Author Topic: design vs AI compared to design vs Players  (Read 17912 times)

0 Members and 2 Guests are viewing this topic.

Offline Michael Sandy (OP)

  • Commodore
  • **********
  • M
  • Posts: 771
  • Thanked: 83 times
design vs AI compared to design vs Players
« on: October 23, 2018, 01:05:08 PM »
So the prompt for this topic was a base I designed, primarily for colony morale, but intended to be useful as well.

    AMM Base class Base    1 000 tons     9 Crew     156 BP      TCS 20  TH 0  EM 0
    Armour 5-8     Sensors 1/0     Damage Control Rating 0     PPV 16
    Intended Deployment Time: 0.1 months    Spare Berths 1   
    Magazine 32   

    PDC AMM launcher PDC Size 1 Missile Launcher (16)    Missile Size 1    Rate of Fire 5  (96 BP)
    Bullet Catcher Missile Fire Control FC18-R1 (1)     Range 18.8m km    Resolution 1   (45 BP)

    This design is classed as a Planetary Defence Centre and can be pre-fabricated in 1 sections

An early magnetic plasma era design, with fire control technology salvaged from Precursors, so it is a bit more advanced.

Pretty cheap, I figured.  Not the cheapest way to get PPV, but at least somewhat useful.  Especially if you had a stockpile of semi-obsolete AMMs to use on the colony.  I gave it a single magazine so it could always have a ready load, and to slightly reduce the micro of reloading from colony.

But here is the thing, it is most effective in dealing with large missile volleys.  Which the AI tends to do a lot, even if they have extra fire controls they won't generally split fire.  But players can do smart things like having some .1 HS MFCs so they have the option of splitting up their volleys to deal with heavy point defense.  Or having nearly identical missiles with very slightly different speeds so their volleys are split as they fire.  Say, a 5.9997 MSP missile, a 5.9998 MSP missile, etc...

So designing versus a player, you will often want a significantly higher ratio of fire controls than versus the AI.

Versus the computer, if your fleet has enough point defense, you don't really need armor or shields on your missile ships.  You can count on the computer to fire missiles at a target until it is out of missiles, at which point they become safe to approach.  Players, however, would reserve missiles for point blank attack, or if their opponent split their forces and provided an opportunity to destroy them in detail.  Versus the computer, once you have determined the weapon mixes of the enemy ship types, you can often safely approach with a beam armed ship that is just slightly faster and slightly longer range and count on methodically tearing them apart.

But players can do things like suddenly have one ship withdraw, or one ship close, and figure out which ship their enemy is basing their range on, and thereby sneak into range.  They can also sandbag in terms of their displayed speed, or simply build a whole bunch of nearly identical ships that are all distinct classes, thereby denying exact information to their opponent.

Players can also both make use of forward observers, and develop tactics for blinding their enemy.  The scouting war can be far more involved for players than for the AI.  Long ranged EM homing missiles, just as an example.

It is fairly easy to design for flawless victories facing the AI.  Use decoy missiles versus their AI, railgun fighters versus conventional missile volleys, and AMM to weaken box launcher volleys, and short ranged missiles to kill enemy beam ships.  Easy peasy.  But players can conceal their max speed, hold missiles to the last second, choose NOT to use any AMM in order to reserve them for offense, and a variety of other tactics that make equal tech fights likely to involve casualties on both sides.

What other designs and tactics are particular effective vs the AI, but are either useless or unreliable versus players?  Or just unbelievably TEDIOUS?
 

Offline Michael Sandy (OP)

  • Commodore
  • **********
  • M
  • Posts: 771
  • Thanked: 83 times
Re: design vs AI compared to design vs Players
« Reply #1 on: October 24, 2018, 06:24:23 AM »
I would also point out that fleets designed for an Arena fight are also very different.  Because normal fleets evolve, components are built with earlier technology, to deal with earlier foes.  Arena fleets can be designed of a piece, without regards for shipyard costs or shipyard retooling.
 

Offline SevenOfCarina

  • Lieutenant
  • *******
  • Posts: 170
  • Thanked: 95 times
Re: design vs AI compared to design vs Players
« Reply #2 on: October 24, 2018, 06:24:53 AM »
Staged missiles, for one. They offer massive improvements in range over single-stage designs while still having a terminal stage fast enough to counteract point defence, and are thus pretty effective against NPRs, which tend to blob their ships in one region. But against players? They get wrecked - all it takes is a single escort ahead of the main group to shoot down the missile buses before they can launch their payloads.
 

Offline Jorgen_CAB

  • Admiral of the Fleet
  • ***********
  • J
  • Posts: 2837
  • Thanked: 673 times
Re: design vs AI compared to design vs Players
« Reply #3 on: October 24, 2018, 07:34:45 AM »
Staged missiles, for one. They offer massive improvements in range over single-stage designs while still having a terminal stage fast enough to counteract point defence, and are thus pretty effective against NPRs, which tend to blob their ships in one region. But against players? They get wrecked - all it takes is a single escort ahead of the main group to shoot down the missile buses before they can launch their payloads.

I find fighter crafts to be much more useful than multi-stage missile, both against AI and in multi-human campaigns. In multi-human campaigns there is a good chance a multi-stage missile is intercepted before it separates. Stand-off ranges from fighters can generally be much further and safer, plus you get the carrier of the missile back to do the same thing over and over.

I also agree with the arena versus practical application in real terms. There are never going to be a super best most efficient design because designs will evolve through the situation and confinements they are formed in.

One rule cheat is to create five exactly the same missiles and fire them from the same FC. This will make each missile its own salvo... very effective against point-defenses.

There are so many things you can do to game the game that simply make no sense and you simply should not do it. If you played against a real human opponent you just need to agree to which degree these things are fair game. To be honest you know full well when you are using a loophole in the game.

Build all hangar stations and have maintenance free sensors, point defenses etc... then build civilian freighters to tow them into combat using tractor beams. Each hangar can then house anything like a modular ships. Sensors and weapons actually can be used inside a hangar or deployed when needed. This is all a byproduct of how the game rules system works and if you were playing in a competitive environment with no house rules that is what you would see.

This game are simply not designed for competitive play, it is designed for role-play.
 

Offline Michael Sandy (OP)

  • Commodore
  • **********
  • M
  • Posts: 771
  • Thanked: 83 times
Re: design vs AI compared to design vs Players
« Reply #4 on: October 24, 2018, 11:50:43 AM »
I think that 2-stage missiles could be effective against a player, but that they would adapt to it far better than the AI ever could.

That also touches on another tactic human players can do that the AI can't.  Humans can bait out missiles and then run out of sensor range, but the computer isn't really good at that.  A player vs player missile duel, you also can't assume you will get to make use of your full missile range, because if you fire at maximum range, your opponent might turn away, reducing the effective range by 20% or so.
 

Offline Jorgen_CAB

  • Admiral of the Fleet
  • ***********
  • J
  • Posts: 2837
  • Thanked: 673 times
Re: design vs AI compared to design vs Players
« Reply #5 on: October 24, 2018, 12:03:30 PM »
I think that 2-stage missiles could be effective against a player, but that they would adapt to it far better than the AI ever could.

That also touches on another tactic human players can do that the AI can't.  Humans can bait out missiles and then run out of sensor range, but the computer isn't really good at that.  A player vs player missile duel, you also can't assume you will get to make use of your full missile range, because if you fire at maximum range, your opponent might turn away, reducing the effective range by 20% or so.

A smart player will use a calculator and fire active searching and/or passive missiles and never reveal your position in the first place. You use passive sensors to fin the enemy and then fire missiles. This will work much better in C# with small sensors getting a huge boost.

But this is all hypothetical since any tactic can be thwarted if you know or suspect it is going to be used, but sensors and scouting is going to be the most determining factor in player versus player conflicts not the weapons you use.
 

Offline Michael Sandy (OP)

  • Commodore
  • **********
  • M
  • Posts: 771
  • Thanked: 83 times
Re: design vs AI compared to design vs Players
« Reply #6 on: October 24, 2018, 12:15:08 PM »
I think players would still be vulnerable to a Pearl Harbor type attack involving 2-stage missiles.  It depends on whether the separation distance exceeds the 50 HS Res 1 detection range.  Something else player vs player designs have to take into account in a sudden transition from peace status to war status.

Players will discover some interesting problems with pvp passive missiles.  Namely, what happens if the target fleet scatters, resulting in the loss of the noisiest ship, but that being about it.
 

Offline Jorgen_CAB

  • Admiral of the Fleet
  • ***********
  • J
  • Posts: 2837
  • Thanked: 673 times
Re: design vs AI compared to design vs Players
« Reply #7 on: October 24, 2018, 04:02:53 PM »
I think players would still be vulnerable to a Pearl Harbor type attack involving 2-stage missiles.  It depends on whether the separation distance exceeds the 50 HS Res 1 detection range.  Something else player vs player designs have to take into account in a sudden transition from peace status to war status.

Players will discover some interesting problems with pvp passive missiles.  Namely, what happens if the target fleet scatters, resulting in the loss of the noisiest ship, but that being about it.

Given how easy it is to monitor Jump Points, especially in VB6 Aurora it is very unlikely you will manage a Pearl Harbor type strike... and I still fail to see what makes multi-stage missiles better than just using either a cloaked ship or fighter to fire the missile beyond enemy sensor range. I also question why active sensors would be on at all if you don't expect an attack, unless it is a station or a planet. Fleets in space should never use active sensors unless they know they are spotted, which defeats the purpose of a Pearl Harbor type attack.

A human versus human game would feature ferocious scouting element and the one who detect the other first will have a huge advantage. First strike is key, no matter how you do it.

In VB6 Aurora it is a bit too easy to find stuff in C# Aurora this will change and it will make fleets split up into different task-forces way more necessary and so combat will become much more complex.

Investing in a 50HS sensor in C# Aurora will be a real gamble, in VB6 Aurora not so much. In C# Aurora it will be way more difficult to launch a multi-stage missile strike with lots of picket defenses around a fleet given how powerful smaller sensors will become. You are going to need to sweep the screen of the fleet before you do anything, more or less. A multi-staged missile need to be pretty big if you want good fuel economy and range, especially in C# Aurora this will make it pretty easy to spot by rather small scouts. Fighter platforms will probably have allot better fuel economy than missiles now, even if they will also have reduced range. The hugely increased fuel consumption of missiles will make separation ranges much shorter and if deployed sensibly sensor scouts can still see pretty far.

If you send in a fighter strike you can abort of you detect a threat or you might send an escort to deal with any anti-fighter threat. Missiles don't have that luxury, once fired you just have to hope for the best.

I'm not saying that multi-staged missiles is bad in all situations, but I don't think it is a very good all purpose solution. It simply is too predictable, it also might reveal the position of your fleet more easily. Which is why I like fighters/FAC over multi-staged missiles. If I ever need to reveal or use on board missiles on my main ships something has gone wrong or I fire from a position of superior strength.

« Last Edit: October 24, 2018, 04:29:07 PM by Jorgen_CAB »
 

Offline Michael Sandy (OP)

  • Commodore
  • **********
  • M
  • Posts: 771
  • Thanked: 83 times
Re: design vs AI compared to design vs Players
« Reply #8 on: October 24, 2018, 07:18:14 PM »
I don't see it as multistage OR cloaked ships, as multistage missiles would make a good weapon choice for a cloaked ship.  But multistage is definitely available a lot EARLIER.  And as such, would influence early ship and fleet design.

A 2-stage missile that separates outside of the range of the largest possible Res 1 sensor available to the target is not very likely to be detected before separation.  Which suggests that 2-stage missiles would have a somewhat limited tech window, as the detection range increases beyond the range a boosted engined submunition is likely to have.

As far as yes, it is easy to monitor Jump points, but a lot of the after action report fiction deals with multination starts.  There is a lot of action within one system.  And while you might be able to hide your fleet, hiding your shipyards and maintenance facilities is a lot harder.  And they need to be protected full time, or else a conventional long ranged missile bombardment could wipe them out.

Yes, you could have extremely strong passive sensors capable of picking up even very small signatures in time to activate active sensors, but in VB6 there may be a rounding issue.  Sufficiently slow and small missiles might have a signature below 1, that would round to 0, and be undetectable on thermals.

Of course, multination starts are so unstable that it is almost a Fermi Paradox variant.
 

Offline SevenOfCarina

  • Lieutenant
  • *******
  • Posts: 170
  • Thanked: 95 times
Re: design vs AI compared to design vs Players
« Reply #9 on: October 24, 2018, 08:32:10 PM »
A 2-stage missile that separates outside of the range of the largest possible Res 1 sensor available to the target is not very likely to be detected before separation.  Which suggests that 2-stage missiles would have a somewhat limited tech window, as the detection range increases beyond the range a boosted engined submunition is likely to have.
It can still be shot down before stage separation. If the hostile task group can be seen, and you have more than one AMM warship, all you need to do is place them a few million kilometres apart and swap their positions every now and then.
 

Offline Jorgen_CAB

  • Admiral of the Fleet
  • ***********
  • J
  • Posts: 2837
  • Thanked: 673 times
Re: design vs AI compared to design vs Players
« Reply #10 on: October 25, 2018, 03:03:05 AM »
A 2-stage missile that separates outside of the range of the largest possible Res 1 sensor available to the target is not very likely to be detected before separation.  Which suggests that 2-stage missiles would have a somewhat limited tech window, as the detection range increases beyond the range a boosted engined submunition is likely to have.
It can still be shot down before stage separation. If the hostile task group can be seen, and you have more than one AMM warship, all you need to do is place them a few million kilometers apart and swap their positions every now and then.

Yes... it is far easier to over engineer AMM fire-controls and range on AMM is not that difficult either not to mention fighter interceptors are very good at shooting down slow multi-stage missiles as well.

A human opponent would have smaller scouts and recon vessels extending the R1 and passive cover around their fleet. No experienced human would just rely on the sensors of their core fleet. You will need to deal with the escort first. In C# Aurora this will even be more so... if you need to deal with the escort you might as well send fighters to fire the missiles anyway.

Multi-stage missiles tend to be really slow which give ample time to both detect and defeat them once detected. Time can be a huge factor if you need to actively paint the target until the missile hits as well.

As I said before... they can be useful in some situations. But if you rely on them for your main ships they will reveal your general location and that is not a good thing versus a human opponent and a human opponent will have many ways to extend their passive and active detection system around any significant fleet.

A fighter craft can move to a position as not to reveal the general position of your fleet and can slip past any passive sensors by moving to an intercept position ahead of time and then fire their missiles and just run away, if you use active/passive targeting that is.

In VB6 multi-staged missiles are more useful than they will be in C# Aurora, for the main reason fleets will need to break up more in C# and rely on more scouting crafts. Big sensors and fire-controls will have less range as well and favor small scouts and good positioning. Fighting will become more up close and personal with reduced missile ranges overall.

I could see multi-stage missiles as a means for extending the range of missiles on any platform as a viable option but they have problem being self guided the way the mechanic in the game works, so you will need to guide them with an active sensor which can be a potential problem unless you located all the enemy escorts and their core fleet. Not to mention extra cost of more powerful missile control systems, both research and build cost.
« Last Edit: October 25, 2018, 03:18:20 AM by Jorgen_CAB »
 

Offline SevenOfCarina

  • Lieutenant
  • *******
  • Posts: 170
  • Thanked: 95 times
Re: design vs AI compared to design vs Players
« Reply #11 on: October 25, 2018, 05:49:41 AM »
Honestly, I'm not certain carriers with fighters will in any way be superior to missile warships, at least in terms of combat efficiency. It's unlikely a fighter will have more than a third of its mass dedicated to launchers, and they frequently need to sacrifice range for the speed needed to outrun warships. And they get only one alpha-strike before they need to reload, which, given C# reloading rules will take a long time. Perhaps enough time for missile warships to locate and engage the large, slow carrier. I mean, all you really need to do is trace their vector as they run away form you. I don't really see carrier-fighter groups winning unless they have a 2 : 1 tonnage advantage or more. The flexibility they offer may make them worthwhile, though.
 

Offline Jorgen_CAB

  • Admiral of the Fleet
  • ***********
  • J
  • Posts: 2837
  • Thanked: 673 times
Re: design vs AI compared to design vs Players
« Reply #12 on: October 25, 2018, 10:57:04 AM »
Honestly, I'm not certain carriers with fighters will in any way be superior to missile warships, at least in terms of combat efficiency. It's unlikely a fighter will have more than a third of its mass dedicated to launchers, and they frequently need to sacrifice range for the speed needed to outrun warships. And they get only one alpha-strike before they need to reload, which, given C# reloading rules will take a long time. Perhaps enough time for missile warships to locate and engage the large, slow carrier. I mean, all you really need to do is trace their vector as they run away form you. I don't really see carrier-fighter groups winning unless they have a 2 : 1 tonnage advantage or more. The flexibility they offer may make them worthwhile, though.

The whole point with fighters is to launch a large enough volley to knock out the enemy high value targets so they need to retreat, that is a Strategic victory which is what is important in war. When you use fighters the whole point is using them as to mask where the carrier is located. This is pretty much how real world carrier operations are meant to function. Approach vector from fighters say nothing on where the carriers are in any way.

The fighters will usually be able to carry more missiles in one volley than comparable ship tonnage, this is what matters in fighter combat, getting that heavy strike to the target and do it in such a way your own strike force remain hidden. They can then dock and if the enemy does not retreat they can reload and strike again. This is of course the optimal way an engagement should go but rarely will.

Fighters are also allot cheaper to research and develop than comparable ships to, not to mention maintain and replace when new technology comes around. Hangars is also very cheap and something that remain cheap throughout the entire game, this make carriers relatively easy to maintain and cheap to upgrade.

I also dislike the combat efficiency term, efficiency means nothing if you can't use it effectively... small ships/fighters are more dynamic. The main benefit of larger ships is their capacity to withstand damage and defend themselves. Fighters and FAC are more dynamic as an attack platform, the large ships are mostly for support at least until you have advanced cloaking systems or very early in the game when you don't have enough fighter technology.

I also don't think any force will win at a 1:1 scenario.... the whole point is to make sure an engagement is as unfair as you possibly can. This means sensors and scouting is they key to success for the most part so you can have the right force at the right place at the right time.
« Last Edit: October 25, 2018, 11:14:04 AM by Jorgen_CAB »
 

Offline Garfunkel

  • Registered
  • Admiral of the Fleet
  • ***********
  • Posts: 2787
  • Thanked: 1051 times
Re: design vs AI compared to design vs Players
« Reply #13 on: October 25, 2018, 01:00:38 PM »
And that's where carriers/fighters shine. You can bring in a bunch of under 500 ton scout fighters with a carrier. Use passive scouts to get the enemy's location, then use waypoints to bring your strike force from an unexpected direction, use active scout to paint the target and launch the alpha strike. Return via same waypoint so the enemy survivors cannot just easily vector back to your carrier.

Only obstacle is fuel, especially now that refueling will take time, but it is very much possible, especially once you get the underway replenishment tech - just have a tanker fighter or two flying with the strike force and, at the way point, have them turn back. This will extend the range of your fighters.

Naturally the way to counter that is to have scouts of your own. This actually reminds me a lot of the Soviet vs American cat-and-mouse game around the Carrier Battle Groups during late Cold War, of Soviet long-range recon planes shadowing CBGs from a distance and American CAP fighters preparing to shoot them down if necessary.
 

Offline SevenOfCarina

  • Lieutenant
  • *******
  • Posts: 170
  • Thanked: 95 times
Re: design vs AI compared to design vs Players
« Reply #14 on: October 25, 2018, 01:14:55 PM »
The whole point with fighters is to launch a large enough volley to knock out the enemy high value targets so they need to retreat, that is a Strategic victory which is what is important in war. When you use fighters the whole point is using them as to mask where the carrier is located. This is pretty much how real world carrier operations are meant to function. Approach vector from fighters say nothing on where the carriers are in any way.

The fighters will usually be able to carry more missiles in one volley than comparable ship tonnage, this is what matters in fighter combat, getting that heavy strike to the target and do it in such a way your own strike force remain hidden. They can then dock and if the enemy does not retreat they can reload and strike again. This is of course the optimal way an engagement should go but rarely will.

Fighters are also allot cheaper to research and develop than comparable ships to, not to mention maintain and replace when new technology comes around. Hangars is also very cheap and something that remain cheap throughout the entire game, this make carriers relatively easy to maintain and cheap to upgrade.

I also dislike the combat efficiency term, efficiency means nothing if you can't use it effectively... small ships/fighters are more dynamic. The main benefit of larger ships is their capacity to withstand damage and defend themselves. Fighters and FAC are more dynamic as an attack platform, the large ships are mostly for support at least until you have advanced cloaking systems or very early in the game when you don't have enough fighter technology.

I disagree.

Let's assume a 10,000 ton missile warship can dedicate 4,000 tons to launchers, magazines, and fire controls. A 1:2:1 ratio between them seems reasonable, so that's 1,000 tons in launchers, 2,000 tons in magazines, and 1,000 tons to one 750 ton sensor and one 250 ton fire controls. Let's also assume that a 10,000 ton carrier, being slower, can dedicate 5,000 tons to hangars, magazines, crew berths, and spare fuel. If the overhead is 20%, then the carrier can support 4,000 tons of fighter. Fighters normally need to dedicate 40% of their tonnage to engines, and miscellaneous stuff is likely to consume another 20%, so they'll only be able to dedicate 40% of their mass to launchers and fire controls. Assuming a 3:1 ratio, that's a total of 1,200 tons of box launcher and 400 tons of fire control, across perhaps eight fighters. So the fighter-carrier can deliver an alpha-strike twice as large as the missile warship. Except the latter can sustain that, since it has perhaps four times the magazine capacity. In order to match magazine capacity, the fighter complement needs to halved, which brings their alpha-strikes more in line with each other. So a carrier effectively becomes slower missile warship with greater range but terrible reload rate.

But we haven't considered the sensor issue. Assuming resolution-optimised fire controls, the 250 ton MFC on the missile warship [root(10)*5=15.81x] will actually be able to engage the fighters at greater range than the fighters' 50 ton fire controls [root(200)*1=14.14x] can engage the warship. And notice how the warship comes equipped with its own sensor, while the fighters must either rely on an external sensor or have a quarter of the carrier's alpha-strike chucked to incorporate a sensor-fighter. Then there's the issue of ECM and ECCM, which larger ships have the tonnage to equip, unlike fighters.

There's also the fact that the faster missile ships will have greater strategic mobility, and can literally blitz through hostile territory faster than carriers can react. Any allied system, especially on a border with a hostile state, is likely to be a host to DST networks that will instantly see anything and everything trying to enter the system, not to mention that most players stick a sensor buoy on every jump point they run across. If you're invading hostile space, I see little reason to deal directly with a carrier fleet when it can be dealt with through siege tactics. The point of a war is to seize enemy worlds, not burn their fleets. The ideal way to deal with a carrier threat is to run for the hostile colony while fending off missile attacks with AMM spam, and then either drop ground forces or bombard the population into submission, gaining access to their DSTs and cutting off the carriers from resupply. Of course, this works for everything.

Perhaps I'm biased, but most of my doctrine revolves around being able to rapidly deploy wherever I need to, and this demands speed. Faster ships are better at concentrating forces and responding to threats, and faster ships demand bigger and more expensive engines, which get expensive to replace. Older ships are quite often far too slow in comparison to contemporary warships to be very useful, so it isn't worth it to keep them around.

Of course, there's no right or wrong way to play Aurora ....

 

« Last Edit: October 25, 2018, 01:22:36 PM by SevenOfCarina »