Author Topic: AI War Beta 7.025-7.028 "Extermination Protocol MkI" Released!  (Read 10243 times)

Offline keith.lamothe

  • Arcen Games Staff
  • Arcen Staff
  • Zenith Council Member Mark III
  • *****
  • Posts: 19,505
Re: AI War Beta 7.025 "Extermination Protocol MkI" Released!
« Reply #15 on: May 12, 2014, 10:27:54 pm »
More information would help immensely, it could be provided a number of ways, from strength to equivalency to a fighter, etc, but almost anything is better then "X times whatever this pulse is" or "this will continue to Y time". I want concrete numbers, not multiples of fuzzy numbers. Doesn't mean concrete numbers are always best, but I hope I convey the spirit of the idea, like how on exos do not provide exactly numbers, but do provide numbers (in the form of large objects) at least give yardsticks to compare other exos.
Exos don't actually tell you how much is going to hit you until it's already launched, but in that case you can't really influence what/whether it launches.

Here you have an ongoing to decision to make, so more precision is appropriate.

I can have it tell you the strength that will launch if the hack stopped immediately.  Unit count is superficially a simpler metric but in practice is a more complex ones: X whats?  Laser Gatlings or Tractor Platforms makes a difference.  Where strength is a much more apples-and-apples thing.

CPAs, in the near future, will probably list strength instead of unit count for that reason (because I want them to go off strength instead of count, and thus not sometimes free 2000 gatlings and sometimes, with the same numbers, 2000 tackle drone launchers).
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 chemical_art

  • Core Member Mark IV
  • *****
  • Posts: 3,952
  • Fabulous
Re: AI War Beta 7.025 "Extermination Protocol MkI" Released!
« Reply #16 on: May 12, 2014, 10:31:44 pm »
More information would help immensely, it could be provided a number of ways, from strength to equivalency to a fighter, etc, but almost anything is better then "X times whatever this pulse is" or "this will continue to Y time". I want concrete numbers, not multiples of fuzzy numbers. Doesn't mean concrete numbers are always best, but I hope I convey the spirit of the idea, like how on exos do not provide exactly numbers, but do provide numbers (in the form of large objects) at least give yardsticks to compare other exos.
Exos don't actually tell you how much is going to hit you until it's already launched, but in that case you can't really influence what/whether it launches.

Here you have an ongoing to decision to make, so more precision is appropriate.

I can have it tell you the strength that will launch if the hack stopped immediately.  Unit count is superficially a simpler metric but in practice is a more complex ones: X whats?  Laser Gatlings or Tractor Platforms makes a difference.  Where strength is a much more apples-and-apples thing.

CPAs, in the near future, will probably list strength instead of unit count for that reason (because I want them to go off strength instead of count, and thus not sometimes free 2000 gatlings and sometimes, with the same numbers, 2000 tackle drone launchers).

I think strength would be best. As strength becomes more integrated across-the-board in various game mechanics, their value as a reference point goes up. Precision isn't as important when they are obstacles that you yourself cannot control, but it is very important if you directly impact when threat is unleashed. It certainly is a step in the right direction.

I had mentioned the "equvilant to X units" so as to provide a second reference point. In my case it would be a triangle unit like the humble fighter. "Response is equal to X strength, or equivalent to Y MK II fighters." This is a holdover for strength still being a...nebulous concept.
« Last Edit: May 12, 2014, 10:33:27 pm by chemical_art »
Life is short. Have fun.

Offline Qatu

  • Full Member
  • ***
  • Posts: 139
Re: AI War Beta 7.025 "Extermination Protocol MkI" Released!
« Reply #17 on: May 12, 2014, 10:49:42 pm »
 I would rather that there was a light but continuous hacking response and all hacking costs were paid upfront and hacking progress could not be allowed to go negative, so for example the superterminal would shut itself down at 0 HaP and a failed hack attempt would have costs but no benefit. Otherwise there's some pretty big exploits available to experienced players but not newer players (like the warhead/martyr+cloaker strategy + superterminal, or superterminal+attritioners+gravity3, or stacking forcefields over superterminal to force all spawns in a tiny point and stunning almost everything with riots2, or pause hacking to get millions of HaP worth of hacking done, or ship farming if they stop being zombies, or useless suiciding zombies if they keep being zombies, or or or or).

 Right now it feels like hacking is generally either way too easy (or even exploitable) or way too hard. That way the very limited amount of HaP directly tied to AIP might be easier to balance.

 Also I hope all normal turrets get the same treatment as the core turrets and the cores get something done to them since they are so extremely overpowering and make the game almost unlosable outside of exo waves. Again kinda feels like an exploit available to experienced players to almost guarantee a win but not known of newer players.
« Last Edit: May 12, 2014, 10:54:58 pm by Qatu »

Offline Winge

  • Hero Member
  • *****
  • Posts: 601
Re: AI War Beta 7.025 "Extermination Protocol MkI" Released!
« Reply #18 on: May 12, 2014, 11:29:37 pm »
More information would help immensely, it could be provided a number of ways, from strength to equivalency to a fighter, etc, but almost anything is better then "X times whatever this pulse is" or "this will continue to Y time". I want concrete numbers, not multiples of fuzzy numbers. Doesn't mean concrete numbers are always best, but I hope I convey the spirit of the idea, like how on exos do not provide exactly numbers, but do provide numbers (in the form of large objects) at least give yardsticks to compare other exos.
Exos don't actually tell you how much is going to hit you until it's already launched, but in that case you can't really influence what/whether it launches.

Here you have an ongoing to decision to make, so more precision is appropriate.

I can have it tell you the strength that will launch if the hack stopped immediately.  Unit count is superficially a simpler metric but in practice is a more complex ones: X whats?  Laser Gatlings or Tractor Platforms makes a difference.  Where strength is a much more apples-and-apples thing.

CPAs, in the near future, will probably list strength instead of unit count for that reason (because I want them to go off strength instead of count, and thus not sometimes free 2000 gatlings and sometimes, with the same numbers, 2000 tackle drone launchers).

I think strength would be best. As strength becomes more integrated across-the-board in various game mechanics, their value as a reference point goes up. Precision isn't as important when they are obstacles that you yourself cannot control, but it is very important if you directly impact when threat is unleashed. It certainly is a step in the right direction.

I had mentioned the "equvilant to X units" so as to provide a second reference point. In my case it would be a triangle unit like the humble fighter. "Response is equal to X strength, or equivalent to Y MK II fighters." This is a holdover for strength still being a...nebulous concept.

A thought on this to see what y'all think:  what if it showed a calculated or estimated strength of reprisal on the tooltip of the hacker itself?  For now, it could show up as a strength number.  Though, as Chemical noted, it would be helpful to eventually translate that into a more user-friendly number.  The superterminal hack would probably have to show up as a message...I'm honestly not sure of a much better way of doing that.  Maybe indicate how much strength in that 'tick' and an estimate for the overall strength of the 'final strike.'


I would rather that there was a light but continuous hacking response and all hacking costs were paid upfront and hacking progress could not be allowed to go negative, so for example the superterminal would shut itself down at 0 HaP and a failed hack attempt would have costs but no benefit. Otherwise there's some pretty big exploits available to experienced players but not newer players (like the warhead/martyr+cloaker strategy + superterminal, or superterminal+attritioners+gravity3, or stacking forcefields over superterminal to force all spawns in a tiny point and stunning almost everything with riots2, or pause hacking to get millions of HaP worth of hacking done, or ship farming if they stop being zombies, or useless suiciding zombies if they keep being zombies, or or or or).

 Right now it feels like hacking is generally either way too easy (or even exploitable) or way too hard. That way the very limited amount of HaP directly tied to AIP might be easier to balance.

 Also I hope all normal turrets get the same treatment as the core turrets and the cores get something done to them since they are so extremely overpowering and make the game almost unlosable outside of exo waves. Again kinda feels like an exploit available to experienced players to almost guarantee a win but not known of newer players.

I understand where you're coming from, but I largely disagree with this (unless your definition of 'light' and 'continuous' differs greatly from mine).  For one, it shouldn't matter if newer players cannot benefit as much from hacking as experts.  They don't get as much use of other game mechanics.  The 'experts' could certainly hack a Superterminal better than I, but we both can use it according to our skills.  Keep in mind that 10/10 by design requires cheese and min/maxing to win (or should; I hear that 10/10 will be getting some bugfixes).  The 7 difficulty, which is the baseline, does not.  Most of what you mentioned has AI counters and/or opportunity costs (except pause hacking, which is a known bug).  Finally, there should be the possibility of a hack going way wrong for the player.
My other bonus ship is a TARDIS.

Offline tadrinth

  • Hero Member
  • *****
  • Posts: 507
Re: AI War Beta 7.025 "Extermination Protocol MkI" Released!
« Reply #19 on: May 13, 2014, 12:12:47 am »
My first game, I was very not-scared of the hacking response up until it started generating Zombie Raid Starships.  I think that was just poor design of my defenses, but those suckers seem way more aggressive about zooming off to find soft targets than any other hacking responses.  That game I had Core Missile turrets and Core MLRS turrets, both of which have bonuses vs Ultra-Light hull, but Mark II+ Raid Starships are immune to missile ammo.  Core Sniper turrets would have countered them nicely, but that controller was out in the boonies.   

Second game I had way better organized defenses so I don't think they were a problem.  That was mostly because I hacked all the capturables that weren't behind my chokepoint, though. 

I like the idea of mini-exo hacking responses.  I always thought the AI should have responses which aim specifically to kill the hacking device itself. 

Also, can the placed hacking devices get AOE immunity?  AI Homeworld + Plasma Siege Starships = no hacking for you while they live, regardless of how many Mark IV shield bearers you pile on top of the hacking device. 

Offline Coppermantis

  • Hero Member Mark III
  • *****
  • Posts: 1,212
  • Avenger? I hardly know 'er!
Re: AI War Beta 7.025 "Extermination Protocol MkI" Released!
« Reply #20 on: May 13, 2014, 01:06:02 am »
My first game, I was very not-scared of the hacking response up until it started generating Zombie Raid Starships.  I think that was just poor design of my defenses, but those suckers seem way more aggressive about zooming off to find soft targets than any other hacking responses.  That game I had Core Missile turrets and Core MLRS turrets, both of which have bonuses vs Ultra-Light hull, but Mark II+ Raid Starships are immune to missile ammo.  Core Sniper turrets would have countered them nicely, but that controller was out in the boonies.   

My favorite was some guy who posted a screenshot to Steam who got 14,000+ Raids. That must have been fun.

Gravity turrets can help, but only so much.
I can already tell this is going to be a roller coaster ride of disappointment.

Offline TechSY730

  • Core Member Mark V
  • *****
  • Posts: 4,570
Re: AI War Beta 7.025 "Extermination Protocol MkI" Released!
« Reply #21 on: May 13, 2014, 01:46:57 am »
Re the superterminal and zombies:

What about a new "zombie level", where they are still unreclaimable, but still show some semblance of intelligence (they can be controlled by the AI for local operations, like with focus firing, prioritization of important targets, etc; but can't join up with higher level forms of intelligence such as threat fleet and possibly seeking out weaker planets)?

A good name for it, IDK. What are some alternate "undead" type things in mythology that don't suffer from quite the level of "brain-deadness" as traditional zombies (could be a foreign (non-western) tradition, as long as it is not too obscure)? That would be a good name for it.
« Last Edit: May 13, 2014, 01:48:31 am by TechSY730 »

Offline ElOhTeeBee

  • Jr. Member Mark II
  • **
  • Posts: 69
Re: AI War Beta 7.025 "Extermination Protocol MkI" Released!
« Reply #22 on: May 13, 2014, 02:23:56 am »
Re the superterminal and zombies:

What about a new "zombie level", where they are still unreclaimable, but still show some semblance of intelligence (they can be controlled by the AI for local operations, like with focus firing, prioritization of important targets, etc; but can't join up with higher level forms of intelligence such as threat fleet and possibly seeking out weaker planets)?

A good name for it, IDK. What are some alternate "undead" type things in mythology that don't suffer from quite the level of "brain-deadness" as traditional zombies (could be a foreign (non-western) tradition, as long as it is not too obscure)? That would be a good name for it.

Ghouls?

Offline onyhow

  • Full Member
  • ***
  • Posts: 122
  • Nuclear powah!
Re: AI War Beta 7.025 "Extermination Protocol MkI" Released!
« Reply #23 on: May 13, 2014, 05:50:40 am »
If you're going to list strength, then you also have to find a way to explain what ship strength mean...

Offline Kahuna

  • Core Member
  • *****
  • Posts: 2,222
  • Kahuna Matata!
Re: AI War Beta 7.025 "Extermination Protocol MkI" Released!
« Reply #24 on: May 13, 2014, 06:13:32 am »
Quote
Specifically, the buffs on 10 are as much as +50%
Hmm
AIP is 181 and wave sizes are only about 2200 even though the AI sends Infiltrators. I remembered wave sizes were about 2500 at ~180 AIP without swarmers. Maybe it's just the Shield Bearers and Eye Bots making the waves easier in this case. (I was playing around with cheats to test a couple of things)
Playground 2.sav

Quote
AI ships are now more deliberate about killing grav turrets that have ensnared them.
Also. Was this supposed to be a nerf to the Gravitational Turrets? Because.. let's say 10 Mark II Gravitational Turrets.. they can tank a lot of damage. Especially if you have Rebuilders rebuilding them. Also their health was buffed. So I wonder how deliberate are the AI ships about killing Gravitational Turrets exactly? Because I was surprised to see how few losses my fleet took.

Zoom in and look at the Bombers and the Gravitational Turrets. It looks like they really want to get rid of those Gravitational Turrets. Now Gravitational Turrets can be used as Force Fields. Kind of.
Playground 3.sav

Other stuff.
I really like the new Warheads. The buffed AOE makes them much more practical than the old ones. Now I haven't tested them vs 20000 ship CPAs yet so I don't know how they would handle those but we'll see. Also thanks to the new Warheads Starships aren't as much of a pain in the ass anymore since I can pop them with an Armored Warhead or two. Normal Mark >= II Lightning Warheads are pretty good at that too.

Hm Mark I Lightning Warhead does 500000000 damage so..
Mark I Lightning Warhead kills 1515 Mark II Fighters
Mark II kills 4545
Mark III kills 7575

A question about the new Carriers.
Why are some Carriers ssoooo damn tanky? Without Warheads It takes forever to get rid of some of them. One had like 700 ships and it was almost as bad as a Mark I Hunter/Killer. Managed to destroy about 30 ships inside it before I had to pop a Warhead. It would have drilled trough the Force Fields protecting my Home Command Station. Also how is the damage calculated? When I shoot at the Carrier the ships inside take damage. Do "all ships inside the Carrier "have Ultra-Heavy hull" like the Carrier"?

aka If an MRLS Turrets shoots a Carrier filled with Fighters would the damage be:
a) 3720*12=44640
or
b) 3720*12*3=133920?
« Last Edit: May 13, 2014, 08:24:39 am by Kahuna »
set /A diff=10
if %diff%==max (
   set /A me=:)
) else (
   set /A me=SadPanda
)
echo Check out my AI War strategy guide and find your inner Super Cat!
echo 2592 hours of AI War and counting!
echo Kahuna matata!

Offline keith.lamothe

  • Arcen Games Staff
  • Arcen Staff
  • Zenith Council Member Mark III
  • *****
  • Posts: 19,505
Re: AI War Beta 7.025 "Extermination Protocol MkI" Released!
« Reply #25 on: May 13, 2014, 10:47:57 am »
I think strength would be best. As strength becomes more integrated across-the-board in various game mechanics, their value as a reference point goes up. Precision isn't as important when they are obstacles that you yourself cannot control, but it is very important if you directly impact when threat is unleashed. It certainly is a step in the right direction.

I had mentioned the "equvilant to X units" so as to provide a second reference point. In my case it would be a triangle unit like the humble fighter. "Response is equal to X strength, or equivalent to Y MK II fighters." This is a holdover for strength still being a...nebulous concept.
Yea, Strength is still nebulous, but I hope to change that.  If there were a better metric I'd use it, but honestly I don't think there is.  "Y Mk II fighters" is a useful reference point but it's not a good "primary description" of how much something is going to hurt.  Partly because fighters are actually some of the easiest stuff to hold off, pound-for-pound (as opposed to, say, bombers).

So what I'd like to do is have the hacking post-response warnings, and the CPA warnings, list Strength on their alert lines, but when you mouseover them it explains that it's simply a measurement and tells you roughly how many MkI fighters that would be (on your cap scale).


and the cores get something done to them since they are so extremely overpowering and make the game almost unlosable outside of exo waves.
If that were the consensus here I'd nerf them, but in past discussions it still seemed heavily in dispute whether core turrets were actually OP.


A thought on this to see what y'all think:  what if it showed a calculated or estimated strength of reprisal on the tooltip of the hacker itself?
You mean the text that pops up when you mouseover the hacker?  Just checking, but why would you want it there rather than the alert box?  Too many things in the alert box already?


What about a new "zombie level", where they are still unreclaimable, but still show some semblance of intelligence (they can be controlled by the AI for local operations, like with focus firing, prioritization of important targets, etc; but can't join up with higher level forms of intelligence such as threat fleet and possibly seeking out weaker planets)?
That's one possibility but it's a lot more work and thus potential bugginess than simply making them normal threat.  Or making them normal threat with a no-reclaim flag.

For now I want to see how much of a problem is there to be solved, rather than hitting it with a complicated solution first.


If you're going to list strength, then you also have to find a way to explain what ship strength mean...
It's simple enough for the tooltip to tell you how many fighters that means, and after a game or two you'll probably have a pretty good idea of how much pain is meant by 1,000, 10,000, or 100,000 strength, etc.


AIP is 181 and wave sizes are only about 2200 even though the AI sends Infiltrators. I remembered wave sizes were about 2500 at ~180 AIP without swarmers.
Hmm.  I suspect other factors causing variance.  I'll make a note to log some waves in 7.024 and then similar waves in 7.025 to see if other changes have watered down the buff (fwiw, there is a wave-size multiplier that used to be 5 on diff 10, and is now 7.5).


Quote
Also. Was this supposed to be a nerf to the Gravitational Turrets? Because.. let's say 10 Mark II Gravitational Turrets.. they can tank a lot of damage. Especially if you have Rebuilders rebuilding them. Also their health was buffed. So I wonder how deliberate are the AI ships about killing Gravitational Turrets exactly? Because I was surprised to see how few losses my fleet took.

Zoom in and look at the Bombers and the Gravitational Turrets. It looks like they really want to get rid of those Gravitational Turrets. Now Gravitational Turrets can be used as Force Fields. Kind of.
Hmm, good point that such heavy-handed "go kill this now" instructions typically lead to non-ideal behavior.  I'll probably have it only do that for half the AI units, maybe 1/4, so the rest will plod on through the molasses in case that's what's best for them to do.


Quote
I really like the new Warheads.
Glad that's working out :)  They are generally more flexible now.


Quote
Why are some Carriers ssoooo damn tanky? (...) Also how is the damage calculated? When I shoot at the Carrier the ships inside take damage. Do "all ships inside the Carrier "have Ultra-Heavy hull" like the Carrier"?
The damage is calculated exactly as if you were shooting the carrier itself.  So the Carrier's hull type and armor and so on are what's used.  The hull type and armor and whatnot of the contained units are completely ignored for this purpose.  They're just a chunk of Health.

At the point in the code where the damage-to-carrier is translated to damage-to-internal-units, it doesn't even know that the damage came from a shot at all (sometimes it didn't), let alone what vs-hull multipliers that shot may have had.

What may be causing such consternation is that you as a player can't tell whether the "next" unit inside the carrier is a fighter or a starship.  If it's a starship it's going to take a lot more hits to knock the carrier's "(102)" down to "(101)".

So perhaps what it should do is use that "next" unit's max health, and the damage done to the carrier since the last internal-unit-kill, for the carrier's health bar.  That way you'll see it zoom down to nothing, then the (102) will count down 1 and the health bar will refill, and then it zooms down to nothing, and it counts down another 1, etc.



Thanks for the feedback, everyone :)
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 RockyBst

  • Full Member Mark III
  • ***
  • Posts: 219
Re: AI War Beta 7.025 "Extermination Protocol MkI" Released!
« Reply #26 on: May 13, 2014, 11:14:06 am »
Ahah, that would certainly explain the seemingly invulnerable carriers which suddenly start dying. We're actually smacking on a MK IV Zenith starship. A little nasty as you can't kill off the low health damage dealers first, but I can't really think of a reasonable way around that. Lowest health first would just be overpowered for the player as things currently stand.

Offline keith.lamothe

  • Arcen Games Staff
  • Arcen Staff
  • Zenith Council Member Mark III
  • *****
  • Posts: 19,505
Re: AI War Beta 7.025 "Extermination Protocol MkI" Released!
« Reply #27 on: May 13, 2014, 12:03:14 pm »
AIP is 181 and wave sizes are only about 2200 even though the AI sends Infiltrators. I remembered wave sizes were about 2500 at ~180 AIP without swarmers.
Hmm.  I suspect other factors causing variance.  I'll make a note to log some waves in 7.024 and then similar waves in 7.025 to see if other changes have watered down the buff (fwiw, there is a wave-size multiplier that used to be 5 on diff 10, and is now 7.5).
Ok, just ran some tests on Diff 10 Vanilla at AIP 180, with a new 7.024 game and then reloading in 7.025.  In 7.024 it was lobbing roughly 550-bomber waves under those circumstances (low-cap test), and in 7.025 it was lobbing roughly 800-bomber waves (with fewer starships).

So I think the overall change worked, though there may be other factors toning it down in other circumstances.  Anyway, we'll see how these numbers play out for a bit and more adjustments can be made later.
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 keith.lamothe

  • Arcen Games Staff
  • Arcen Staff
  • Zenith Council Member Mark III
  • *****
  • Posts: 19,505
Re: AI War Beta 7.025 "Extermination Protocol MkI" Released!
« Reply #28 on: May 13, 2014, 01:08:26 pm »
FYI: my plan is to get 7.026 out as a small release Wednesday sometime before noon eastern-time, mainly to implement some of the hacking changes discussed earlier but there are some other fixes and such already in as well.

Gotta take care of some other stuff during the day, but will probably be tinkering with the hacking stuff this evening so further feedback (on that and other things) is welcome :)
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 tadrinth

  • Hero Member
  • *****
  • Posts: 507
Re: AI War Beta 7.025 "Extermination Protocol MkI" Released!
« Reply #29 on: May 13, 2014, 03:25:02 pm »
When carriers deploy their contents due to the AI having insufficient ships in the air at once, how do they prioritize which ships to deploy?  If they always deploy high strength ships first, that seems like it would help with carriers appearing to shrug off damage because you're beating on a starship and don't realize it.

I'm not sure how carriers unload when there's more than one carrier, though.  Patch notes suggest only one can auto-deploy every 2s per planet, so if the same one autodeploys repeatedly, then the other carriers might not get a chance to dump their starships for a while.