Any tips for final university project?

megalmn2000

Hi guys!
I'm finally at my last year of university, hoping to find an internship and a job soon.
I'm currently working on the pre-production phase of my final university project, ( which consist to make a game ).
I would like to ask you guys if you have some tips to recommend?
Any workflow (Agile, Scrum, etc...), software (free or commercial), good habits, deadline, project management app, SVN, tools (Skype, MSN, AIM), etc...

Our current process:
- team of 4-5 members (1 designer / programmer, 2-3 modelers, 1 animator)
- 4 months of production
- CryEngine 3 SDK for our main game engine
- PC is the target platform

Thanks A LOT for your support!

Dan!

I don;t know the skill level or work ethics of your group members but I have some general advice. Keep your scope relatively small. Example If you are doing a puzzle platformer restrict yourselves to something like 2 polished levels and maybe ~4 enemies +boss. I'm sure the programmer needs to meet certain programming goals like score and level loading and progressing stats so that should be enough to satisfy that criteria. It also gives the artists time to focus on making better art for portfolio content (also if the game itself turns out to be a stinker due to whatever may happen at least you have something to show). Good luck!

Noodle!

I completely agree with keeping the scope small.

Definitely use Scrum, as it's a great process. You can modify it to fit your team. Tools like SVN or Perforce are essential as well, you really don't want to be in a situation where you lose data that is not backed up.

You said Skype, MSN etc. Will you be sitting separately? I'd strongly advice you to try to sit together as a team.

For a small project I'd recommend you to plan your game quickly using placeholders, getting the final game design down as soon as you can. Playtest with placeholders. Then put everything in and iterate on the assets. That way you'll know quite accurately how much you have left to do.

Deadlines are good, work in sprints or deliveries. It'll give you focus throughout the project.

And a hard one is to try to keep a coherent art style. Even if your individual models look great they can clash pretty hard if they've got distinct styles. This might be a no brainer, but I thought I'd put it out there.

Good luck!

chrisradsby

Yeah keep it small, make sure to create a game that works before even thinking about adding more content ie Make sure gameplay works in a small environment before continuing. Let the designer and the artists cooperate to ensure a modular workflow for quick environments.

Try to get someone to focus on textures, making sure the fit together so that the art-direction is the same. I don't really think you guys will be able to set up and use perforce but try to figure out some way to make sure that you always have backed up data.

At the end of the workday, make sure everything runs, works fine and maybe back-up the stuff so that you guys don't lose like a week of worth of progress.

Look at other small project games, we have some schools in Sweden that create games in just a couple of weeks as well. Link here:

http://www.futuregames.se/student-project-archive/

Check that out maybe, they do it in UDK but maybe you guys can find some tips or something there.

Good luck and make sure to post your progress or questions here

glottis8

Yea... once more. keep the scope of the project small. The smaller the nodes, the more polished it will feel like. Specially if you keep things to a confined area and less open world. Things like that can be mocked to have it appear as a big env, but its very linear and constricted.

The scrum is very useful, you will know what other people are working on, if people are blocked by something they need, or thoughts on things one might need to think about.

Lots of communication all the time.

Good luck tho!

Farfarer

Yep - small scope. Can't push that aspect hard enough. The rest of the stuff is from personal experience of the same sort of uni module.

I'd recommend you devise a single gameplay feature/gimmick that is unique and that you can "bolt on" to a well established genre or game's mechanic. That means you can fall back on genre conventions when it comes to the design and controls of the game and allow you to spend your time focusing on fleshing out the unique feature - which is the thing that will get your game noticed.

Puzzle-oriented or multi-player/deathmatch games will likely be your best bet of getting something clever, accessible and able to be contained in a small amount of enclosed levels (thus giving you a finite and reasonalbe amount of artwork to do and gameplay to polish).

Prototype everything in primitives first. Get the gameplay ironed out then worry about the art.

Any time estimate you make of how long something will take - DOUBLE IT OR TRIPLE IT. Just assume outright that everything will always take twice as long as it's estimated to. Then your schedules might end up half-way realistic :P

CryEngine3 is probably not the way to go. Pick an already well established and flexible engine with lots of documentation and knowledgeable folk around that you can pick the brains of ( I'd recommend UDK or Unity ). My guess is that unless you're building a AAA realstic FPS game in a few months, CryEngine3 is likely too new and specialised to be of much use to you.

Ditch one of your modellers and get another programmer :P

That's my 2c. Good luck, dude!