Escape from Zeta Aurigae - Demo version 0.3.0

Oh boy, is this a big one. This could be considered the first actual release of the demo, even though it's been available for a week or so now. 

First, the changelog!

Changes in 0.3.0

• New Feature: Construction Point blocking
  • Construction points may now be blocked by construction of certain structures
  • This is to indicate that the ship uses up capacity running a complex structure
  • Blocked construction points can be recovered by destroying the building using it
  • Future updates will have ways to gain more construction points.
• New Feature: Damage can bleed over to subsequent health pools
  • When damage depletes the topmost health pool, it can do damage to the one below.
  • This is calculated as a ratio; 50% overkill = 50% of damage potential on the next layer
• New Structure: Repair drones!
  • Repairs damaged structures across the starship
  • Each drone bay has 5 drones, and only 1 can work per damaged structure
  • When a drone is idle it will look for the most damaged building and head there
  • Drones automatically return to the drone bay to recharge
• New Turret: Pulser energy weapon.
  • Semi-wide tracking cone
  • High shield damage, but medium armor damage
  • Rather fast tracking rate
• New Turret: Melter energy weapon
  • Very long range, narrow cone
  • Slow tracking
  • Rather strong against armor and hull, weaker against shield
• New Turret: Impactor cannon
  • Slow firing rate, long range
  • Hard hitting, but not suited for shooting at fast enemies
  • Punches through armor and hull effectively.
  • Not great against shields
• Other changes:
  • All buildings are now using much less power
  • Hull now uses 1 power, to limit the overall size of a ship.
  • All weapons should now have sounds on shooting
  • A lot of little UI fixes and tweaks:
    • Construction panel now indicates the open menu
    • Construction panel buttons have indicators for hotkeys
    • Resource panel in bottom left now use new icons for energy, CP, and shields.
    • Certain score elements removed from game over screen. There will be updates to this.
    • Selecting ships now show all their weapons. There is no grouping yet
• Random statistics since last release (0.2.1+34, December 11, 2023):
  • Files changed 108
  • Lines inserted: 1202
  • Lines deleted: 234
  • Time spent:
    • Godot: 21h
    • Playtest: 5h
    • Git (Fork): 1h
    • Aseprite: 20m
    • Discord: 29h
  • Cups of tea brewed: Numerous
  • Mugs of chocolate: 4
  • Coffee? Yes, many.

Statistics

Most of the time since the last demo has been spent developing a way to get statistics on how well weapons do damage against predefined scenarios. To do this, I dummy enemies along a set of predefined paths, with a configuration of turrets set up, one damage type at a time. This would log all damage to a comma separated file with the time of damage and how much damage was taken.

Then I used, with the help of a friend (you know who you are!), a python script that uses pandas and matplotlib to process this data into 3 second slices and then doing 1000 samples from the data, finally producing a graph as the one below.

This graph will tell me the approximate damage output of each weapon against slow moving enemies at close distance (so they spend shorter time in the weapons targeting cone). This measures damage to armor only, letting me see how well the weapons perform against armored enemies. As you can see, most are quite balanced, while stalker mines have a rather higher output of damage.

These statistics are not a quick solution to finding balance between weapons, not by far, but it helps me get an idea how weapons perform in given scenarios. I have a series of those scenarios, and will be able to run this whenever I do changes to the damage output of turrets in the future.

This is the same scenario except that this is for hull damage. As you can see, stalker mines are bit further down compared to armor damage. Shredder and coiler excel as intended against hull, though the coiler has a rather wide set of outliers. This could be because of time spent changing targets as another target becomes more interesting. I will make a dev log at some point explaining the complexity of my targeting system.

I will for sure use this information to get a grasp for how well the turrets behave on a baseline situation, but it also requires a lot of play testing before a balance can be found.

If you have questions, reach out to me in a comment or look for me in the Godot discord where I am quite active on most days.