7 Sprints later, here we are! The final post you'll be hearing from me about our now finished game Citadel! This post will serve as a post mortem from a production standpoint on the project, and of course as a chance for me to show off some amazing work from throughout the project that was completed by my team. I'm very excited to go over everything, so let's not delay!

To start, as always and for the last time, I'm happy to show you our final Burndown chart!

The final burndown chart shows our initial struggle at the start of production, which was a result of myself not assigning enough work at the start. Once I got a handle on this, and once the team started adjusting to the workload needed to complete the project on time, you can see the team very quickly catch up during sprint 3, and consistently stayed ahead until the very end of production, when they started to even out with my projections and completed the project right at the end (with the exception of some cut features). The final statistics are as follows:

Points Assigned: 77

Points Moved To Verify: 69

Points remaining In Verify: 0

Points Moved To Complete: 69

Points Assigned But Not Moved: 8

I'd also like to highlight some interesting statistics from the entire project's lifespan overall that I went ahead and calculated just out of curiosity! The most points completed in 1 Sprint came out to 107, during that 3rd sprint that saw the team not only catch up but get ahead of the project, and on average we saw 73 points per sprint, which is about 7 points per person. This is extremely close to my goal of ensuring that at least 8 points per person were assigned per sprint, with most, if not all 8 of those points hopefully completed.

With that, I'd now like to move into the actual post mortem section of this post. I'll be adding a few images here or there of work that I'd love to show off from throughout the whole of production. There's a lot I'd like to mention, so I'm going to do my best to keep things concise, and I'll be listing everything off to keep it nice and clean. Let's start off with what went right during the project!

• The overall tone of the game was extremely consistent, and I feel it's half of what makes the game really enjoyable to play.
• Spells are very fun to use, are effective in combat, and they look and sound great!
• All the primary systems in the game aren’t too confusing and tend to be picked up and understood well enough by the majority of players
• The game wasn't too easy or too hard. It was in a decent middle ground, and we rarely had anyone complain about difficulty one way or another.
• The Room Generation System resulted in what truly felt like unique castle runs, as it was able to utilize over 100 different room templates that could be put together to create unique pathways in the castle.
  
• By the end, the player was able to successfully earn gold in the castle, return to the surface with it, buy more stuff to get more powerful, then go back in feeling more powerful. it was a successfully implemented full gameplay loop!
• The player tips in the mines proved effective.
• Along with providing steady income, the mines proved VERY engaging, and loads of people would spend lots of time in the mines just clicking away.
• Plenty of the castle rooms was filled with personality, due to our creative 3D Modelers and Level Designers.
• Sound effects and music created an amazing atmosphere in-game, increasing player retention to the game itself

But nothing is perfect of course, and there are plenty of things that didn't go as well as they could have. Here's what I felt could have been done better on the project, and as a result what I would have done different, especially from a production standpoint:

• Work was getting assigned out of order, due to certain workflows not being planned out properly. This wasted a lot of time.
• Workflow for level designers especially was never quite consistent, it wasn't until well after the halfway point of the project that they finally had a consistent pipeline that they followed, which was far too late.
• Early Sprints were slow work-wise, and it took time for both myself, and as a result the whole team, to adjust to the amount of work competition that was needed to ensure the completion of the game.
• Annotated maps for levels were not used early on, due to me not assigning them as work, and the levels suffered for I. Annotated maps are very important for planning and for implementation, and this was a huge oversight on my behalf.
• Bugs were rampant in our game, and it would sometimes take entire sprints for them to be removed, just for them to come back later due to poor communication on Github management.
• Communication between the disciplines on the team could have been slightly better, and while the team had zero issues working together, we could have had the different disciplines working together on implementation far more than we did.
  
• We really needed more consistent merging between branches after bug fixes and implementations, to prevent repeated work or re-emerging issues. I mentioned this slightly when I brought up bugs, but this was an issue that effective multiple aspects of development.
• Player feedback in-game, while improved in some aspects towards the end of development, was still far weaker than what would have been best.
• The shop was consistently broken all throughout development until the very end, and it resulted in issues such as the in-castle version of the shop not making it into the final game, and of course not being able to properly playtest certain aspects of the shop, leaving it far weaker than it could have been.
  
• UI information would constantly be inaccurate and would constantly need to be fixed, and it wasn't until the end of development that it was truly finished, once again making early playtesting difficult.
• Some levels still felt too cramped, even after polishing them, and this sort of ties back into playtesting issues, and annotated map issues.
• the final boss Squigmarr, more specifically the fight with him, would break quite a lo, once again making playtesting difficult.
• For longer than an entire sprint, the level generation was broken and would lead to a dead-end, once again making playtesting difficult

and now to conclude this postmortem, I'll now go over what I would do again in this project, as it had a positive impact on the project and with how it turned out.

• My Trello Set-up was very efficient, clear, and made it so that it was easy for my developers to know what they were working on, and it made putting together my burndown charts very easy
• My consistent communication with team members, and how I encouraged collaboration, within disciplines at least, during implementation
• Our branch setups in GitHub prevented specific issues that arose from completely destroying the game, and I far prefer Github over other methods of version control.
• While implemented late into the project, the use of required team meetings for working specifically on builds I now consider a must and would require again in the future, as it was very healthy for each build that had dedicated meetings.
• I was not afraid to force the game to scale back or to cut features, and I would be just as willing to do it again, if not, even more, to ensure more important features would get implemented into the game.

And with that, that's Citadel! I thank you for reading this post, and if you have, for reading any of my other posts, and I hope you enjoy the game! I'll officially end this post with some final images from work on the game, and if you haven't already, be sure to check out The Adonai Invitational made by our sister team here at CAGD. Thank you again!