Stellaris Dev Diary #79: Ship Component & Balance Changes

  • We have updated our Community Code of Conduct. Please read through the new rules for the forum that are an integral part of Paradox Interactive’s User Agreement.
I still have brutal difficulty with economy management. i dont know how the AI manages to have massive fleets with plenty plenty of research, minerals, credits, etc coming in. I've seen many an empire have far greater fleets with seemingly a much smaller amount of territory. If someone could point me in the direction of tutorials meant for retards like me I'd appreciate it.
 
Short answer: no

Longer answer: Your larger ship would have to deal with the Square Cube Law with regard to multiple systems, including the superstructure and the thruster exhaust. As the volume and expected mass of the ship increases by the cube of the multiplier, the surface area of the cross sections of the superstructure (if increased proportionally) only increases by the square. This means that a larger ship, assuming it has the same proportional superstructure, cannot handle as much acceleration without failing (either in the direction of main thrust or in turning) as a smaller ship. If the thrusters scale in the same way, they also do not provide as much proportional exhaust pressure as a smaller thruster. If the larger ship wants to commit a significantly larger percentage of its internal volume to superstructure, it can handle as large or larger compression and lateral stresses; it could then also apply a much larger percentage of its rearward-facing surface area to thruster exhaust in order to actually catch up to and pass a smaller ship.

Normally the only time the superstructure issues wouldn't apply is if the materials used are so strong as to take up such a negligible percentage of the ship's overall cross-section that massive increases in their absolute amount have virtually no impact on the usable volume. Unless you have reactionless drives, in which case neither would apply, but that's not how they're defined thus far within the game.

Unless of course someone wants to claim that since we have space dragons and psionics, we don't have to worry at all about anything approaching science, engineering, physics, math, logic, etc. Ever.

Interesting. I don't knew about this "square cube law".
 
While a fully fitted out corvette with all the best technologies might be two to three times more effective, it also costs up to ten times as much due to the fact that component mineral cost scales far more than their actual effectiveness, particularly for weapons.

In order to address this, we've gone over the cost of essentially every component and ship hull in the game.

They should cost the same mineral-wise. The problem is that Stellaris assumes that industrial capacity is available in basically endless quantities - not an unreasonable assumption, given the setting - and raw materials, represented by minerals, is the bottleneck on production, which is inconsistent with scaling the mineral cost of newer components.

Instead, make components cost energy as well, representing the amount of processing required to produce it, and scale that instead. This not only makes more sense, but would make energy more important towards the end game, making things like Dyson spheres more important.

This could then be further developed, for example by using energy to hurry production.
 
The armour and evasion formulas are so damn lazy it makes the whole combat system unbalanceable. But apparently we need to know exactly who domestic slaves are servicing at all times. There are only 3 weapon mechanics that differ by more than numerics and that is already apparently too hard for PDX. I would suggest that we need some energy storage and ship abilities to actually make it interesting but I don't think PDX would be able to balance that in time for the game to start in 2200. Despite having a system that tracks the location of every ship and every non-instant projectile as well as fighters, the actual complexity of the combat is below most freemium web games. It achieves a complete tedium and lack of interest in every aspect of naval warfare from technology to ship design to fleet building to logistics to fleet composition to battle deployment to battle control. On top of this sterile system is a toothless AI that spends all game sniffing glue so that RPers can jizz themselves about their great empire and buy more DLC.

The issue is a lack of depth - meaningful choices. Choosing between four different hull sizes is meaningless when one hull size is strictly superior. Choosing between five different weapons types is meaningless when three of them are worthless. Stellaris presents the player with lots of things to click on, but little in the way of meaningful choice.
 
Instead, make components cost energy as well, representing the amount of processing required to produce it, and scale that instead. This not only makes more sense, but would make energy more important towards the end game, making things like Dyson spheres more important.
Components already cost energy, namely the energy it takes to run a mine.

Their change to cut down on tier improvements difference is a good step in the right direction of having tech/research matter as well, while keeping the energy cost the same means you still can't overbalance weaponry & shield/armor on a ship.
 
I still don't understand why there is only one line of combat computers per ship type. Why not let the player decide whether they want to design their destroyers as picket, artillery, or frontline vessels?
 
I heard they used to exist but they were removed to enforce that each ship has a particular job.
They used to have an offensive & defensive computer with different stats, but they were all the same options for all ships. Thus, if you wanted fire rate on your artillery ship, you needed an offensive computer and it would charge in to soak all damage.

I'm trying to write down all possible tactics for ship 'designs' (artillery, broadside, carrier, torpedo boat,...) and maybe build a mod for those specific ones per ship size.
 
Wouldn't it be plausible to make the targeting algorithm more 'optimal', the higher it's equipped AI is?

The problem is that the "optimal" solution was actually bad, because "shoot all your guns at the one that produces the highest expected damage" means ships keep changing target and not removing enemy guns from the battle.
 
The problem is that the "optimal" solution was actually bad, because "shoot all your guns at the one that produces the highest expected damage" means ships keep changing target and not removing enemy guns from the battle.
...I actually think I would love to see idea standing behind that type of algorithm.
 
...I actually think I would love to see idea standing behind that type of algorithm.

It's in the defines.

Basically, the ship works out the expected damage (accounting for evasion, armour, and weapon bonuses) for every target it can reach for every gun it has, and shoots all the guns it has at the one target that produced the best outcome for any gun. (Weighted by other factors like the size difference between the ship and target, range to target, and the last target it shot at)

There's quite a large weighting in the system for bonus shield damage (especially because of armour), so weapons with bonus damage to shield tend to be the one that produces the highest target score and so direct all the fire. Which means as soon as the shields on one target run out they find a new one that has shields. Regenerating shields make this even worse because the old target gets loads of shields back because it isn't being shot any more.

In 1.8 the weighting for the last target the ship shot at will be increased so ships should stick to their target more even with weapons with shield damage bonus.

XL weapons are exempt and pick their own target, and also pick the target the ship points at (though not necessarily what all the other guns shoot at).


This was also why Kinetic Battery/Artillery was the best anti-shield weapon to use, because it fires much more slowly than plasma and doesn't take over the target preference until its turn to fire, a ship with that gets to fire two volleys of plasma at a target, and because of focus fire that usually means it gets to do proper hull damage.