilling soundtrack, corridors with dirty worn plaster walls, an angry lady looking at you from an oil-painting, a claustrophobic atmosphere, unexpected sight-seeing’s... Audio-Visual matter very much for a horror game. Half-finished, untextured ugly looking prototype maps just can't simulate this. Every missing detail or broken bit can completely ruin the experience. In other words, you need a finished environment to test if it's scary. And throw the whole thing away if it isn't. Not exactly effective.

Now (older) Silent Hill games weren't known for superb fancy graphics. But they were consistent. In fact, the foggy, dark-contrast "gritty" visuals made the game so... nightmarish. But if they pulled up the fog-curtain, it suddenly wouldn't be that scary anymore, probably. Consistency my friends.


Need stuff
All in all, putting down the maps is a challenge. The first thing that struck me, are the proportions itself. I made a 80 x 80 square meter boundary for this playable demo. Sounds reasonable for a tower shaped building, right? Fortunately with the new Engine22, I can hit play any time, so the editor launches the real game.exe so you can "walk" (erh, shove like a sturdy bulldozer) through your maps. But using real-life proportions (like a 150 cm wide corridor, 205 cm tall door, or 60 cm deep kitchen block) doesn't automatically result in good results. I was afraid of making the rooms unrealistically large, and the corridors so long that walking through them becomes a boring pain. But testing the actual game, it seems to be the contrary. Nice architectural details like pillars, small height differences or hidden corners are easily overlooked as they are too small. In a matter of a second you already passed them. And it took me only 20 seconds to travel from A to B, which is about 30% of the playable demo area size already.

But before simply stretching the maps, I should note that size can be deceiving. Ever entered an empty new house? Rooms without junk and without measurable reference objects like a table, door, or another human, may appear small. Also bad UV mapping (huge bricks stretched over the wall) will downsize the room visually. Having no obstacles, and a walk-speed being too high, doesn't help either. And of course, tweaking the camera FoV angle –thus having a different perspective- may enlarge or shrink the room. Simply putting the camera 20 centimetres lower ("Toddler view" as one of the guys here once said about the T22 "Radar Demo" ) already makes a perspective difference.

Yet again, it's pretty hard to make conclusions based on empty, unfinished rooms. I believe that simply by doing it, the experience will grow. A list of do's and don'ts, and some well-tuned references should roll out. I'm not looking forward to stretching up the whole demo area though, as all the rooms are connected somehow and packed densely in the available space :(

Expected this corridor to "feel" a lot longer. Is it the unfinished look? Should I increase camera FOV? Slow down the walk-speed? Or do I really have to re-map the damn thing (and shift all neighbor rooms as well)? That's the kind of stuff you probably didn't think about when generating terrific horror-ideas in bed for your game!