Wednesday, December 11, 2013

Post Mortem and Moving Forward

Post Mortem

Each member of the group was tasked with posting their thoughts about the good and bad aspects of this term. There is of course additional feedback (view-able below), but the summarized version of these results are as follows:

The Good

Dylan Yates - The entire team has agreed that Dylan was definitely one of the biggest factors leading to the success of this project. He did a disproportionally large amount of work, and that was due to our extreme confidence in his ability to use Unity. Without him, the project just wouldn't be what it is now, and we owe him big time.

Steady progress, regardless of setbacks - We had a new build each week that was noticeably improved over the previous week's. Regardless of assets being late or not turned in at all, we were able to push forward. This gave us a huge confidence boost as we moved forward. This was definitely the result of good team work and time management on an individual level, as well as group-wide.

Good overall team flow - This is broken up into a few categories.

  • Casual communication - Our group seemed to meld together very well. It didn't seem like anyone was afraid to put their voice into the project, and we were able to communicate well because of this. There were some issues, as any group will face, but we were able to overcome them and deliver a great final product.
  • Group diversity - We had a large enough group to give each person a specialized role. This allowed everyone to contribute their best skills to the project instead of having to take time to learn something they weren't necessarily as comfortable with. In a fast paced project like this, we didn't always have the luxury of spending a week or two to learn something new.
  • Individual Cooperation - Our team did a pretty good job working together and getting stuff done for most of us having not worked with one another before. There were of course issues at times, but even then the issues didn't completely stop the flow of workand could have been much worse. For the most part, we met all of our deadlines and provided high quality work, and we are proud that our team was able to do that.

The Bad

Level Design and Audio - We didn't have a designated level designer or sound person. The level design was mostly a team error, with no one wanting to take the lead on it, but we were not assigned a sound person and had to find someone outside of the class. Due to the lack of a dedicated game designer, the final game level is a little dry. Without a dedicated audio designer, we also missed out on a lot of atmospheric tension.

Management and team consistency - This is more of an overall disappointment in Evan and myself rather than the rest of the team. There were times earlier on where we think we didn't micromanage as much as we should have, and it may have ended up hurting the quality of the project at times. This is could be due to our inexperience with managing, but personally there were times when I didn't want to push other teammates to the point of them getting frustrated with me. I thought that this would break the flow of the team and be detrimental to our success. We now know that that is the curse of being in a management position, however, and we definitely stepped up our game towards the end of the term.


A few things that should have been stressed early on:

  • Naming conventions - Most of the people ignored the naming conventions that we agree on. It didn't show the downside now, because we don't have too many assets. It would become more and more of a problem as the project goes on.
  • Strict deadlines - There were issues with getting things turned in on time to the point where a formal policy had to be written up. This led to situations where code might not work with the newest assets. Path-finding was a large problem, with Ryan rarely having enough time to re-calibrate after new level designs were made. This left Dylan with barely enough time to pump out a build, and even less time for playtesting and creating presentations based on the results.
  • Communication - This was one of our biggest problems early on. At times, it wasn't clear what people were supposed to be working on and when it came to the time it was due, nothing was handed in. Eventually we figured out that all of our communications needed to be posted publicly so everyone knew what needed to get done.
  • Distributing work - It seemed at times that not everyone was putting in the same amount of time and effort into the project. This may have been due to an individual just not performing as well as they could, but there was also a lack of even distribution. This is not entirely due to the group communication, there were just times when something came up on the fly and it was just decided by someone to do it themselves rather than send it to someone else in the group.
Git - We had our fair share of technical difficulties with Git and Unity. Doing pretty much anything with Git would involve the repository getting messed up and we were forced to re-clone it. There was also the issue of Git not knowing how to handle merging Unity binary files Each time someone wanted to work on the project they had to notify everyone else to make sure no one was working on it, then they had to manually download the latest version, make their edits, and re-post. Git worked well for scripts, but since Dylan and Ryan's work rarely overlapped, we didn't get to leverage that. Git also worked very poorly for models, given the nature of their file structure.

Overall, the group agrees that we worked together well. And even though we had our issues, it was definitely as great learning experience for all of us.


Moving Forward
Our group met up for a discussion on what we would like to see moving forward.

Administrative

  • Spend a week going through the project to clean everything out and fix up any inconsistencies.
  • Master asset list.
  • Team communications need to be improved.
  • Stricter deadlines.
  • Plan for when people miss deadline.
  • Stronger structure.
  • Teamwork PM rather than Facebook.
Art
  • Art "manager" to ensure consistent art direction
  • Naming conventions need to be followed.
Programming
  • A list of what we need to program to distribute work better.
  • Comment code.
General
  • Have the Creature setting "seeds" or something to that effect, making it evident to the player that they are trying to cultivate the station for themselves to live in.
  • Puzzles.
  • More interesting environments.
  • Multiple monster types.
  • Better "cat and mouse" mechanics
  • Audio! Consistent and higher quality audio. We need a full audio director/engineer.
  • Dedicated time to meet up and work together.

No comments:

Post a Comment