Turret balancing experiments

[BadgerBadger]

I'm planning on doing some experiments this weekend where I will put turret defenses down, observe the AIP then trigger a wave to attack me. I want to get some real numbers for "I've spend X on science and get crushed by a wave at AIP Y" to allow for more informed balance decision making.
I might wind up also just throwing things into the test chamber.

Do people have any suggestions for some experiments they'd like to see in particular? This might mostly a question for Keith, but I figure other people would have some ideas as well.

[etheric42]

I'd be curious to see various turret placement effectiveness tests.  For example, wormhole clustering, versus lines to the planet controller (AIWC style) versus distributed around all the power distribution nodes.

[Magnus]

I'd also want to check the effect of turret range on effectiveness. My hunch is that, right now, the power of a turret has more to do with its range than the actual dps number reported in the tooltip. Reason: being able to fire before getting return fire from its intended target.

[BadgerBadger]

Experiments. I spent 2K science on Turrets. I then built 1000 power worth of turrets defending 1 wormhole (the planet I had was only 1K power); two Mark II turret types and 1 Mark 1 turret type were unlocked. I had one turret type each that was best against Armor, Jammers and Evasion. This gave me a Strength of 8K. 4K worth of AI ships came through the wormhole and obliterated my turrets.  I tried with my turrets clustered around the wormhole the AI came through, with my turrets clustered around another wormhole, and also with a line of turrets leading from the AI spawn wormhole to the next wormhole they would go to. In both cases the battle was brutally one sided in favor of the fleetships.
 
Note that I changed the Time Before A Structure Can Rebuild to 60 seconds so turrets wouldn't just rebuild instantly; it looked really dumb. But overall turrets still feel very underpowered.

I also feel like tractor turrets are pretty useless. Sure they can hold ships in place, but turrets don't outrange the fleetships. So the AI ships I've trapped don't actually need to move at all to destroy everything.

AI ships tend to just hang out around my destroyed turrets waiting for them to respawn. This is an annoying bug

[keith.lamothe]

A quick change that would help (and deal with a mantis thing I saw go by earlier) is to replace

Code: [Select]

                    ArcenPoint exitPoint;
                    if ( entity.PlanetFaction.Faction.Type == FactionType.Player )
                        exitPoint = otherSideWormhole.WorldLocation;
                    else
                    {
                        //figure out where to have the unit actually "exit wide"
                        FInt allowableDeviationMultiplier = FInt.FromParts( 0, 080 );
                        FInt minValue = FInt.One - allowableDeviationMultiplier;
                        FInt maxValue = FInt.One + allowableDeviationMultiplier;
                        FInt actualValue = FInt.Create( Context.QualityRandom.Next( (int)minValue.RawValue, (int)maxValue.RawValue ), false );
                        AngleDegrees angleToExit = angleToWormhole * actualValue;
                        int distanceToExit = ( ( (FInt)distanceToWormhole * Context.QualityRandom.NextInclus( 98, 102 ) ) / 100 ).IntValue;
                        exitPoint = Engine_AIW2.Instance.CombatCenter.GetPointAtAngleAndDistance( angleToExit, distanceToExit );
                    }

with just "ArcenPoint exitPoint = otherSideWormhole.WorldLocation;"

That will kill the "AI ships popping out far away from the wormhole" thing.

It made a lot more sense when the wormholes were all along the edge of the grav-well, and the intent was for ships to visually "zip in" from the distance rather than just popping in.

[BadgerBadger]

So here's a new save game. I just had about 5-6K strength of turrets wiped out by 3K AI strength from a wave on Armus. Now, the 3K of AI strength was about 4000 Mark 1 ships. So yeah, they did quite a number on my turrets. Now, technically the AI did retreat (the battle went on long enough that the first set of turrets they killed managed to rebuild themselves, which feels bad. The AI could easily have actually won the fight by focusing down my Planetary Controller, and it just feels super weird mechnically/flavorfully that my turrets could rebuild themselves in time to rejoin a fight they had already died in.

I actually kinda feel like turrets fare really poorly against massive numbers of fleetships. If you build something like Lightning Turrets then they are outranged by every fleetship and never seem to do damage. I almost wonder if the AI should get more guardians and fewer fleetships in waves/as reinforcements?

As I think about it, I don't actually think there are any current counters in the human arsenal to massed fleetships besides warheads them. Are there actually any good AOE units that are always accessible?

[Magnus]

As I think about it, I don't actually think there are any current counters in the human arsenal to massed fleetships besides warheads them. Are there actually any good AOE units that are always accessible?

No. This also underlines another problem: starships are nearly useless offensively because they can only hit one target at a time, which becomes insignificant when you have thousands of weak ones to shoot down.

[keith.lamothe]

Siege Starships use Plasma Torpedoes, which are AOE. Same with Plasma Turrets.

Flak and Lightning turrets are AOE, though very short ranged.

That said, thanks for the save, it will probably help figure out why, in the math, the turrets are doing so much more poorly than their stats would indicate.

[zeusalmighty]

As I think about it, I don't actually think there are any current counters in the human arsenal to massed fleetships besides warheads them. Are there actually any good AOE units that are always accessible?

The human resistance fighters have access to Flak Starships; perhaps we should have access to these too? And/or the lightning starship? (If that's an actual thing)

[BadgerBadger]

The HRF use some AI Guardians that they've captured. They don't know how to make new ones

[zeusalmighty]

With the removal of the shield starship, the list is pretty sparse. The siege starship deals solid AOE but it deals structural damage. Would be great to have a starship that dealt some Jammer AOE damage (and armor for that matter)

[Magnus]

Siege Starships use Plasma Torpedoes, which are AOE. Same with Plasma Turrets.

Flak and Lightning turrets are AOE, though very short ranged.

That said, thanks for the save, it will probably help figure out why, in the math, the turrets are doing so much more poorly than their stats would indicate.

Siege Starships = structure damage X 10.
Plasma turrets = structure damage x 10.

What if I get hit by 10.000 jammers or armors?

Also, I need 500 tech investment for ONE siege starship. Which doesn't really change things when it's one against 10.000.

Flak and lightning turrets have shorter range than the things they're supposed to kill. Total waste of resources.

[zeusalmighty]

Flak and lightning turrets have shorter range than the things they're supposed to kill. Total waste of resources.

Radar dampening is definitely not going to be a thing? That would be one way of making these more viable if the simple solution of increasing their range isn't

[keith.lamothe]

Radar Dampening as a mechanic was constantly causing problems in AIWC's code and especially in unit behavior; I don't think there will be any range modifiers here.

On Flaks/Lightings, if you put them in a cluster around a wormhole whatever comes out is definitely in range. And has to stay there if you have enough tractors.

Fair point in non-wormhole scenarios, though.

[BadgerBadger]

The AI will send thousands of fleetships, which readily seems to overwhelm your available tractors.