Trying to understand range in Stellaris

  • 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.

Cordane

GW/SC/PD/Flak Wonk
18 Badges
Sep 25, 2013
673
402
  • Cities: Skylines Deluxe Edition
  • Tyranny - Tales from the Tiers
  • Stellaris - Path to Destruction bundle
  • Crusader Kings III: Royal Edition
  • Crusader Kings III
  • Stellaris: Distant Stars
  • Stellaris: Apocalypse
  • Hearts of Iron IV: Expansion Pass
  • Hearts of Iron IV: Death or Dishonor
  • Europa Universalis IV
  • Stellaris: Leviathans Story Pack
  • Stellaris: Digital Anniversary Edition
  • Tyranny: Archon Edition
  • Hearts of Iron IV: Cadet
  • Stellaris
  • Pillars of Eternity
  • Cities: Skylines
  • Stellaris: Synthetic Dawn
I recently had a thread where I discussed a range-based Tracking/Evasion system, but I'll admit I made some pretty large assumptions about the current operating environment with regard to ranges. The system also assumed that typical combat ranges could comfortably be measured in whole or appreciable fractions of light-seconds (approximately 300,000 km for one light-second), and the system offered Evasion based on the length of time between when a weapon is fired and when it would impact, along with the thruster capability of the ship. While Tracking in this system does have some consideration for how quickly a given turret can traverse and elevate, direct-fire (DF) weapons operating at those ranges are going to use extremely small movements and just have to be ultra precise.

As the range gets tighter, the length of time available to evade will shrink toward nothing for beam weapons and, because the designers won't likely completely gimp the effectiveness despite what would realistically be relatively slow flight speeds, DF projectiles (plasma balls, kinetic slugs) will still likely be close to beam weapons. However, angular changes will start to creep in a bit, but wouldn't be significant until ranges are so close as to make beam weapons impossible to evade and even projectiles will offer little opportunity to do so. In order to take advantage of a turret that can traverse 90 degrees in less than a minute (or deal with those that can't do it), ranges would have to be within hundreds (not hundreds of thousands) of kilometers.

I'm not asking (right here and now) for the graphic engine to properly render space battles at the few light-second scale, instead of the current light-HOUR scale (light from our Sun takes about 43 minutes to get to Jupiter, and I've easily seen Stellaris battles cover beyond that relative distance), but knowing what range "ballpark" we're supposed to be in makes it easier to brainstorm corrections and variations to it.
 
I'm not asking (right here and now) for the graphic engine to properly render space battles at the few light-second scale, instead of the current light-HOUR scale (light from our Sun takes about 43 minutes to get to Jupiter, and I've easily seen Stellaris battles cover beyond that relative distance), but knowing what range "ballpark" we're supposed to be in makes it easier to brainstorm corrections and variations to it.
There is this concept of "Intentionally Abstract". For example, population is intentionally abstract.
If Paradox ever gave a figure how many human equivalents one pop is worth, people would bombard them with post on how "migration X" or "pop related system Y" is unrealistic.
And they are just smart enough to avoid that bleep :D
 
True... And if too much reality were to be injected into this then the real ranges would make solar systems so incredibly vast that space battles would be... real... but very different.
In essence, there ends up being only two space battle scenarios. Meeting engagements where both sides decide to fight, and forced engagements (static and pursuit engagements where one side (usually inferior) cannot escape the engagement.
At that point, range means very little except to when various participants are no longer part of the engagement... which is roughly what we see now.
 
Range is not nearly as good as it sounds, because every ship captain in Stellaris has a raging stiffie for close quarters fighting.

Tl;dr on OP, but this problem really should be fixed.
 
There is this concept of "Intentionally Abstract". For example, population is intentionally abstract.
If Paradox ever gave a figure how many human equivalents one pop is worth, people would bombard them with post on how "migration X" or "pop related system Y" is unrealistic.
And they are just smart enough to avoid that bleep :D
I'm not asking for specificity, just a rough ballpark figure - ships shooting at each other with a million kilometers in between fight entirely different from those a hundred kilometers apart. The display of the battles that we have currently is absolutely not what it is supposed to represent - the distances and time frames are both WAY off - but what it's supposed to represent isn't well known either.
 
Reviewing the design of the data files... that may be because people tend to mix too many ship types into fleets... And the aggregate becomes a confused admiral AI.

I am going to do a test with a could of very long range combat computer fleets fighting along side some very close range combst computer fleets to see how they do in combat and whether they keep their roles sorted out.
 
I'm not asking for specificity, just a rough ballpark figure - ships shooting at each other with a million kilometers in between fight entirely different from those a hundred kilometers apart. The display of the battles that we have currently is absolutely not what it is supposed to represent - the distances and time frames are both WAY off - but what it's supposed to represent isn't well known either.

Would the graphic showing a deep zoom in make it feel better?
 
Would the graphic showing a deep zoom in make it feel better?
No, because I know what the current representation on screen is, and it's obviously not trying to represent any actual battle. Do you honestly think that they're imagining most fights starting at 3/4 of a BILLION kilometers away and having each weapon fire only once every 1-5 MONTHS? Those aspects are completely for ease of gameplay, but they're likely extrapolated out from a more realistic basis that can be modified and then blown back out to playable scale.
 
If they wanted space battles to resemble anything close to reality, they'd have to break down the skirmishes into their own little "encounters" a la Endless Space 2. The scale of space is too friggin' huge otherwise.

It's probably best to think of the galaxy map as an abstraction of the galaxy rather than take it at face value. Think of it as a one of those little holographic 3D maps you see on every scifi movie or something.
 
If they can shoot from across the system then that's like shooting from pluto to the other side of the solar system.
Pluto is 49 AU away (really far) from the sun. That's 6.7 light hours from the sun. Even from one end to another it's still less than a day so it's not actually FTL breaking to shoot across the entire system in less than a single day.
 
If they can shoot from across the system then that's like shooting from pluto to the other side of the solar system.
Pluto is 49 AU away (really far) from the sun. That's 6.7 light hours from the sun. Even from one end to another it's still less than a day so it's not actually FTL breaking to shoot across the entire system in less than a single day.
All I can say is, holy shit. With speeds like that, every hit should be an instakill. Including planetary bombardment.
 
All I can say is, holy shit. With speeds like that, every hit should be an instakill. Including planetary bombardment.
That would shorten the game... oh and probably create galactic peace. I played a game of risk with my buddy and our two girl friends once. He and I beat each other up... and the girls moved in and killed us and then declared a global matriarchy and world peace...
A less than satisfactory conclusion.
 
All I can say is, holy shit. With speeds like that, every hit should be an instakill. Including planetary bombardment.
Lasers literally have speeds like that. ;)
 
Range is not nearly as good as it sounds, because every ship captain in Stellaris has a raging stiffie for close quarters fighting.

This happens because when ships were told "park at your longest range" they park at the longest range of any of their weapons and most of them will be out of range.

So the various computers park at either ranges 40, 60, or 80 from their chosen enemy.

Except corvettes that keep zooming around.

Range is not really an accessible balance factor between ships in the current model.
 
This happens because when ships were told "park at your longest range" they park at the longest range of any of their weapons and most of them will be out of range.

So the various computers park at either ranges 40, 60, or 80 from their chosen enemy.

Except corvettes that keep zooming around.

Range is not really an accessible balance factor between ships in the current model.

Those values should be percentages of the max range the ship's guns have. That should help the ship understand how far it is supposed to be.
 
Those values should be percentages of the max range the ship's guns have. That should help the ship understand how far it is supposed to be.

Right, but 80% of the range of a Giga Cannon is 120. Still out of range for every other weapon in the game except Neutron Launchers.

Which isn't a problem for degenerate exploit builds but is for the AI.

80 was chosen because it's just inside the range of the shortest range Large weapons (once you add the 20% that the computer grants), so a ship can't be a dumbass and park out of range of most of its fire.

There isn't really an effective way of doing anything else in the current model.
 
Right, but 80% of the range of a Giga Cannon is 120. Still out of range for every other weapon in the game except Neutron Launchers.

Which isn't a problem for degenerate exploit builds but is for the AI.

80 was chosen because it's just inside the range of the shortest range Large weapons (once you add the 20% that the computer grants), so a ship can't be a dumbass and park out of range of most of its fire.

There isn't really an effective way of doing anything else in the current model.
Correction, pick the range that is most prevalent by dps among the ship's loadout by the intended behavior:
On an artillery Battleship (1 GIga+ 4 Neutron launchers) that would be neutron launchers.
On a L-gun Cruiser 2L guns and 2 Ms, that would be 2L.
On an arty destroyer (1L + 2 P), that would be the 1L

Corvettes would follow a different logic, as they are supposed to perform a different role. They have picket and swarm available. Both are short range, so i would differentiate them by emphasis on attack/defense. Swarm would be aggresive and picket as the one more focused on self-preservation in combat.

Amongst behavior types that leaves line combat, which i would consider thus:
Advance till you can fire all your weapons, except PD, be between the enemy and your Arty ships. This would basically be the best one for cruisers and destroyers against missile/ strikecraft fleets.

Thoughts?
 
Correction, pick the range that is most prevalent by dps among the ship's loadout by the intended behavior:
On an artillery Battleship (1 GIga+ 4 Neutron launchers) that would be neutron launchers.
On a L-gun Cruiser 2L guns and 2 Ms, that would be 2L.
On an arty destroyer (1L + 2 P), that would be the 1L

All of those make the L/M mix ships to park out of medium weapon range. (noticably everything forced to have M slots also can't have artillery behaviour so always parks closer).


Thoughts?

In the current model my approach would be to go back to having 1.9 style computers and just set Destroyers to park at 25, Cruisers at 50, and Battleships at 80 with most of the M slot sections disabled. That would stop the AI's autodesigns being as dumb as they are.

(Also the idea of kinetics for shield and energy for armour needs to go because the AI is still using 1.0 preferences for one or the other, so it easy to counter and always ignores whatever fancy synergies you try and design in)

In the long term I think a more abstract combat model would be better.
 
Not that I'm suggesting any of this but I do have few solutions in mind, which are pretty radical.

1. get rid of space battle entirely, only existing as statistics like other games (EU4, CK2 etc)
2. Add another view. We have galactic map view, to system view. Then there would be another one where you can zoom into the actual battle where ships are so shrunk in size. Time also flows differently in this view, by the hour to be exact. Disabled in multiplayer of course.
3. shrink the ship sizes, range of weapons and the battle space itself. But they are going to have to be tiny.
 
Alright, there is a completely separate set of conversations going on in here that I had no intent on starting - let me explain what I'm looking for:

I know how to play Stellaris - have for over 900 hours. I understand how the purely game-related elements interact with other game-related elements. I know that the galaxy and system views, and the planets and star ships moving about in them, are simplistic representations of vast, complex systems. I. Got. It. Thanks!

What I'm looking at is how turrets would work in a real world equivalent to what's imagined here. Think about it this way: one ship can "see" an opposing ship in the distance. It wants to fire at the enemy but its turret needs to follow the target as it moves. How fast does the turret have to move?

Math involved: I've seen Stellaris fleets cross solar systems (early game) in two months time, and that gets faster as the game progresses. I thought to myself that the major planets (Pluto's disputed status notwithstanding) would mark the hyper-limit for that, so I just ran with the average distance from our Sun to Neptune (30.11 AU), doubled it, and divided by 60 days. That approximate 1 AU/day converted over to a little over 1,700 km/s (bookin, I know), and then I thought about what that speed would look like moving along the edge of a circle with a radius of 1,000,000 km (approximately 3.3 light-seconds) - our first range to target.
  1. At that speed and at that distance, the enemy ship has an angular velocity (the same speed the turret has to match) of 0.10 degrees/second. That translates out to the turret needing to turn 90 degrees in 15 minutes - not hardly moving at all. Virtually any size turret is going to keep up with that. The range is where it's difficult - even full lightspeed weapons are going to be hard pressed to hit an agile target. Regardless of how fast you think an actual rail gun projectile is going to be, it certainly will not be lightspeed or probably anywhere close, so those are going to usually miss even a somewhat mobile target. Never mind any lightspeed lag from your sensors.
  2. Taking that same speed and making the radius/range 100,000 km (1/3 of a light-second) still has the angular velocity at 1 degree/second, or 1.5 minutes to turn 90 degrees - not too fast, but we're also talking about decent sized turrets. But our 1/3rd light-second lag each way makes each shot easier, with little time for lightspeed weapons to be dodged, and rail guns getting a bit closer to respectability.
  3. Moving in another notch cuts our range down to 10,000 km (1/30th of a light-second), and cuts the time for a turret to turn 90 degrees down to 9 seconds. That's pretty close to the rotation speed of a WWII tank's main gun, which (relative to a starship) is probably pretty dinky - for the larger turrets, hahaha. Again, lightspeed weapons hit in a blink and projectiles travelling at even just 1% C only take about 3 seconds shot-to-impact.
  4. One last move in to just 1,000 km (1/300th of a light-second) and bringing the turret around 90 degrees needs to take less than a second. Accuracy/Tracking is going to take a major penalty as swinging the turret around that fast is going to conflict with moving it precisely to hit something 1,000 km out. Might be able to dodge a K-slug if it's "slow" but higher techs nullify any chance - that's assuming it was going to hit anyway. (/math)
  • When the range is long, Tracking is limited only by sensor lag, as turrets do not have to swing around quickly. Evasion is relatively easy, because even lightspeed weapons take time to get to their targets.
  • When the range is shorter, turret turn speed becomes the larger factor in Tracking, as there's virtually no sensor lag. Evasion can still be a factor, but it's not going to work for any ship that can't accelerate hard off-vector or isn't quite small. Understanding the relative differences in ships' profiles helps inform Evasion, as the potential area/volume of where a targeted ship can be relative to its previous vector and its profile, in the amount of time the weapon gives the target, is the biggest factor in making a shot miss.
Using the above information, which of the four options do you think is most likely to be the one that the Devs are using to determine the rules that we're supposed to be following in our simulated environment? I could see Strike Craft getting into that 1K range, making them damn tough to be hit by anything - any closer and they're more likely to impact themselves on the target versus hitting it with their weapons at those speeds. Corvettes and Destroyers might have to stay moving in the 10K range to have a chance, while anything larger than a Battleship is still going to be kind of screwed if they try to stay around 100K from their main targets. Looking at it now, I don't really see the 1MM range being viable for anything but versus-station fighting.

Please note that I still have the velocities at early game levels and cruising speed to boot - turret capabilities will only have to get better as techs progress. Also I'm just winging it on the hard numbers as (like I said) the Devs haven't given us any to actually work with.