Should fighters really act as cannonfodder?

For what's not in 'Top Priority Game Design'. Post your ideas, visions, suggestions for the game, rules, modifications, etc.

Moderator: Oberlus

What should ship weapons usually target?

Planets, ships and fighters equally
1
10%
First target ships and planets. If no planet or ships: target fighters
3
30%
Target only ships and planets. (Also remove fallbacks for other weapon types)
5
50%
Something different (please explain)
1
10%
 
Total votes: 10

Message
Author
User avatar
Krikkitone
Creative Contributor
Posts: 1559
Joined: Sat Sep 13, 2003 6:52 pm

Re: Should fighters really act as cannonfodder?

#16 Post by Krikkitone »

I think you should have weapons distinguished on what they Can target rather than what they Tend to target.

Standard Weapons Only target ships and planets, even if all that is left is fighters they can't attack them
PD Weapons Only target Fighters/bomber/interceptors and missiles/torpedoes if they go in (they cannot attack ships/planets)

Interceptors and flak would be PD weapons
Bombers would be standard
Fighters could be PD+attack ships (but not planets)

JonCST
Space Kraken
Posts: 157
Joined: Sat Dec 15, 2018 4:28 am

Re: Should fighters really act as cannonfodder?

#17 Post by JonCST »

Hi all. Morlic said:
Morlic wrote: Sun Sep 15, 2019 6:43 pm Fighters as cannonfodder would be fine with me if they were actually removed as combat targets after being shot down. Since they are removed at the end of turn, they can block an arbitrary number of shots.
And Moric corrected that in a later post to "bout/round". The result i just got in a multiplayer game:

Two carriers (mine) got into a fight with 12 laser ships (theirs).

First round:
8 fighters launched.
Both carriers destroyed, no damage to laser ships, no shots fired at fighters.

Second round:
8 fighters * 8 damage = 64 points of damage to laser ships.
4 laser ships destroyed.
All eight fighters destroyed.

Third round:
No combat at all.

So, if i understood what this is saying, the fighters do not block an arbitrary number of shots.

What i believe they do is exist as a target for any ship to choose until they are destroyed.

Since ships fire simultaneously, fighters can be "overkilled", i.e. draw more fire than required to destroy them in that round.

But, they disappear at the end of the round in which they are destroyed.

Does that make sense?

I don't object to this behavior, any more than i object to Many SGHs all targeting a single opposing ship and "overkilling" it.

It's kind the nature of the combat system.

Jon

User avatar
Oberlus
Cosmic Dragon
Posts: 5714
Joined: Mon Apr 10, 2017 4:25 pm

Re: Should fighters really act as cannonfodder?

#18 Post by Oberlus »

JonCST wrote: Tue Sep 17, 2019 6:42 pmSo, if i understood what this is saying, the fighters do not block an arbitrary number of shots.

What i believe they do is exist as a target for any ship to choose until they are destroyed.
Until they are destroyed at the end of a combat round.
So, in your example, the 8 fighters were shot down by 26 (or 39) shots of laser ships, and each of them took an arbitrary number of hits.
One fighter can take several hits in the same round, that is the problem pointed out.
I myself don't have a strong position here.

JonCST
Space Kraken
Posts: 157
Joined: Sat Dec 15, 2018 4:28 am

Re: Should fighters really act as cannonfodder?

#19 Post by JonCST »

Oberlus wrote: Tue Sep 17, 2019 8:32 pm Until they are destroyed at the end of a combat round.
So, in your example, the 8 fighters were shot down by 26 (or 39) shots of laser ships, and each of them took an arbitrary number of hits.
One fighter can take several hits in the same round, that is the problem pointed out.
Right, but isn't that true of any ship? It can be hit by multiple shots far beyond it's destruction? Then disappears at the end of the round?

The fighters were targeted while they still existed. If a (normal) ship is destroyed, the shots don't get re-targeted, correct? Why should fighters be different?

Am i missing something?

J

User avatar
Oberlus
Cosmic Dragon
Posts: 5714
Joined: Mon Apr 10, 2017 4:25 pm

Re: Should fighters really act as cannonfodder?

#20 Post by Oberlus »

JonCST wrote: Tue Sep 17, 2019 11:43 pmRight, but isn't that true of any ship? It can be hit by multiple shots far beyond it's destruction? Then disappears at the end of the round?

The fighters were targeted while they still existed. If a (normal) ship is destroyed, the shots don't get re-targeted, correct? Why should fighters be different?
Maybe they should, because they are able to soak much more damage in relation to their initial number of hit points (1).

JonCST
Space Kraken
Posts: 157
Joined: Sat Dec 15, 2018 4:28 am

Re: Should fighters really act as cannonfodder?

#21 Post by JonCST »

Oberlus wrote: Wed Sep 18, 2019 1:02 am Maybe they should, because they are able to soak much more damage in relation to their initial number of hit points (1).
If that's the concern, maybe the weapon targeting should change, rather than when ships are destroyed?

Maybe only flak target fighters, "real weapons" aim for "real targets"?

I seem to remember a couple of discussions like this in other threads. Similar concerns about SAC, too. Something like "why should a SAC target a little ship when there's a planet or a big ship available?"

I've also heard something about classifying weapons as "point defense" or "offense".

Point defense: flak, interceptors, fighters
Offense: bombers, MD, Laser, PC, DR, SAC

So, a prioritization scheme for weapons, something like:

Point Defense: Fighters, ships
Offense: Ships, planets, fighters.

And some would request:

SAC: Planets, ships, fighters.

and change targeting logic to something like:

Case weaponType A:
if object in targetClass t1 exists, target it
else if object in targetClass t2 exists, target it
...

Case weaponType B:
...

Not sure how awful it would be to implement this, nor what the run-time burden would be.

J.

Ophiuchus
Programmer
Posts: 3433
Joined: Tue Sep 30, 2014 10:01 am
Location: Wall IV

Re: Should fighters really act as cannonfodder?

#22 Post by Ophiuchus »

I added the suggestion from Krikkitone as an option.

This reset the current poll which was

Planets, ships and fighters equally
1
14%

First target ships and planets. If no planet or ships: target fighters
3
43%

Something different (please explain)
3
43%

The voices for different I think were from Krikkitone (which i added as an option), alleryn (which I read as "give me more info about the options"), and ???

Please everybody, revote!
Last edited by Ophiuchus on Wed Sep 18, 2019 2:40 pm, edited 1 time in total.
Any code or patches in anything posted here is released under the CC and GPL licences in use for the FO project.

Look, ma... four combat bouts!

Ophiuchus
Programmer
Posts: 3433
Joined: Tue Sep 30, 2014 10:01 am
Location: Wall IV

Re: Should fighters really act as cannonfodder?

#23 Post by Ophiuchus »

JonCST wrote: Wed Sep 18, 2019 2:16 pm Point defense: flak, interceptors, fighters
Offense: bombers, MD, Laser, PC, DR, SAC

So, a prioritization scheme for weapons, something like:

Point Defense: Fighters, ships
Offense: Ships, planets, fighters.

...

Not sure how awful it would be to implement this, nor what the run-time burden would be.
The way you imagine implementation would be awful (because freeorion conditions do not work that way).
But it is actually easy to implement the intended result and run-time burden would be neglible.

Also please compare with Krikkitone's suggestion - it sounds a lot like what you write.
Any code or patches in anything posted here is released under the CC and GPL licences in use for the FO project.

Look, ma... four combat bouts!

User avatar
Oberlus
Cosmic Dragon
Posts: 5714
Joined: Mon Apr 10, 2017 4:25 pm

Re: Should fighters really act as cannonfodder?

#24 Post by Oberlus »

Ophiuchus wrote: Wed Sep 18, 2019 2:35 pmThe voices for different I think were [...] and ???
And me. I'm rather busy and didn't get the time to come back and post my thoughts. I haven't got the time to sit and reflect on this to really get to a position. Now is the time.

The alternatives I consider worth trying:

Krikkitone's:
Each weapon class has a subset of possible targets. All possible targets of a weapon class are targetted with uniform probability. Other targets can't be targeted.
- Cannons (including SAC) target ships and planets (pure Offense)
- Bombers target ships (and planets but not ignoring shields?) (pure Offense)
- Interceptors and flaks target drones (pure PD).
- Fighters target ships and drones (mixed role).
This is mostly standard FRO combat system (before preferential combat system) with some adjustments (make cannons do not target drones). It is indeed the most KISS alternative, and I can't find combat scenarios with a exploit. Having removed drones as targets from cannon weapons remove the seemingly OP damage-soak property of the drones.

Preferential approach (Ophiuchus'?):
Each weapon class has a subset of primary targets and a subset of secondary targets. Primary targets are targetted first, secondary targets only when there are no primary targets.
- Cannons: primary ships and planets, secondary drones.
- Bombers: primary ships, secondary drones.
- Fighters: primary ships and drones, no secondaries.
- Interceptors: primary drones, secondary ships.

Probabilistic approach (mine some time ago?):
- Each weapon class has a subset of primary targets and a subset of secondary targets. Primary targets are targetted with double (or triple) probability than secondary targets.

Exampli gratia:
10 cannon ships VS 10 carriers and 20 bombers
* Standard system:
- cannons target 33% carrier, 66% bomber.
- bombers target 100% ship (there is anything else to target, this is the same in every system).
* Krikkitone's:
- cannons target 100% carrier (and once there are no ships, bombers are still untargetted).
* Preferential:
- cannons target 100% carrier (once there are no ships, bombers are targetted 100%).
* Probabilistic:
- cannons target 10*2/(10*2+20*1) = 50% carrier, and 20*1/(10*2+20*1) = 50% bomber.

I could live with Standard system if drones where removed from combat round as soon as it gets one hit, so that following shots in the same round can't target it (i.e. no multi-hit soaking). That is meassing with combat implementation.
I could live with all other designs while drones are still allowed multi-hit soaking, but Krikkitone's is the simplest (KISS) and straightforward to implement (I think so), plus it has one good advantage: it gives actual importance to flak cannons, currently under-powered. The only advantage I could see of the Probabilistic method over Krikkitone's, and I'm not sure it is a real advantage, is that it allows for extra randomness in combat, but it comes with the actual drawback of a more complex implementation (and maybe more CPU usage).

So I go for Krikkitone's approach.

Ophiuchus
Programmer
Posts: 3433
Joined: Tue Sep 30, 2014 10:01 am
Location: Wall IV

Re: Should fighters really act as cannonfodder?

#25 Post by Ophiuchus »

Ophiuchus wrote: Wed Sep 18, 2019 2:35 pm I added the suggestion from Krikkitone as an option.

This reset the current poll which was

Planets, ships and fighters equally
1
14%

First target ships and planets. If no planet or ships: target fighters
3
43%

Something different (please explain)
3
43%

The voices for different I think were from Krikkitone (which i added as an option), alleryn (which I read as "give me more info about the options"), and ???

Please everybody, revote!
Not many revotes. There are 2 votes now for Krikkitone's suggestion. I guess these are from krikkitone, and oberlus. The one vote for the second option is still mine.

So I guess the other votes stand:
One vote for equality (pro smokescreen)
Three votes for preferred targets (contra smokescreen)
Two votes for specialised targets (also contra smokescreen)
One vote (alleryn) for something different/need more input. (maybe/maybe not smokescreen)

So most are against smokescreen, so i start working on that.
Any code or patches in anything posted here is released under the CC and GPL licences in use for the FO project.

Look, ma... four combat bouts!

User avatar
swaq
Space Dragon
Posts: 384
Joined: Tue Aug 20, 2019 1:56 pm

Re: Should fighters really act as cannonfodder?

#26 Post by swaq »

Is it possible for there to be no enemy ships but still fighters? (option 2)

User avatar
Oberlus
Cosmic Dragon
Posts: 5714
Joined: Mon Apr 10, 2017 4:25 pm

Re: Should fighters really act as cannonfodder?

#27 Post by Oberlus »

swaq wrote: Thu Sep 26, 2019 12:33 pmIs it possible for there to be no enemy ships but still fighters? (option 2)
Yes.
On round one the carriers launch their fighters. Those fighters won't be targetted that turn, but the carriers will. So it can happen that the carrier launches fighters and is destroyed in the same turn, and those figthers without carrier will be fighting until they are destroyed or the combat ends. Then they will try to be "rescued" by any carrier with space in their hangars.

Ophiuchus
Programmer
Posts: 3433
Joined: Tue Sep 30, 2014 10:01 am
Location: Wall IV

Re: Should fighters really act as cannonfodder?

#28 Post by Ophiuchus »

So first results. If boats are not valid targets they need a buff in order to compete.

Balancing wise I start with the assumption that Laser Fighters should lead to a similar cost efficiency as similar MD4 design against unshielded targets.

Balancing is all affected by the number of boats in a hangar, the launch profile, the damage per boat and the cost of hangars and launch bays.

I have also two starting points - UI wise it would be good if e.g. the damage value could give an impression of power level. E.g. if laser fighters are on a level as MD4 they should have a damage value of 5 or 6.

The other starting point is - costwise should carriers be rather for expensive or cheap hulls?

I will assume three bouts so a boat launched in bout-1 does 2*dam and 1*dam if launched in bout-2. I would prefer to up the number of bouts to four because it will help balancing the boat types but i think that is definitly out of scope for 0.4.9.

And the last axis is balancing slots and hangar vs launch bay costs.

I will have a look at a few designs. And think how to split and reason the two combats systems (E.g. ships and planets shoot ships and planets; boats shoot boats and ships).
Any code or patches in anything posted here is released under the CC and GPL licences in use for the FO project.

Look, ma... four combat bouts!

Ophiuchus
Programmer
Posts: 3433
Joined: Tue Sep 30, 2014 10:01 am
Location: Wall IV

Re: Should fighters really act as cannonfodder?

#29 Post by Ophiuchus »


I updated my spreadsheet based combat simulator to be able to handle normal ships in combination with fighters.

I ran some values for a small fleet of six robos against a opponent shooting 10 shots each turn. I used zortrium/laser4/laser fighters levels for the values.

For determining how many ships survive the exact amount of damage is not necessary (only one damage level). For calculation of combat effectiveness it needs the values to see how much damage the fighters and the ship weapons do combined.

One important thing i noticed: it is quite good for the robo fleet to have half up to two thirds fighter carriers. Much more survival and doing just a bit less damage than a robo fleet without carriers.

But the normal cost efficiency formula sais the non-carrier is four times as cost efficient as the carrier (0.026 vs 0.082).
Even if i tweak the cost efficiency formula to add 22 structure per launched fighter (i.e. one overkill on L4) to the fleet and assume the fighters do not get shot down the carrier the cost efficiency is way worse (0.063 vs 0.082).

Has some errors in my simulator, this needs an update
Any code or patches in anything posted here is released under the CC and GPL licences in use for the FO project.

Look, ma... four combat bouts!

Telos
Space Squid
Posts: 50
Joined: Thu May 09, 2013 4:46 am

Re: Should fighters really act as cannonfodder?

#30 Post by Telos »

From my perspective, there are two big problems in this ballpark, and I don't think eliminating drone-targeting would help either of them.

First, there's the problem that decoys (esp. comsats, but also cheap componentless flux ships), have way too big of an effect on combat outcomes for their cheap price. Decoys are already present in Round 1 (unlike drones that take time to launch), and decoys are much cheaper to mass, so their damage-absorption impact is much larger than that of drones. In my eyes, decoys raise all the same damage absorption worries that drones do, except to a much greater extent. It may be that the best solution to the decoy problem is to introduce some sort of preferential/restricted targeting scheme. If so, I would suggest that any change to the eligibility of drones for being targeted should be bundled together with a change in the eligibility of decoys for being targeted. E.g., I could imagine lumping unarmed troopless ships together with drones as being low priority for standard weapons, though I'd say they should still be eligible as secondary targets.

Second, there's the problem of stealth carriers, who can launch their drones while facing no danger themselves. Making drones be untargettable by standard weapons would make this problem even *worse*, since then fewer drones would get killed in round 2, so they'd do even more damage in round 3, and would survive to subsequent turns. So I think you REALLY SHOULDN'T make drones untargettable until you've fixed stealth carriers, e.g., by making launching drones de-stealth the carriers. (Other options were discussed in this thread.) I guess it wouldn't worsen this problem if you made drones be *secondary* targets, since they're typically the *only* visible targets in a stealth carrier attack, so making them "secondary" wouldn't actually change anything.

Aside from these two big problems, I actually don't have any problem with drones working as a sort of decoy. The carriers are still very vulnerable in Round 1, and the opportunity costs are high enough that drones seem balanced enough to me.

Post Reply