Author Topic: Rebalancing normal AI guard posts  (Read 7238 times)

Offline keith.lamothe

  • Arcen Games Staff
  • Arcen Staff
  • Zenith Council Member Mark III
  • *****
  • Posts: 19,505
Rebalancing normal AI guard posts
« on: March 04, 2013, 12:01:55 pm »
Ok, it's been brought up a lot in the past and now it won this current worst-unit poll, so I'm working on rebalancing the normal AI guard posts into something a bit more dangerous.  I had a few ideas of varying craziness to change how they're handled conceptually and all that, but decided that they don't really need that, at least at this time.  So mainly focusing on just changing stats and so forth.  Anyway, I did have a few questions to get clearer on what exactly seems to need the attention.

May as well go by guard post:


Group 1:

Short Range Guard Post
MLRS Guard Post
Missile Guard Post
Passive Guard Post

I'm guessing these are just overall underwhelming in the kind of threat they bring to the table (the "passive" one in particular, heh).  My thought is to:
1) rename the "Short Range" into "Needler"
2) rename the "Passive" into "Laser"
3) balance all four along the lines of the four main lines of human turrets (needler/laser/mlrs/missile) with the same hull-type-bonuses and probably same bonus magnitudes, dunno), with each post equalling some % of a ship-cap of the equivalent turret type/mark.  Exactly what that % should be, I'm not sure.  Something tells me you don't want a single mkIV guard post to be as powerful as a full cap of mkIII (or mkIV, if they existed) turrets :)  Probably not even as powerful as a full cap of mkIV fleet-ships (turrets are balanced as 3x as powerful as fleetships, for reference).


Group 2:

Arachnid Guard Post
Command Station Shield Guard Post
Stealth Guard Post

I honestly don't hear a lot about the Arachnid.  It has a lower seeding chance than most.  Does it need similar firepower to whatever group 1 winds up at?

On the command-shield-protector post, these seem sufficiently influential as-is since they make taking the planet a more multi-stage procses; do they need to actually kill stuff too?

Stealth guard posts are already pretty substantial thorns in the side in some cases, and having them shoot stuff would just break their cloak.  Do these need to be nastier?


Group 3:

Spire Shield Guard Post
Counterattack Guard Post

I'm guessing these are already plenty frustrating as-is :)


Group 4:

Special Forces Guard Post

Not seeded in the normal way and the SF mechanic is really what's threatening about SF, not the posts.  They also don't have multiple mark levels so they'd need to be balanced as a particular mark (I'd say mkIV).  Do these need to be bumped up to the power of whatever group 1 winds up at, or higher?


Thoughts?
Have ideas or bug reports for one of our games? Mantis for Suggestions and Bug Reports. Thanks for helping to make our games better!

Offline Faulty Logic

  • Hero Member Mark III
  • *****
  • Posts: 1,194
  • Bane of the AI
Re: Rebalancing normal AI guard posts
« Reply #1 on: March 04, 2013, 12:20:45 pm »
Group 1
I'd want these as the equivalent of a cap of fleetships (of their mark) in firepower (with corresponding multipliers), and about 1/4 the health. Not quite sure what that translates to in terms of turrets, though.

Short-range => fighter
MLRS => missile frigate
Passive => bomber
Missile=> sentinel frig (half-cap)

I would also like for them to have the hulls of their fleetship, bringing the RPS aspect of the triangle back into some semblance of relevance.

Group 2
Arachnid => these should be anti-starship (medium/short range, 10 million damage against 1 target every few seconds).
Command shield => leave as is.
Stealth => ERASE FROM EXISTENCE. These don't change my tactics, except to add a tedious tachyon sweep at the end of planetary conquest.

Group 3
Fine as is.

Group 4
Remove from eye networks (so you can kill eyes without AIP). This slight player buff is moderated by the fact that the eyes got tougher from the other buffs. They don't need firepower, they do other things.
If warheads can't solve it, use more warheads.

Offline keith.lamothe

  • Arcen Games Staff
  • Arcen Staff
  • Zenith Council Member Mark III
  • *****
  • Posts: 19,505
Re: Rebalancing normal AI guard posts
« Reply #2 on: March 04, 2013, 12:39:27 pm »
Group 1
I'd want these as the equivalent of a cap of fleetships (of their mark) in firepower (with corresponding multipliers), and about 1/4 the health. Not quite sure what that translates to in terms of turrets, though.
Just taking the standard fighter as the baseline for that, this would result in all 4 posts going from 700k base health to 3M base health, and non-bonus-dps going from at-most-13k (that's the current MLRS post value; short-range and missile posts have values under 2k) to 49k.  I'd probably go with new bonuses parallel to the turret types of roughly the fighter's magnitude (6x).

I would ask for a sanity check, but I already know there is none.  Insanity check?

On changing their hull types, yea, I could do that.

Quote
Arachnid => these should be anti-starship (medium/short range, 10 million damage against 1 target every few seconds).
So it would have the cannot-target-small-stuff flag?  Actually I guess it would be easier if it didn't have that, as you could saturate its target list with small stuff to keep your starships alive.

Quote
Stealth => ERASE FROM EXISTENCE. These don't change my tactics, except to add a tedious tachyon sweep at the end of planetary conquest.
I'm thinking I can probably just make it not cloak itself, but still cloak the stuff near it.

Quote
Group 4
Remove from eye networks (so you can kill eyes without AIP). This slight player buff is moderated by the fact that the eyes got tougher from the other buffs. They don't need firepower, they do other things.
Hmm, if the code isn't too intransigent on that one, and I suppose it wouldn't be too confusing if I added a note to the special forces post description about not needing destroyed to kill an eye (or maybe amending the eye description).  Though special-forces-captain AIs use a lot of these posts in place of normal ones, so it could lead to some pretty-easily-killed eyes.  Not a big deal, I guess, as SF is its own intergalactic eye ;)
Have ideas or bug reports for one of our games? Mantis for Suggestions and Bug Reports. Thanks for helping to make our games better!

Offline Faulty Logic

  • Hero Member Mark III
  • *****
  • Posts: 1,194
  • Bane of the AI
Re: Rebalancing normal AI guard posts
« Reply #3 on: March 04, 2013, 12:53:24 pm »
Quote
I would ask for a sanity check, but I already know there is none.  Insanity check?
Are those the mkI numbers? If so, excellent. Otherwise, more firepower.

Quote
so it would have the cannot-target-small-stuff flag?
Either way would work. Ideally, it would target starships if possible and fleetships otherwise, but starship-only would be fine.
Adding a scout-0 multiplier might also be a good idea, so transports don't get instagibbed.
If warheads can't solve it, use more warheads.

Offline Diazo

  • Master Member Mark II
  • *****
  • Posts: 1,717
  • I love/hate Diff 10
Re: Rebalancing normal AI guard posts
« Reply #4 on: March 04, 2013, 12:54:35 pm »
I'm actually not too concerned about the power of the guard post itself.

If I get a guard post alone against my fleet, it's dead, no if ands or buts. If the guard post still has its defenders, those defenders are still the priority. Unless the guard post is buffed to Core levels (which they are not going to be), the defenders are always the bigger threat.

My concern is buffing their health and eyes.

You get to mid game and you've got a Mk III world that has been on alert for a bit while you deal with other things.

It has an eye and several hundred other ships defending it but because of the eye you have to raid the guard posts.

The start and the end will be easy. (At the start several hundered AI ships means you can bring your entire fleet. At the end a solo guard post will just die.)

I'm concerned about the middle. The AI is down to not that many defenders left so you can only bring low numbers of ships in to raid. Currently because of their state you can get in, take the losses to focus fire the guard post down and get out. If the guard post gets a major health boost this will become a lot more problematic.

Now, this is not necessarily a bad thing. The point I'm trying to make is that guard posts are one of the few units that are directly tied to other units and so changes made to them will have a significantly larger ripple effect then previous changes made to units that existed more on their own.

I know you've already said you are not going to do it, but I'd like these left on the long term list as the threat is not the guard post itself, but the guard posts defenders. I'd prefer the raw stats to stay the same but each guard post gained abilities that helped their defenders somehow.

Again, a guard post that is alone dies against my fleet. You can't boost them high enough to get around that without turning every system into a core-world assault level battle and that would grind the game to a halt very fast.

D.

Offline keith.lamothe

  • Arcen Games Staff
  • Arcen Staff
  • Zenith Council Member Mark III
  • *****
  • Posts: 19,505
Re: Rebalancing normal AI guard posts
« Reply #5 on: March 04, 2013, 01:08:10 pm »
Are those the mkI numbers? If so, excellent. Otherwise, more firepower.
MkI numbers, everything linear from that.  I'd toyed with the idea of breaking the normal progression and having mkIII posts be 2x as strong as linear would make them, and mkIV like 4x-8x as strong as linear would make them... but I think that could get nasty.

Quote from: Diazo
Again, a guard post that is alone dies against my fleet. You can't boost them high enough to get around that without turning every system into a core-world assault level battle and that would grind the game to a halt very fast.
Right, these can't ever be fortresses-in-their-own-right or the game just halts.  The question is what price you have to pay for that relatively-quick-kill, and whether you have to think about how you do it.  If they get base DPS and bonus multipliers like a cap of fighters, a mkIV MLRS post (with bonus vs light) will do 50k*6*4=12M damage per second to a pile of fighters.  That's enough to kill a mkI cap of fighters in a little over a second (salvo time is 4 seconds and projectile travel time is non-zero, so not that fast, but you get the point).  So you wouldn't want to just mob those with your blob unless you didn't need the fighters anymore, etc.
Have ideas or bug reports for one of our games? Mantis for Suggestions and Bug Reports. Thanks for helping to make our games better!

Offline Diazo

  • Master Member Mark II
  • *****
  • Posts: 1,717
  • I love/hate Diff 10
Re: Rebalancing normal AI guard posts
« Reply #6 on: March 04, 2013, 01:14:30 pm »
Hmmm.

In that case, what about a small overall stats boost, I'm only taking 20-25% or so, but increase the magnitude of the multipliers.

Maybe reduce the number of multipliers some of the posts have (do any have more then 4 multiplier types?)

What this would do is not make the post that much more of a threat overall, but it would encourage thinking about your fleet composition so that you did not attack a post with a x15 light multiplier with fighters.

On the other hand, we are now getting into the increasing micro territory?

D.

Offline Faulty Logic

  • Hero Member Mark III
  • *****
  • Posts: 1,194
  • Bane of the AI
Re: Rebalancing normal AI guard posts
« Reply #7 on: March 04, 2013, 01:22:43 pm »
Quote
50k*6*4=12M
Off by an order of magnitude, there.

Quote
On the other hand, we are now getting into the increasing micro territory?
Yes. But almost any planet you could take without micro you still can, just with higher cost.
If warheads can't solve it, use more warheads.

Offline keith.lamothe

  • Arcen Games Staff
  • Arcen Staff
  • Zenith Council Member Mark III
  • *****
  • Posts: 19,505
Re: Rebalancing normal AI guard posts
« Reply #8 on: March 04, 2013, 01:27:04 pm »
In that case, what about a small overall stats boost, I'm only taking 20-25% or so, but increase the magnitude of the multipliers.

Maybe reduce the number of multipliers some of the posts have (do any have more then 4 multiplier types?)

What this would do is not make the post that much more of a threat overall, but it would encourage thinking about your fleet composition so that you did not attack a post with a x15 light multiplier with fighters.
I try not to have multipliers over 10x, and rarely that high.  Anyway, a simple operation on the existing attack stats won't really work since the existing numbers are pretty pathetic and vary by a factor of 10 from post type to post type.  So here's an example of what I'm doing right now (these aren't in stone, largely based on my gut feeling and Faulty Logic's feedback above) :

Quote
** Short Range Guard Post:
*** Renamed To Needler Guard Post.
*** Hull Type from Medium => Light.
*** Base Health from 700k*mk => 3M*mk (about 1/5th a cap of fighters).
*** Base Attack Range from 6000 => 7500 (same as needler turret).
*** Shots-per-salvo from 1 => 5.
*** Base DPS from 1267*mk => 50k*mk (right around a cap of fighters).
*** Vs-Hull-Type bonus list from "Heavy (2x), Artillery (3x)" => "Heavy, UltraHeavy, Structural, Artillery (all 6x)" (same types as needler turret, same multiplier as fighter).


Quote
On the other hand, we are now getting into the increasing micro territory?
Having to consider hull types is part of the "approved micro", you could say.  Obviously it's not a zero-micro game, and having the visual recognition of "ah, that is an MLRS guard post" map to "ah, that will eat my fighters as a light snack, better send in the other stuff" isn't too big a deal.


Quote from: Faulty Logic
Off by an order of magnitude, there.
... meddlesome facts!

;)

So a mkIV MLRS post would take out about 1/3rd of a cap of mkI fighters per 4-second salvo.

Ok, so maybe not a light snack, maybe a quick lunch.  You'd still probably have a good reason to check the label on the chipper-shredder you were shoving your fleet into, to adapt deployment.

Have ideas or bug reports for one of our games? Mantis for Suggestions and Bug Reports. Thanks for helping to make our games better!

Offline KDR_11k

  • Hero Member Mark II
  • *****
  • Posts: 904
Re: Rebalancing normal AI guard posts
« Reply #9 on: March 04, 2013, 02:03:26 pm »
I wouldn't make the guard posts too strong, the main threat are supposed to be the guarding units. The AI guardians are already the AI's turret equivalent.

I'd rather see the guard posts be modifiers for their guards so the post itself stays rather harmless (though maybe a bit harder to kill) while the enemies you face near it are a bit different depending on the post.

Offline Aeson

  • Full Member Mark II
  • ***
  • Posts: 176
Re: Rebalancing normal AI guard posts
« Reply #10 on: March 04, 2013, 02:09:06 pm »
I personally think that current guardposts are fine, and that the defenders of a guardpost are where their real threat should come from anyways. The idea that a single guardpost would have equivalent firepower to a same-mark cap of equivalent fleet ships could make fighting the various AI vessels near it very bad, especially if they've had a chance to build up their numbers or the Special Forces show up (both of which can already be fairly difficult situations).

I'd be fine with the moderate increase in power that Diazo suggests (though maybe not quite the multiplier suggested, if that turns out to be along the lines of "obliterate entire caps of fleet ships because they didn't notice what kind of guardpost was close enough to the wormhole to hit ships coming through it"), but increasing guardpost firepower nearly fourfold seems excessive to me. Three million health per guardpost isn't something I mind, since usually guardposts die with the first volley from the fleet anyways, but significantly increasing both health and firepower means that I'm more likely to lose significant amounts of my fleet from straying into range or fighting in range of the guardpost (and I can't not fight in range of the guardpost if I want to clear the system of AI ships, regardless of what I want to do to the guardpost).

There's also a bit of a problem if the guardpost and the ships the AI spawned to protect it happen to have complementary types - e.g., fighters spawning on something that obliterates missile frigates.

With regards to the Passive Guard Post - this is essentially just a guard post for stalling the player a bit. Either it needs to have enough health and armor to really stall the player long enough to be inconvenient (like a Wormhole Guard Post), or it needs to have some effect which would cause a player not to want to kill it (e.g., Special Forces Alarm, Counter-Attack on random planet, causes an AI progress increase if killed, triggers a CPA on death, spawns 1000 threat ships on death). Also, if you go in for boosting its health significantly, it really depends on how you want it to inconvenience the player as to what else needs to happen with it - if you want it to inconvenience a player neutering a system, then high health is sufficient; if you want to inconvenience players who take a system, requiring it to spawn on resource nodes and allowing it to survive command station destruction would be sufficient, as would giving it the command station shield effect.

I also think that the range on the Short-Range Guard Post might be too short or that they have bonuses against the wrong things. I most often use fighters and bombers to dispose of any guardposts because they're the fastest of the triangle ships and more importantly fighters are cheap to replace and so make good cannon fodder for drawing fire away from the bombers. If the Short-Range Guard Post is to keep its current bonuses, there needs to be a reason why I might bring something with that hull type into its firing range, but missile frigates do just fine at clearing any fighters from around the post from outside its range, and fighters and bombers are better against the post anyways.
« Last Edit: March 04, 2013, 02:10:49 pm by Aeson »

Offline keith.lamothe

  • Arcen Games Staff
  • Arcen Staff
  • Zenith Council Member Mark III
  • *****
  • Posts: 19,505
Re: Rebalancing normal AI guard posts
« Reply #11 on: March 04, 2013, 02:16:20 pm »
Quote from: KDR_11k
I wouldn't make the guard posts too strong, the main threat are supposed to be the guarding units. The AI guardians are already the AI's turret equivalent.
The guardians are supposed to be, but they're pretty easy to bait.  AI turrets were mainly a problem due to large numbers (and thus CPU consumption) and being replenished by reinforcements.  Guard posts don't have that issue.

Anyway, they're not being made all that strong, just brought up to a level where they actually make some kind of difference in a fight.

Quote
I'd rather see the guard posts be modifiers for their guards so the post itself stays rather harmless (though maybe a bit harder to kill) while the enemies you face near it are a bit different depending on the post.
That would be nice, and I could do munitions boosting and perhaps armor or speed boosting (but those are hard to do in a localized area and have them be significant), but the "covers nearby units with forcefield" and "cloaks nearby units" roles are already filled.  For that matter, so is the "protects command station" role.  There's also "repairs nearby units" but that tends to not matter much mid-combat.  Making up new mechanics for a post to buff its nearby ships would be pretty consuming of both dev-time and CPU cycles.

All that said, I agree that the path to moving the guard posts into the "truly interesting" category is not by making them main combatants, but through other, more nuanced stuff.  For now I'm settling for moving them out of the "mostly irrelevant" category :)


Quote from: Aeson
There's also a bit of a problem if the guardpost and the ships the AI spawned to protect it happen to have complementary types - e.g., fighters spawning on something that obliterates missile frigates.
I tend to look upon such accidental "emergent" combinations as a source of variety and entertainment, actually ;)

Drawing the post's defenders away or simply smashing the post with a single salvo shouldn't really be much harder than it used to be, it's just not as trivial.


So, here's what I have for the guard posts; not on the wiki changelist yet as I haven't done the commit.  Let me know if anyone thinks something is just way off or if there's something else they want done (no promises, but now's a good time to say it).

Quote
* AI Guard Posts (not the core ones, the normal ones), in honor of tying for first in the latest what-needs-a-buff poll (after many months of feedback that these were not threatening enough) :
** Short Range Guard Post:
*** Renamed To Needler Guard Post.
*** Hull Type from Medium => Light.
*** Base Health from 700k*mk => 3M*mk (about 1/5th a cap of fighters).
*** Base Attack Range from 6000 => 7500 (same as needler turret).
*** Shots-per-salvo from 1 => 5.
*** Base DPS from 1267*mk => 50k*mk (right around a cap of fighters).
*** Vs-Hull-Type bonus list from "Heavy (2x), Artillery (3x)" => "Heavy, UltraHeavy, Structural, Artillery (all 6x)" (same types as needler turret, same multiplier as fighter).
** MLRS Guard Post:
*** Hull Type from Heavy => Medium.
*** Base Health from 700k*mk => 3M*mk.
*** Base Attack Range from 7000 => 12000 (same as MLRS turret).
*** Shots-per-salvo from 11/21/31/41/51 => 30.
*** Base DPS from 1700*mk => 50k*mk.
*** Vs-Hull-Type bonus list from "Light (16x), Swarmer (8x), Neutron (4x)" => "Light, Swarmer, Neutron, UltraLight, CloseCombat (all 6x)" (same types as MLRS turret).
** Missile Guard Post:
*** Base Health from 700k*mk => 3M*mk.
*** Seconds-per-salvo from 7/6/5/4/3 => 3.
*** Base Attack Range  from 17k/27k/37k/47k/57k => 27k (same as Missile turret).
*** Base DPS from 13k*mk => 50k*mk.
*** Vs-Hull-Type bonus list from "UltraLight (8x), Medium (6x)" => "UltraLight, Medium, Polycrystal, Neutron, Composite (all 6x)" (same types as Missile turret).
** Passive Guard Post:
*** Renamed To Laser Guard Post (hey, at least it's not now the Passive-Aggressive Guard Post).
*** Hull Type from Structural => Heavy.
*** Base Health from 1.4M*mk => 3M*mk.
*** Shot type from flare => laser.
*** Base Attack Range from 8000 => 9000 (same as Laser turret).
*** Shots-per-salvo from 1 => 4.
*** Base DPS from (negligible) => 50k*mk.
*** Vs-Hull-Type bonus list from (well, nothing) => "Refractive, UltraHeavy, Polycrystal, Heavy (all 6x)" (same types as Laser turret).
** Anti-Starship Arachnid Guard Post:
*** Hull Type from Light => Polycrystal.
*** Base Health from 1.4M*mk => 3M*mk.
*** Base Attack Range from 10k/12k/14k/16k/18k => 14k.
*** Seconds-per-salvo from 15 => 5.
*** Base DPS from 10k*mk => 2M (bear in mind these cannot target small stuff, it's an anti-starship sort of thing).
*** Vs-Hull-Type bonus list still empty, but now has 0x multiplier vs Scout hull so it doesn't just gank your transports.
*** Shot type from laser => energy wave (so now like a sentinel frigate shot; just aesthetic for the most part, makes it stand out better).
** Stealth Guard Post: didn't actually need a buff, but:
*** Renamed to Cloaker Guard Post.
*** Are now un-cloakable; they still cloak stuff near them but will always be revealed themselves.
** Special Forces Guard Post: also didn't need a buff (the SF ships are the important part of SF, really), but these no longer count for purposes of the "if no AI guard posts are present, an AI Eye self-destructs" rule.
Have ideas or bug reports for one of our games? Mantis for Suggestions and Bug Reports. Thanks for helping to make our games better!

Offline LordSloth

  • Sr. Member Mark III
  • ****
  • Posts: 430
Re: Rebalancing normal AI guard posts
« Reply #12 on: March 04, 2013, 02:29:32 pm »
Here's a thought for the special forces guard posts: make them rarer, appearing in about one of every three or five systems. In exchange, they make the local guarding fleetships function similar to the Tag Teamer AI.

An idea I will regret: change short-range guard posts into gravity guard posts.

In addition to it's anti-starship capability, the Arachnid could be given a small amount of interceptor drones to combat fleetships, still vulnerable to missile frigates. Although these are light hull enemies, I usually take them out with fighters during the stage 1 assault where I clear the medium hull guardians. Certainly, the missile frigates could pop them in a sixth the time, but short of a transport I usually make up the time with the fighter's higher speed.

Edit: This post is outdated, whoo.
« Last Edit: March 04, 2013, 02:32:05 pm by LordSloth »

Offline Hearteater

  • Core Member
  • *****
  • Posts: 2,334
Re: Rebalancing normal AI guard posts
« Reply #13 on: March 04, 2013, 02:47:53 pm »
My rough feeling is if I have a cap of Bombers the same mark as the Guard Post, I should lose roughly 10% of my Bombers killing the Guard Post.  Without changing the hull type, a cap of Bombers is going to basically one-shot a Guard Post even with a health bump.  So maybe give them a small bump from 700k to 2m (at Mark I), and then work from that.  Given that health bump, a Mark I Guard Post will need to output at least ten shots dealing 1.56m total damage in the time it takes the Bombers to close to range and their attacks to hit.

Group 1:
Pretty much the basics.  Around 2m health and they should kill 10 Bombers of equal mark before dying.

Group 2:
All buffed to 2m health for consistency.  No change to CS Shield.  Give Archnid about x10 damage (3.5m damage so it would 2-shot a Heavy Bomber Starship).  I don't see a need for restricting it to small ships only given its 15 second reload.  Stealth Guard Posts should be reworked.  Make them Stealthed Barracks that get their reinforcements inside and they release them all when you build a CS in the system.  Basically a trap for taking the system.  Also, they don't count for keeping Eyes alive.  For a bonus, let them system-wide EMP player ships for 5 seconds/mark when they pop.  You could always search them out before and possibly killing them in that manner might avoid the EMP (you'd still need to kill the ships just like with popping a carrier).

Group 3:
No change.

Group 4:
SF Guard Posts should be unpleasant.  I'd like to see a good damage buff, in line with at least group 1 and possibly a special effect like Translocation.

Offline LordSloth

  • Sr. Member Mark III
  • ****
  • Posts: 430
Re: Rebalancing normal AI guard posts
« Reply #14 on: March 04, 2013, 02:51:51 pm »
Oh, and Arachnid guard posts should at least be a threat to Shadow Frigate Champions, so a structural multiplier would be good at clearing down those force fields.