Macro-manage Late game large scale

A forum for chatting about in-development game features.

Moderator: luciderous

Ashbery76
Posts: 76
Joined: Tue Feb 03, 2015 9:04 pm

Re: Macro-manage Late game large scale

Postby Ashbery76 » Mon Aug 07, 2017 8:23 pm

This problem seems to happen in all 4x games.Just go big with a few PD ships..The battle systems don't have a reason for task forces as its just a wall of fire.There are no subs,ect that require specialist mixed fleets.

Land bases tactical game don't suffer from this.You play TW and you need variation in units..

zolobolo
Posts: 439
Joined: Fri Nov 25, 2016 3:49 pm

Re: Macro-manage Late game large scale

Postby zolobolo » Mon Aug 07, 2017 9:26 pm

Ashbery76 wrote:This problem seems to happen in all 4x games.Just go big with a few PD ships..The battle systems don't have a reason for task forces as its just a wall of fire.There are no subs,ect that require specialist mixed fleets.

Land bases tactical game don't suffer from this.You play TW and you need variation in units..

Agreed, but I have seen more balanced fleets with the two sizes here then in any other space 4X game (even if the AI mostly doesn't mix in a third category)

The solution is the same as in land battles yes.
Land battle games: archer>spear; spear>cavalry; cavalry>archer (optional: infantry good against spear and archer)
Space battle solution analogue to above: destroyer>capital; cruiser>destroyer; capital>cruiser (optional: small craft good against cruiser and capital)

The player is thus encouraged to generally build fleets of all 3 types represented = more variety = more unceartanty in engagements = more fun :)

User avatar
faijeya
Posts: 41
Joined: Thu Apr 16, 2015 6:57 pm

Re: Macro-manage Late game large scale

Postby faijeya » Mon Aug 07, 2017 9:59 pm

zolobolo wrote:I would love to see such a thing :)

Sparse stars plus marathon pace do the trick.

zolobolo wrote:<Real life comparisons skipped>

You're wrong in most of your examples.
However, it's a forum to discuss the game, not measuring grounds.

The only relevant thing for the game is that the vehicles tend to get heavier, larger and costlier until they overgrow their niche.
For example, modern MBTs are that expensive, that IFVs are slowly turning into light tanks.
The only thing that stops them from resurgence is the ungodly amount of T-54/55ABCDE/Type 59/Type 69 in the third-world countries which otherwise would be stomping grounds for these things.

zolobolo wrote:This game already emraces this concept of roles (particularly since railgun rebalance).

Not even close.
However, switching the design paradigm mid-game from ship-centric to fleet-centric is quite a bad idea.

Anyway, the combat (or any conflict whatsoever) in 4X is there primarily to validate the correctness of player's strategic decisions.
If player's tactical genius may (CREEE-E-E-E-E-D!) influence the outcome, well, it's fine, but not necessary.
And due to a relatively small percent of player's attention budget the combat may take, any scheme a single variable too complex will fail.

zolobolo
Posts: 439
Joined: Fri Nov 25, 2016 3:49 pm

Re: Macro-manage Late game large scale

Postby zolobolo » Tue Aug 08, 2017 8:54 am

faijeya wrote:
zolobolo wrote:I would love to see such a thing :)

Sparse stars plus marathon pace do the trick.

Ok, let me give this a whirl and see what falls out of it.. but what difficulty do you use? Normal, Hard, Harder?
Please also let me know how many star systems are set and how many AI opponents

faijeya wrote:
zolobolo wrote:This game already emraces this concept of roles (particularly since railgun rebalance).

Not even close.
However, switching the design paradigm mid-game from ship-centric to fleet-centric is quite a bad idea.

Examples:
- AI sends 8 Destroyers against my Battleship which is technically more advanced - Battleship looses
- AI sends 2 Carriers and 2 Heavy Cruisers against my Fleet Carrier and Battleship and the later loose
- AI sends Battleship and 50 Frigates against my inferior tech Battle cruiser around 3 Medium Cruisers and a few destroyers and AI fleet get wiped out

faijeya wrote:Anyway, the combat (or any conflict whatsoever) in 4X is there primarily to validate the correctness of player's strategic decisions.

Mostly agree, but if you include a tactical combat system in your game, it will better offer the possibility to turn the tide on close battles besides offering a show. What I am trying to say is that it needs to have an impact and this is difficult to achieve if there are only a couple of battleships duking it out as that situation will lead to a foregone conclusion

faijeya wrote:And due to a relatively small percent of player's attention budget the combat may take, any scheme a single variable too complex will fail.

Combat does take only a couple of minutes with a handful of ships (sorry I still have to see these big fleets), so this is covered.
If a player is into 4X I would argue that they have at least the will to invest a lot of attention into the game. The details are what make this genre compelling for most and not the visuals or exciting action.

The variables need to be transparent, and while I agree they should not be overly complex per-say but they need to be infused with enough of uncertainty to make the decisions interesting even after the player has established an effective strategy otherwise it becomes boring, and random events are not always a good solution. An obvious solution would be to not allow for a perfect tactical approach (rock-paper-scissors mechanic) and mix it up with different faction dispositions

zolobolo
Posts: 439
Joined: Fri Nov 25, 2016 3:49 pm

Re: Macro-manage Late game large scale

Postby zolobolo » Tue Aug 08, 2017 12:39 pm

Ran into great test-cases and probed around with them to see how much fleet composition and player decision matters:
Short result: very much

Case 1: Two Yoral Feets facing off with the same tech (missile, torpedo, shield and guns level) in a space battle (no planetary defense involved)
AI fleet: 1 Battleship 1 Heavy Destroyer (estimated SUM: 1200 production, 2100 metal)
Player fleet: 2 Light Cruisers 1 Torpedo Destroyer 1 Destroyer and 3 Frigates (SUM: 1260 production, 2170 metal)

Autocombat result: Player defeat. AI looses Heavy Cruiser and Battelship takes around 35% damage
Player tactic 1: Player defeat . Offensive maneuver concentrating around 70% fire fire on Heavy Cruiser first, enemy Battelship took around 50% damage
Player tactic 2: Player wins . Defensive maneuver utilizing missiles and torpedo salvos beyond range first before moving into range. Concentrate 100% fire on Heavy Cruiser first. Player lost 1 destroyer

Case 2: Two Yoral Feets facing off with the same tech (missile, torpedo, shield and guns level) in a space battle (no planetary defense involved)
AI fleet: 2 Battleships 2 Heavy Destroyers (estimated SUM: 2400 production, 4200 metal)
Player fleet: 1 Carrier, 3 Light Cruisers 2 Torpedo Destroyers 3 Escort Cruisers (SUM: 2450 production, 4150 metal

Autocombat result: Player defeat. AI looses both Heavy Cruiser but both battleships remain undamanged
Player tactic : Player wins . Defensive maneuver utilizing interceptors, missiles and torpedo salvos beyond range first before moving into range for bomber runs. Concentrate 100% fire on Heavy Cruiser first. Player lost 1 Light Destroyer

Conclusion:
1. Using autocombat, the default tactic favors less mixed fleet composition (though furhter testing should be done to confirm this)
2. Player tactic has a huge impact on the outcome when different fleet composition is used not only turning the tide of battle but also limiting losses to a minimal amount.
3. When player develops advanced understanding of the tectical combat mechanics, resources are more efficiently spent on fleets with mixed roles then on capital ships

In both player tactics, Firgates served as PD only, and destroyers as torpedo and missile platforms.

Thus it seems to me that the games tactical combat mechanics benefit mixed fleets.
Personally I find this more fun then concentrating production on captial ships alone and would greet a solution which resolves large fleets problematic withouth changing this mechanic

Uncle_Joe
Posts: 37
Joined: Wed Jul 19, 2017 10:55 pm

Re: Macro-manage Late game large scale

Postby Uncle_Joe » Tue Aug 08, 2017 4:14 pm

Agreed 100%. Player skill in combat can and does matter very much.

I would love to see it released as a 'one-off' multiplayer combat sim. ;)

I tend to keeps smaller ships up and running for about 60-75% of the game. Once I'm in the mop-up stage, I rarely bother with them anymore and just crank out the largest ships and keep them moving to the front. By that point, fleet composition rarely matters and it's just a matter of piling more firepower to the front.

But for that first 60-75% of the game, fleet composition matters very much and weapon mix also matters. In my last game as Colonial, I was fighting the Ashdar Imperials and pretty much whomever massed up more carriers tended to win a big way (felt very much like WW2 in the Pacific). But since the Fleet carrier were still really metal-intensive, I was still building up Battlecruisers with direct fire weapons for fire support and using Light Cruisers for cheap PD support. I fought one huge battle vs a similar AI fleet and all of his Battlecruisers were heavily set up for PD and it was a RUDE awakening lol. Most of my fighters died in the initial wave and if I hadn't had a nice core of the gun-armed BCs backed by the PD CLs, I would have been heavily rolled. As it stood, I won, but barely. The AI should have retreated it's carriers when they were low on fighters since they and their supporting PD BCs lacked the firepower to take out my BCs. If it had retreated, I would have suffered a severe material loss at that point and that was with equal-ish tonnage, but differing weapon mixes.

Long and short is that the battles are still generally interesting enough to be fighting them manually throughout most of the game. Sure, the walk-over battles I just auto-calc, but for another approaching parity, it's a lot of fun to go in and pit my designs vs theirs. I'd VERY much like to try that against other human players too. ;)

User avatar
faijeya
Posts: 41
Joined: Thu Apr 16, 2015 6:57 pm

Re: Macro-manage Late game large scale

Postby faijeya » Tue Aug 08, 2017 4:30 pm

AI does target a single ship if it can.
It means the whole torpedo/missile/interceptor salvo will be aimed onto a single ship, usually the closest to the AI forces.

Autoresolve plays it fair, but a human player may and probably should cheese the missiles mechanics by placing ships in front of the targeted ship first, then firing all main guns, then turning the targeted ship to the side and moving it behind the defense line, to drag the incoming missiles through all the PD the player's fleet has.
Doing it each battle, however, is boring, so sometimes you just click on autocombat because "King has many".

This way carriers are deadly enemies in autoresolve and a pitiful nuisance in a tactical battle.

I personally don't know why would you build anything other than torpedo destroyers as Yoral.
Even if you don't upgrade the torpedoes (but why?) you can overlay your two torpedo salvoes (by firing, then moving a bit, and firing the next turn) overwhelming even the late-game point defense.
The alpha damage rules in this game, and Yoral have the cheapest and the most easily accessible glass cannon, that ceases to be a glass cannon even with deflector shields and turbolaser PD.

I haven't found a single case in which a mixed fleet composition outweighed logistics hurdles.
Quantity has a quality of its own.

Uncle_Joe
Posts: 37
Joined: Wed Jul 19, 2017 10:55 pm

Re: Macro-manage Late game large scale

Postby Uncle_Joe » Tue Aug 08, 2017 4:41 pm

And this is why the game needs multiplayer. ;)

All we can do is theorize...we can't put any of it to the test and see what can truly beat what if properly handled.

zolobolo
Posts: 439
Joined: Fri Nov 25, 2016 3:49 pm

Re: Macro-manage Late game large scale

Postby zolobolo » Tue Aug 08, 2017 5:09 pm

I personally do not apply the two exploitation methods: spinning ships around to utilize side and rear shields and pulling in missiles into PD cover.
Well, maybe the later, when moving escort cruisers forward first which more or less is the same results but puts my ships in harms way and consider it to be a fair move.

Yoral torpedo destroyers are great no question but I would counter them with:
- Fighters: effective at taking torpedoes down and faster them the torpedoes so they can make several runs
- Bombers: again faster then torpedoes so if those destroyers do not have escort they are toast in 2 turns
- PD might also work especially when you get to Primary Beans, but even before that, 16 Rapid Fire PD lasers and 4 anti missiles can probably balance out the 8 torpedoes all the while costing 25% less metal and still packing a punch in close combat with their non-depleting laser guns :)
- Capitals can also be stacked full of missiles which would equally be effective (Yoral Battleship can easy unleash 20 misiles compared to the 8 torpedoes of the destroyer). Though admittedly the capital variant is much less cost effective, it can take more punches, and deflect what is left from the torpedo salvoes after its PD salvo with shields

Recommended a slight alteration on Carrier behaviour as it has been bugging me a big as well when the AI sends in their fighters and bombers into a devastating PD screen. Carriers should be more conservative when deploying them and thus become much more unpredictable instead of unleashing all their load at Turn 1 and be "easily" countered

zolobolo
Posts: 439
Joined: Fri Nov 25, 2016 3:49 pm

Re: Macro-manage Late game large scale

Postby zolobolo » Wed Aug 09, 2017 6:46 pm

So this is my first attempt to see a large stack like that in action:

Here are the galaxy settings:
https://www.dropbox.com/s/uja2typ1b0v64mc/Settings.JPG?dl=0
Number of opponents: 5

This is the stance at turn 235:
https://www.dropbox.com/s/a3ezc311adu6xzw/Galaxy.JPG?dl=0

Haven't seen any indication of fleets amassing more then usual:
Humans (bottom left) only have a single fleet consisting of 1 Heavy Cruiser, 2 Light Cruisers, 4 Scouts, 2 Troop Transports and 1 Colony Ship. The other AI opponents seem to have similar fleets with one exception: An Ashdar Imperial fleet has been sighted containing 15 Light Cruisers

Played the game as a pacifist, no war with anyone so far and also the AIs themselves didn't seem to get around fighting each other.

The problem seems obvious:
- Sparse systems means, the AI cannot expand fast enough. I would have though that their fleets will still build up, but they do manage to throw them away at "Monster nests". Eventually they might still pile up but probably need to increase difficulty to the highest level so they get enough resource handicap (as they are very slow in colonization)
- Starting location: Player has started from the middle of the map. I do love this as a starting position, as normally it is more difficult, but in this case it just enables the player to colonies the whole galaxy from a prime location, while the AI is limited by distance of stars and it shortcoming:
- AI does not utilize Outposts nor Lane Amplifiers. Thus the whole Galaxy is ripe for player conquest. Normally Marauders would break the player but I have modded them out. This normally gives the AI a big edge, but in this case it just removes the one force holding the player back

Thus this game in won in around 300 turns and if the AI gets an additional 100 turns or so it might start building up, but that will seem irrelevant from a game perspective as it is over at that time even when playing as pacifist.

The issue I see with this scenario is more with the map layout and AI behaviour.

Solution is obvious:
1. Slightly decrease median distance between star systems so that the AI can expand
2. Let the AI build outposts to systems where it cannot colonies so it can expand further

Next, I need to increase to maximum difficulty and remove the mod on Marauders though I suspect this will again cut back AI progress as I have seen in previous games

Uncle_Joe
Posts: 37
Joined: Wed Jul 19, 2017 10:55 pm

Re: Macro-manage Late game large scale

Postby Uncle_Joe » Wed Aug 09, 2017 7:07 pm

What size fleets are you looking at for the AI? I saw 3 Yoral BBs supported by over 45 DDs in one of my games a bit ago (slightly modded).

And the Ashdar Imperials fielded 10+ Fleet CVs escorted by PD BCs (maybe a half dozen) and another dozen or CLs. That was the biggest battle of that game (and it was epic! lol).

Other than that, I typically a dozen or so combatants by mid-game.

For reference, my mod:

Diplo-changes as per your difficulty mod
Marauder changes per your difficulty mod
Rare Pirates
Normal Galaxy (50 stars)
5 players
Econ/Research set to about halfway between Normal and Hard and halfway between Normal speed and Epic
Ship Cost multiplier at .9

Without doing direct exploit of the AI, it gives a decent game without feeling like cheat spam. I play the Humans most times so it's generally an uphill battle most times. My game with the Ashdar Colonials was a victory, but it wasn't a cakewalk.... ;)

zolobolo
Posts: 439
Joined: Fri Nov 25, 2016 3:49 pm

Re: Macro-manage Late game large scale

Postby zolobolo » Wed Aug 09, 2017 8:08 pm

I am trying to get to those fleet sizes and composition including super-dred and disco-ball like shown in the beginning of this thread.

According to the info so far, it happens on sparse-star maps, where the player is pacifist and at around turn 400 or so.

All I have seen so far in my games was one instance where I met also a Yoral fleet with two battleships and around 50 Frigates. Nothing came close to that when it comes to numbers. Typically a few battleships late-game, and some cruisers or destroyers thats it.

I have been playing all the faction 2-3 times on medium and large maps unmodded, and a dozens of games as Gremak and Colonials on small and 170 stars modded. The largest fleets and best composition I saw with modded version so far as Marauders didnt cut AI down, but always played on non-sparse maps sinceI saw in the beginning that AI doesnt do Outposts.

It would be sweat if we could transport some of these fleet sizes and mixes into our normal gameplay but am afraid it requires a unique bland of galaxy settings (sparse maps and possibly other speicif settings), player strategy (pacifist) and AI bonuses (I have reduced the research bonus for AIs and that might also be holding it back).

I am also happy with the modded difficulty so far though that also requires avoiding sparse start systems and is recommeded to give plenty of breathing space for AI, more and better enemy fleats to fight (on a reasonable tech level) would alwas be welcome though :)


Return to “Testing”

Who is online

Users browsing this forum: No registered users and 2 guests