It's Yir Auld Da Update 1

The Idea

For some reason I decided the game I really wanted to work on was a 2D beat em up. And, after watching Jim Trinka's Auld Da videos; I wanted it to be an Auld Da Game. So after making some basic sprites, and just to clarify; all the assets in the game are placeholder, I got to work on a prototype.

Combo System

My first attempt at a combo system didn't quite work. It was rhythm based free-flow combo system like you would find in the Batman Arkham games. Except in two dimensions. After a week of tinkering, it wasn't quite working. 

So now the Combo is a more traditional button masher. Each input is added to an array of previous inputs, which is then checked against predefined combo arrays. The system itself is complete, but there are only two combos currently. However making a new combo won't require any more coding, only making new sprite sheets for attacks and setting up the attacks in a spreadsheet.

Slow Motion

Slow motion is fun and opens up scope for more enemy types. Implementing slow motion in Unreal Engine 4 is fairly easy. There's a global time dilation variable that effects all actors in the world. Where as each actor has it's own custom time dilation variable. So to get the Auld Da moving at super speed; his custom dilation is set to an appropriate speed (five as max is fairly fast) while the world time dilation is set to one over that value. So every actor moves at one over the Auld Da speed while the player character moves at his normal value. As (1/x)*x is 1.  

Enemy AI

This is obviously still in very early development. I'm aiming to have 3 or 4 different enemy types that will have different reactions to the slow motion power so you're not just button mashing enemies to death.

The nice thing about the behaviour tree system in UE4 is that it slows down in keeping with global time dilation. So the enemy 'brain' slows down in slow motion. 

Graphics

The world of Auld Da will be 3D, but the gameplay will be locked to a 2D plane, and all the characters will be 2D sprites. UE4 sprite flipbooks can have materials applied to them. So 2D sprites can still receive 3D lighting information by using normal maps. One again, these are just placeholder assets. The normal maps were created very quickly in crazy bump. 

TODO

So next up I've got to make multiple enemy types, some basic level concepts and new animations for the combo system. Stay tuned.

Quick Dog Walking Simulator AI Lesson

Here's a short video that explains how I made the AI behaviour tree and environment query for Dog Walking Simulator.

I made the game to teach myself about the environment query system. It's clearly a very powerful system so I hardly scratched the surface this time around.

Matt

UE4 March Game Jam Submission : Irrelio

Had a great time this weekend working on a game jam submission.

The theme was 'set it free', so we had a good think and settled on an atmospheric horror game set in a maze, that requires you to lead different coloured butterflies to open gates and get out.

This is what we ended up with at the end of the weekend. I feel like we created a good first level of a game here, the atmosphere is pretty good, the maze isn't too complex and the concepts about how butterflies interact with the gates, pools and other butterflies is put across to the player in an easy to understand way, with any hand holding.

We liked the feel and concept so much it'll be getting worked on for the next few months to see if we can make it into something better.

So our list of stuff to do is:

All new models, textures and materials

The ones we are using are decent and open source, but they don't come together as a cohesive art style.

New music and sounds 

Pretty self explanatory. The music was knocked out in LMMS in a morning and the quality certainly reflects the time that was put in. And the sound effects were made in the most ghetto possible way, the foot steps where a ball of socks pushed into a tupperware container of gravel.

Failure state

Obviously a game without a failure state is a game for filthy casual scumbags. All joking aside, a failure state in a horror game is a tough thing to get right. The monsters in Amnesia here scary right up until you charged at them and saw that the textures were of a middling resolution and the animations were a little janky. (I love both those games, before you start).

Two games that get the potential horror of a failure state right is The Last Of Us and Resident Evil 4. Seeing Leon have his head cut off with a chainsaw and watching his twitching body drop to floor stuck with me for a while and had me running from those chainsaw enemies in terror. And as for The Last Of Us, well...

So that's going to be the largest challenge early in development.

The aim I think will be to make 3 or 4 different environments in order to keep the player interested in exploring the world. Making a short game of only 2 or 3 hours also keeps the player interested. It means you can introduce mechanics and environments at a decent rate without having to stretch out your content. Something that a certain triple A Alien game was guilty of.

There will obviously be updates every month or so with the hope of having a game released by the end of the year.

You can download the game we had by the end of weekend here. It's quite short, and may not be super performant on lower end machines.
Game Jam Submission Link

Matt

Dog Walking Simulator Out Now

As your first game release should be small and manageable; I took a project I made to learn about Unreal 4's AI behaviour trees and environment query systems.

It is available on android tablet devices here: Dog Walking Simulator Link

I learned a few important lessons about shipping a game here. There's always more to think about and I'm not 100% happy with what I shipped. But the point was to force myself to actually release something. 

So there you have it. A simple game where you play fetch with a wee doggy.

Matt.