Sunday, 10 October 2010

Ruins of the Mayans: Quick-Point Postmortem

Well I have finally gotta to the Beta stage of my map (Hurrah!).  I'm having a few teething problems with publishing the map (mainly materials missing GRRR!) so I have decided to look into this a bit further down the line, but for now I'm happy to stick some screenshots of the Beta version to my map on my portfolio site and upload a video later in the week.

As I have learned a hell of a lot on this project I thought I should take some time to reflect a moment on what went wrong and what went right in a quick-point postmortem.

  • The original design was a bit one directional - players traveled up to the pyramid and once up there there wasn't much to do on the back entrance apart from travel down again.  This meant I had to redesign the map so it took me longer than expected to get past the blockout phase.
  •  Setting up team start points as opposed to player starts - I failed to realise that in there were two different types of player starts.  Word to the wise for all noobs, when you are following a tutorial for creating map, make sure the tutorial relates to the type of map you are creating so that you add the correct player starts or in my case the team start points.
  • Creating kill volume in the water restricts the player's path and makes them stay on course - This was a great feature to add as I could manipulate the players paths instead of them using the water to swim away.
  • Setting the player starts to different teams - Again another niggling little issue that stumped me for a while.  Thankfully I found some helpful tutorials on the net to aid me with this.  Essentially set each team start to a different number either 0 or 1 for a two teamed game.
  • When blocking out, place everything before getting the players to playtest - They won't get the full experience if only half of the game features have been added.  Seems stupid but I did it so I think it's worth noting.
  • Doing this project whilst working through an Alpha and Beta period on a AAA title is quite demanding, physically and mentally - Thankfully I worked on the majority of this project during my lunch hours because after I got home I was extremely tired.  Unfortunately this meant that I was restricting myself to only five hours work per week for a while.  I did put a lot more hours into the development of this map during the latter stages though.
  • Publishing map - I have recently found that some materials have been missed from the cooked map.  Whether this was due to Unreal crashing near the end of this or that the materials I used just weren't accessible I don't know, however I will be looking into this further. 
  • Fully define my goals and expectations for the game before I begin.
  • I need to have a much fuller sense of the gameflow and pick ups needed at what point in the game.  To make sure that I do this I will be making more detailed maps before I even touch the editor.
  • Adjust the spawn rate of the pickups - As this will prevent players camping at locations where the better items are spawned.  By decreasing or increasing the spawn rate of the pick ups will also allow me to change the difficulty of the game.
  • Research architecture and environments more thoroughly so that when using restricted assets like only UT3 ones, I have alternative ideas of what to implement within the environment to make it more interesting instead of flat walls and empty space.  And if models are needed know exactly how things will look and where I can get these from.
  • Playtesting sessions and getting feedback are a godsend! - I would have never have gotten to the level I am at now without the feedback I had from the WoLD forums and my playtest sessions.  Mainly the entire flow of the game was fixed and changes that were necessary were made to make sure the game was fun.
  • Add cover spots so AI can also use cover and add 'priority' path nodes so that the AI move to those locations when the flags or significant objects are obtained by the player.
  • Re-work the scale that I use to produce a better sized map.  The scale that I used was way too big.  This was able to be seen clearly especially within corridors and exterior areas. 
  • Beyond Unreal Forums are ace!  A great forum that I used when I got stuck with small kismet issues.
  • Make more sniper locations for FPS - If I ever create another multiplayer map I will have to create more sniper locations.
  • The amount of playthroughs I did was good and gave me very informative feedback that helped to make my level much better (4 or 5)
  • Optimisation - I need to make sure that I have used the minimum amount of static meshes, lights and brushes to ensure when I create larger levels that the frame rate or loading times of the levels are not affected.  If I get into this habit now then hopefully I should be ok when it comes to my more ambitious projects.

In conclusion, I've learned that research needs to be thorough and across all aspects of your level.  But playtesting with experienced players of your genre will give you invaluable feedback and help to improve your game highly.

1 comment:

  1. Great write up. Its a great idea to do one of these at end of each level I should really do one on mine when I'm done.

    I ran into a lot of similar problems. I can really relate to the researching architecture one I've been stuck so many times because of the limited assets I had.

    Also your right about feedback and play testing.

    I'll have a look over your portfolio today and talk and see if I can give you some feedback :)

    Keep up the good work