Duplicate Stellaris - The second point defense slot of the battleship carrier section breaks artillery computer range selction.

[Ranamar]

Description
The second point defense slot of the battleship carrier section breaks artillery computer range selction.

Game Version
2.8.2 (947b)

What version do you use?
Steam

What expansions do you have installed?
Synthetic Dawn, Utopia, Leviathans Story Pack, Apocalypse, Megacorp, Distant Stars, Ancient Relics, Lithoids, Federations

Do you have mods enabled?
No

Please explain your issue is in as much detail as possible.
As tested, I discovered that including point defense on an artillery battleship causes it to charge in to PD firing range from its targets before swinging around to go back out to its desired maximum range. It appears to be specifically point defense which causes this to happen, and I assume it happens with other weapons, as well.

Steps to reproduce the issue.
Create both of the attached ship designs.

Send fleets of each of the attached ship designs (separately, to avoid formation-flying confusion) to attack a stationary target, such as a starbase or enclave.

The fleet of ships with fighters but not point defense will move in to the desired range and stop, while the fleet of ships with point defense but not fighters will charge all the way in before turning around and flying back out to the same range as the other fleet.

The expected behavior is that neither ship class should move closer than Missile range, because that is the median weapon range.

Upload Attachment
File(s) attached

Edit: updated title with more accurate description

 

[Ranamar]

Okay, I did some more testing, and it appears that, specifically, it is the second PD slot of the Carrier Core hull which seems to screw this up.

Akagi-onepd, pictured here, will fly close...

... but Akagi-1pd-b, pictured here, will not.

 

[Ranamar]

After further testing with various designs, I have not found any designs which charge in when they shouldn't except for ones that are using the second PD slot on the battleship carrier core.

So, this is much more limited than I thought, but there are definitely things which are breaking artillery computer desired range calculations.

 

[MagicalChicken]

Technically not a bug as far as I can tell, artillery uses "median" as its formation distance, so it will charge in to the range of its median weapon, then back off to its "max" range weapon.

I do wish median's calculation would change however, it's a very weird way of doing it and causes stuff like this. Carrier computer currently avoids this issue as it uses "max" for all its distances, making it a better artillery computer in a lot of cases.

 

[SeekingEtermity]

OK, first of all that's a bug (if of nothing other than tooltip text) for multiple reasons:

1. Artillery computers claim they advance to the longest-range weapon, not that they will ever deliberately go closer under any circumstance. That's obviously incorrect.
2. Artillery computers specifically don't say anything whatsoever about median anything.
3. "The fifth slot, per a somewhat opaque and completely arbitrary numbering scheme, on a ship with 9 slots" is not any sane person's definition of "the median weapon" in any context, but especially not when the context is range (and the slots are not sorted by range).

With that out of the way, can I just ask: what the <REDACTED>? How did this get past QA? What about it was deemed a reasonable design? Why in the world would the "formation" distance ever be allowed to be less than the maintained distance? What was wrong with just making the artillery computer do the obvious thing, like the carrier computer? (Or is that the bug, which somehow you haven't fixed in multiple releases despite it being a 1-line change?)

 

[MagicalChicken]

OK, first of all that's a bug (if of nothing other than tooltip text) for multiple reasons:

1. Artillery computers claim they advance to the longest-range weapon, not that they will ever deliberately go closer under any circumstance. That's obviously incorrect.
2. Artillery computers specifically don't say anything whatsoever about median anything.
3. "The fifth slot, per a somewhat opaque and completely arbitrary numbering scheme, on a ship with 9 slots" is not any sane person's definition of "the median weapon" in any context, but especially not when the context is range (and the slots are not sorted by range).

With that out of the way, can I just ask: what the <REDACTED>? How did this get past QA? What about it was deemed a reasonable design? Why in the world would the "formation" distance ever be allowed to be less than the maintained distance? What was wrong with just making the artillery computer do the obvious thing, like the carrier computer? (Or is that the bug, which somehow you haven't fixed in multiple releases despite it being a 1-line change?)

The tooltips are just very vague (and can be somewhat misleading in the case of picket and line)

Since artillery's "formation distance" (where it starts wanting to go) is the median weapon's range, it'll go to that and as it gets there will then try to back off and attack from its preferred attack range (which is its weapon's max range) and attempt to stay at that range (not always consistently though, but ship AI as a whole has issues regardless), just wanted to add some more detail.

It's not a bug, it's just a very weird handling of things. As soon as artillery's formation distance is set to max instead of median, it acts pretty much like carrier does and constantly tries to stay away at all times. Unsure why they want it at median currently, maybe they had some reason for it, but I don't know what it would be.

Tooltips of combat computers in general could use some touching up though if nothing else.

 

[Peter Ebbesen]

@MagicalChicken

This weird behaviour and its reason has been known since the ship role revamp (and presumably hasn't been fixed because it is way down some priority list), and though you are right that it is mechanically the right way of behaving given how the game mechanics are implemented, that does not mean that it is not a bug

When battleships with hangar sections set to artillery behaviour engage by swarming the enemy, then turning tail and flying away until they regain the range to stand off with artillery, rather than standing off with its artillery while the fighters sweep in, it is a bug, as that is not intended functionality of battleships, carriers, or artillery computers.

 

[Rooker75]

And despite all this, they also saddled them with a firing arc and a minimum range. One of the devs really had a grudge against battleships.
Is it only the battleship that does this or do other ship types also act stupid when fitted with PD?

 

[MagicalChicken]

And despite all this, they also saddled them with a firing arc and a minimum range. One of the devs really had a grudge against battleships.
Is it only the battleship that does this or do other ship types also act stupid when fitted with PD?

This happens with any ship, it's based on how median selects a weapon. Most combat computers use median in some way, just not in the same way. It only hurts bigger artillery ships more because when they get in close just once, it's much harder for them to get away after that.

I have a feeling there was just a miscommunication on the dev team in some capacity, because the tooltips of picket and line for example are not indicative of what range they like or what they do.

 

[Ranamar]

And despite all this, they also saddled them with a firing arc and a minimum range. One of the devs really had a grudge against battleships.
Is it only the battleship that does this or do other ship types also act stupid when fitted with PD?

It's only the one slot. Every other PD slot works fine, at least on battleships. Give it a hangar bow and both of those work fine. IIRC, I even tested three up front and the one not-broken one on the central section, but I didn't screenshot it because I was getting tired of permutations. I didn't test cruisers because, again, I was getting tired of exhaustively poking things.

I felt like I was going insane when I emptied the second slot (to find a minimal test case) and suddenly the ship behaved the way it was supposed to.

 

[Ranamar]

The tooltips are just very vague (and can be somewhat misleading in the case of picket and line)

Since artillery's "formation distance" (where it starts wanting to go) is the median weapon's range, it'll go to that and as it gets there will then try to back off and attack from its preferred attack range (which is its weapon's max range) and attempt to stay at that range (not always consistently though, but ship AI as a whole has issues regardless), just wanted to add some more detail.

It's not a bug, it's just a very weird handling of things. As soon as artillery's formation distance is set to max instead of median, it acts pretty much like carrier does and constantly tries to stay away at all times. Unsure why they want it at median currently, maybe they had some reason for it, but I don't know what it would be.

Tooltips of combat computers in general could use some touching up though if nothing else.

I only know what it was supposed to be because there's another thread around here somewhere with the results of someone else doing code-diving.

Also, I'd say that the median->max thing usually works quite well for artillery. The ship flies up and switches to autokite mode as soon as it gets inside the max weapon range under most circumstances. Sometimes it does this thing when approaching at an angle where it can't stop fast enough and ends up wobbling back and forth across the max range line in a circle, which looks super weird, but usually it just flies up and stops.

 

[MagicalChicken]

This weird behaviour and its reason has been known since the ship role revamp (and presumably hasn't been fixed because it is way down some priority list), and though you are right that it is mechanically the right way of behaving given how the game mechanics are implemented, that does not mean that it is not a bug

When battleships with hangar sections set to artillery behaviour engage by swarming the enemy, then turning tail and flying away until they regain the range to stand off with artillery, rather than standing off with its artillery while the fighters sweep in, it is a bug, as that is not intended functionality of battleships, carriers, or artillery computers.

I still wouldn't personally call it a bug, as it was likely an intentional choice on their part and is technically working as it should, whether or not it was a good choice (I agree that it's not a good one and should be changed anyway) Although I will consider the chance that whoever did that change just possibly forgot how median or formation distance worked, but I'll give them the benefit of the doubt there for now, as it's something that's used with most combat computers so I would hope they know how it works (or at the very least know more than me, which should be easy)

Maybe I'm being too picky with the term bug, I don't know.

It's only the one slot. Every other PD slot works fine, at least on battleships

Well this at least confirms to me that it chooses the median weapon based on the middle slot, and not equipped middle weapon that I initially thought (which still wouldn't have been ideal, but would have been better at least) I will continuously wonder why they did it like that, it's such a weird way of doing things.

 

[SeekingEtermity]

It's not a bug, it's just a very weird handling of things.

I still wouldn't personally call it a bug, as it was likely an intentional choice on their part and is technically working as it should, whether or not it was a good choice (I agree that it's not a good one and should be changed anyway)

it's such a weird way of doing things.

If it
A) Doesn't make sense (as in, no sensible person would call the observed behavior correct)
B) Doesn't do what the user-/player-facing documentation (tooltip) says it does
then it's a bug. It's quite possible that the bug is in the game's definition of "median" rather than the behavior for the artillery computer (though honestly both are pretty questionable), but that doesn't matter, the resulting behavior is still wrong. "It does what the code says it should" is meaningless; computers always, at some level, do what the code says they should. The question is does the code say to do the right thing and here it just clearly does not.

 

[MagicalChicken]

If it
A) Doesn't make sense (as in, no sensible person would call the observed behavior correct)
B) Doesn't do what the user-/player-facing documentation (tooltip) says it does
then it's a bug. It's quite possible that the bug is in the game's definition of "median" rather than the behavior for the artillery computer (though honestly both are pretty questionable), but that doesn't matter, the resulting behavior is still wrong. "It does what the code says it should" is meaningless; computers always, at some level, do what the code says they should. The question is does the code say to do the right thing and here it just clearly does not.

I think B depends, because whoever wrote the tooltips for line and picket for example just straight up got them partly wrong at no fault of the combat computers themselves. Tooltips are their own (important) issues entirely that do need looking over, seems like there's often a disconnect between what a tooltip says and what was actually intended (I'm also looking at you, hull/armor regeneration numbers)

It's quite possible that the bug is in the game's definition of "median"

I am at the very least certain that median itself is actually working as they intended it, it's just a bad way of selecting the "middle" range as mentioned, made worse by the fact that it's not made clear anywhere to the user how it does so or even where it's used, which is an issue.

I just think our definition of bug varies in this case, as this to me is just a miscommunication of functionality and the behavior itself not being a very good fit for what it's on, which is all I've really been trying to say. I should have just led with that to be honest, I always say more words than I need to.

 

[Zares]

Are you sure about your game version?

That's an interesting finding... it would probably explain why my Cruisers with Line computers (median formation distance) wouldn't stop in time. And I my Cruisers had one of two sections from NSC mod - either P2M6 or S4P2H1 - funny thing, if it was second Point Defense slot doing this.

 

[TTFTCUTS]

Belated reply, but with this topic gaining traction recently as it was brought up again...

Picking the range of the "middle" filled slot would be the exact behaviour if you were trying to calculate the median range of the weapons by sorting them by range and picking the middle entry, but accidentally picked the middle entry of the unsorted list instead. My money's on something like that being the culprit.

 

[the_unnamed]

@MagicalChicken

This weird behaviour and its reason has been known since the ship role revamp (and presumably hasn't been fixed because it is way down some priority list), and though you are right that it is mechanically the right way of behaving given how the game mechanics are implemented, that does not mean that it is not a bug

When battleships with hangar sections set to artillery behaviour engage by swarming the enemy, then turning tail and flying away until they regain the range to stand off with artillery, rather than standing off with its artillery while the fighters sweep in, it is a bug, as that is not intended functionality of battleships, carriers, or artillery computers.

Now that you mention it, I’ve seen it happening too. Instead of starting to fire at longest range and then stay behind or keep on circling around the enemy, they first advance at them and then retreat to their actual position.

 

[GhostDanny]

Long range computers should try to stay at long range indeed.

Point defence are just for when missiles and strikecraft try to get close.
And should there for be considered a form of shield/armour, not a weapon.

 

[currylambchop]

We really need to keep this alive so it eventually gets fixed, maybe someone could report it again as I’m quite busy

 

[Feedback_Drone [Qloc]]

Hello!
Thank you for your vigilance and perseverance!
We are aware of the issue, it has already been added to our bug database.