Massively Immersive

Aaron E. Freedman's personal blog. Game design. Coffee. Games.

teamcordata:


Update_0.6
Hello tumblr. It’s a been over a week since we all returned from GDC. We’re full of vitamin D, over-priced eats and laden with business cards. The largest event from this past week was running our second fully networked play test. This formed with core of our production discussion this week. For the sake of clarity, let’s list some of the key features in the current build:
Teams of 3 players.
3 fighter classes - drone, tank & mech.
2 types of weapons for each class.
Players score points for each kill.
Points can be multiplied by picking up kill confirms.
Points can be deposited at capture points.
Players loose all points on death.
Players can change fighter class on re-spawn.
There are a few remaining features not yet implemented in our working build but let’s move on the results of our testing. These are some of the issues we discovered last week:
Discrepancy between local and remote versions of player positions. Likely this is a problem that will never totally go away but it can be improved greatly. One option is to establish a predictive algorithm that creates a more accurate tracking of the player position but can produce some network lag and “jumping” whenever the server re-syncs player locations. Another option is to lerp remote and local versions of player position by a different ratio so that this discrepancy becomes less noticeable. You will undoubtedly hear more solutions as we tackle this issue.
Maps are too big. Ah the great wide open. Plenty of space means players are zoning defensively and there is too much down-time between firefights. Smaller levels with more complex geometry for our next test.
Player types need iteration. Everyone loves playing the mech-type (pictured). A good medium class with average speed, it supports classic strafing and feels balanced. The drone feels a bit like a recklessly fast unicycle and always shoots in the direction it is traveling. The tank is like a very slow car. We received suggestions to make the two light and heavy classes more extreme in order to isolate the benefits of playing these classes.
Visual communication.  There are a host of visual communication questions. The most glaring of these is the symmetry of tank and drone models confusing the direction players think they are facing. We got requests for drawing the mouse position (controls turret aim in mech and tank models) and requests to display your life stat (currently only ammo, points and aim reticle are shown). There is a ton of UI work to do that will take precedence as we near our soft-launch date so you’re sure to be hear about this further.
Various Bugs. Weapons not reloading. Falling off the world. The usual.
Phew, that’s a lot. Because this has taken up so much space I’m going to leave what we worked on after the test till our next post but we are more or less all on the same tasks we had last week. Check back in soon for an update on our third play test, level design and player class iterations.
Oh yeah, almost forgot, our game has a name!
Welcome to Zero.Mark

teamcordata:


Update_0.6

Hello tumblr. It’s a been over a week since we all returned from GDC. We’re full of vitamin D, over-priced eats and laden with business cards.
The largest event from this past week was running our second fully networked play test. This formed with core of our production discussion this week. For the sake of clarity, let’s list some of the key features in the current build:

  • Teams of 3 players.
  • 3 fighter classes - drone, tank & mech.
  • 2 types of weapons for each class.
  • Players score points for each kill.
  • Points can be multiplied by picking up kill confirms.
  • Points can be deposited at capture points.
  • Players loose all points on death.
  • Players can change fighter class on re-spawn.


There are a few remaining features not yet implemented in our working build but let’s move on the results of our testing. These are some of the issues we discovered last week:

  • Discrepancy between local and remote versions of player positions. Likely this is a problem that will never totally go away but it can be improved greatly. One option is to establish a predictive algorithm that creates a more accurate tracking of the player position but can produce some network lag and “jumping” whenever the server re-syncs player locations. Another option is to lerp remote and local versions of player position by a different ratio so that this discrepancy becomes less noticeable. You will undoubtedly hear more solutions as we tackle this issue.
  • Maps are too big. Ah the great wide open. Plenty of space means players are zoning defensively and there is too much down-time between firefights. Smaller levels with more complex geometry for our next test.
  • Player types need iteration. Everyone loves playing the mech-type (pictured). A good medium class with average speed, it supports classic strafing and feels balanced. The drone feels a bit like a recklessly fast unicycle and always shoots in the direction it is traveling. The tank is like a very slow car. We received suggestions to make the two light and heavy classes more extreme in order to isolate the benefits of playing these classes.
  • Visual communication.  There are a host of visual communication questions. The most glaring of these is the symmetry of tank and drone models confusing the direction players think they are facing. We got requests for drawing the mouse position (controls turret aim in mech and tank models) and requests to display your life stat (currently only ammo, points and aim reticle are shown). There is a ton of UI work to do that will take precedence as we near our soft-launch date so you’re sure to be hear about this further.
  • Various Bugs. Weapons not reloading. Falling off the world. The usual.


Phew, that’s a lot. Because this has taken up so much space I’m going to leave what we worked on after the test till our next post but we are more or less all on the same tasks we had last week. Check back in soon for an update on our third play test, level design and player class iterations.

Oh yeah, almost forgot, our game has a name!

Welcome to Zero.Mark

Prototype Studio Preview

Play »

Here’s a preview of another project I’ve been working on over the past few months. It’s just a collection of hacky prototypes I’ve been making while fleshing out my Unity prototyping kit. This one is a short test of a scripted dialogue system.

I’ve got some other stuff, like a clicker game and a space game and other whatsits that I’ll post once I whip up a launcher and fix a few broken things.