View Issue Details
| ID | Project | Category | Date Submitted | Last Update | |
|---|---|---|---|---|---|
| 0022364 | AI War 2 | Note To Test | Dec 13, 2019 9:44 pm | Dec 25, 2019 1:06 pm | |
| Reporter | RocketAssistedPuffin | Assigned To | RocketAssistedPuffin | ||
| Status | resolved | Resolution | fixed | ||
| Product Version | 1.017 Random AI Types | ||||
| Summary | 0022364: Puffin misc ToDo list | ||||
| Description | Putting here both as a reminder and let anyone else input. 1): Refine NastyPick (Ion Cannons, Mass Drivers, Eyes, etc) more with new system. (Done) 2): AIs like Ensnarer to have proper non-turret-defense groups. (Done) 3): Metal to claim and metal to repair super units sorted. (Done) 4): AI planets to have notably scarier defenses on higher difficulties (emphasis on the scaling - do not touch 7 and below). (Kind of done). 5): Make AI planets more puzzle like, punishing if you FRD on higher difficulties without being overwhelming. Bonus multipliers potential safe-ish lever. 10x minimum? Dires are 20x. 6): Investigate Factory refleet speed - casualties on offense are replaced immediately at too fast a rate, allowing for players to attrition worlds. Note: For tutorial, Homeworld needs a unique Factory with increased speed, so it still flows nicely, but doesn't slow progress down. Must test with automation settings. (Code thing from Badger that impacts this). 7): Connected with above, counterattacks rarely fire and are game ending when they do. Rarely fire, due to people simply being able to continue to attrition the planet while stalling it (i.e falling back is often the sup-optimal option). (Same as above). 8): Military Stations to not be so amusingly absurd versus things it can engine stun. (Technically done though still somewhat problematic). 9): Investigate AI budget allocation for any oddities/flaws. Warden overflow being sent to waves causes issues with newbies being confused at enormous waves coming in out of seemingly nowhere - very hidden mechanic. Perhaps defense budget should be solely for that. (Done, nothing of note) 10): Investigate Reserves. Very chance based, easy to cheese apparently? (Badger fixed the cheese bit). 11): War Harvesters / Siegers - slow to gain new groups of these, but technically stackable as they never leave and are now replenishable. Might be abusable if one was patient enough. (StarKelp smited this). 12): Sort out Dyson Bulwark and Bastion weapons and strength. (Done) 13): Possibly: Consolidate Dark Spire units. Very tanky, numerous, low DPS. Unsure if much benefit to this. Would look cleaner. (Meh). | ||||
| Tags | No tags attached. | ||||
|
|
To issue 6&7: I don't think reinforcement speed needs to be any lower, nerfing the Engineers has mostly halted that. Instead I think it would be a good idea for the AI using selected ships (good vs high-albedo, high-mass, etc, on higher difficulty) and going directly against the centerpieces so the player needs to protect their centerpieces. This means that the players can't reinforce if they don't take care of their transports. Usually these can even be used to soak up damage from enemies when putting them close enough, since they are rarely focused by the enemy as far as I can see, by other things than Sabots and OMDs. |
|
|
They were intentionally made lower priority for various reasons, like being used entirely as meat shields and the AI not killing any actual combat units. I think about the reinforcement speed due to some cases I've seen where it'll go from losing pretty hard to winning trivially, just from spamming more ships at the thing. You can essentially get to a point where the health-per-second you add into a fight outpaces the AIs DPS, because you're building MK6-7 units for cheap. Take a MK1 Vanguard for instance. It has 24k health, and costs 4,500 metal to build. A Factory outputs 2,000 metal per second. So roughly every 2 seconds, a unit with 24k health comes out, meaning this single Factory is outputting 12k health-per-second into a battle. Now a MK7 Vanguard. MK7 has 4x health, and the cost does not increase, so that Factory outputs 48k health-per-second by itself. Add in a Combat Factory or two (which level up and output more), some more Factories, Engineers, and it gets pretty darn high. A single Military World, and the starting Rejuvenator Combat Factory leads to something like...8,000 just from the Factories, then 2,000 from the 10 Engineers from Rejuvenator, so 10,000 total metal per second, at only MK1. If it was MK3, the Rejuvenator is doubled, and the Engineers are tripled, from doubling their speed then increasing cap by 50%. That'd then result in 10,000 from Factories, then 6,000 from Engineers, so 16,000. That's 8 Factories output, so...96k health per second, if building MK1 Vanguards. For MK7 Vanguards, 384k health-per-second. For reference, the Overlord Phase 2 is 280k DPS, not including the Overdriver which isn't really able to be counted. That's just one Combat Factory and one planet. The Vanguard is admittedly an extreme ship to use, but it highlights fairly well I think how nuts it can be. Of course, this is draining metal, though I've seen people with insanely high storage and income. That isn't including ships regenerating shields, or Engineers repairing for a fraction of the cost (shield repairs are free, and hull is half cost - with units being usually 2/1 ratio, this means the cost to repair is usually a third of the cost of a new one). It also isn't including drones, or zombifying anything. Hence the comment on the Counter-Attacks never triggering, or when they actually do, being game ending. If you have enough metal lying around, you can slap the planet to death. EDIT: You could use Vanguard Hydras actually. The unit + the heads has the same health, but also far higher damage output + the heads are temporarily invulnerable. That would be an amusing attrition cannon. |
|
|
Now that you put it that way, you're right. I hadn't even noticed it before, I like to crush a system with more ships than the enemy has, or special fleets good against whatever defenses are in there. But it sounds like one could solve this in a number of ways: - Reduced factory speed (although that might make it more difficult to fleet up even if not under attack), so instead how about reducing factory speed if the target reinforced is on another planet? That makes combat factories more important to have on the same planet, and as far as I know the AI hates those a LOT and goes for them as primary targets, so it would be pretty difficult to keep them alive in a (without reinforcements) losing battle. - Added metal cost per MK level, like in AIWC. Then maybe some people would actually buy the metal income tech... Now that I think about it, how about this as a combination: - Up to 2.5x higher metal cost at Mk7 - Factories reinforce 50% slower if the centerpiece reinforced is on an adjacent planet - Logistical Command Stations cause their factories to have 2 max jumps instead of just 1 as reinforcement range, however at that range the reinforcement speed is only 25% of what it would normally be. That sound reasonable? |
|
|
Reducing the speed is simple, but...aye, I'm concerned of all the Classic "netflix time" problems. A penalty for the centerpiece being on a different planet would lead to people just storing them on the planet with the Factories, I think. Units would be taken out of Officers and put into custom Flagships to avoid the penalty there. Then you'd just do mass waves of units rallied in manually. If I understood that right? For the metal cost per MK level, that was a thing but was removed early on in Fleets. Information here: https://wiki.arcengames.com/index.php?title=AI_War_2:_The_Arrival_of_Fleets#No_More_Mysterious_Metal_Increases_As_Marks_Go_Up The 2 max jumps is interesting, though I recall being told the code for that kind of thing is...ow? At least, it was for supply when that was a thing. I had an idea about extended range and it was shot down in under a minute. One from me, one from a friend who I mentioned this to: 1): Higher mark units simply take longer to build. Don't cost more, just take longer. MK7 is 4x health, so it'd take 4x time to build. This already has a concern raised that it'd make things like Frigates agonisingly slow to rebuild. (Me) 2): Fleets in combat (i.e losing units) reinforce slower. This'd impact attritioning, without touching the refleet after a battle, or acquiring a new Fleet, etc. (Friend) |
|
|
That's possible too, though increased time to construct is not that different from more cost too, and goes into exactly the same "netflix time". Gotta think some more about that. |
|
|
Metal income wouldn't be a problem there however, which I recall was one of the main blocks in Classic. But yes, it is a concern. The second one I like the most myself. |
|
|
How about this, as a mix of all the things: -> Factories construct as a base only 1000 metal/second. Home Command and Logistical Command Station gain the new Advanced Factory which constructs 2000 metal/second and reinforces up to 2 hops away instead of just 1. -> Factories rank up with the Engineering Tech and if the planet is Mk'd up, gaining 15% more construction (thus around 2x speed, exactly 205% at Mk7). -> Factories reinforcing fleets on a different planet receive a modifiers to construction speed based on certain factors. --- If the fleet is in combat (the centerpiece is ineligible to repair), multiply construction speed of all ships of that fleet by 0.5x --- An additional multiplier per hop. This also counts the planet the fleet is on UNLESS(!) the factory is on the same planet too: ------ If the planet is owned by the player or by a faction they are allied with, the factor is 0.9x ------ If the planet is neutral the factor is 0.75x ------ If the planet is hostile the factor is 0.5x So for an extreme example: Reinforcing a fleet in combat on a hostile planet with an additional hostile planet between the factory and the fleet: 0.5*0.5*0.5 = 12.5% normal reinforcement speed. As an example for being under attack at an adjacent, player-owned planet, the fleet also being in combat: 0.5*0.9x = 45% normal reinforcement speed. With the addition of not counting the negative hop modifier if the fleet and the factory are on the same planet this means that a combat factory doesn't automatically run at an additional -50% speed just if it's on a hostile planet too, but that would just be my opinion. It's already hard enough keeping these squishy things alive. |
|
|
I like the notion of increasing the build time (but not metal cost) for higher mark units |
|
|
This is indeed very smart -- what a great lateral solution! |
| Date Modified | Username | Field | Change |
|---|---|---|---|
| Dec 13, 2019 9:44 pm | RocketAssistedPuffin | New Issue | |
| Dec 13, 2019 9:44 pm | RocketAssistedPuffin | Status | new => assigned |
| Dec 13, 2019 9:44 pm | RocketAssistedPuffin | Assigned To | => RocketAssistedPuffin |
| Dec 15, 2019 5:33 pm | RocketAssistedPuffin | Description Updated | |
| Dec 16, 2019 3:46 pm | RocketAssistedPuffin | Description Updated | |
| Dec 16, 2019 4:00 pm | RocketAssistedPuffin | Description Updated | |
| Dec 17, 2019 8:22 am | NRSirLimbo | Note Added: 0055018 | |
| Dec 17, 2019 9:30 am | RocketAssistedPuffin | Note Added: 0055019 | |
| Dec 17, 2019 9:30 am | RocketAssistedPuffin | Note Edited: 0055019 | |
| Dec 17, 2019 9:41 am | RocketAssistedPuffin | Note Edited: 0055019 | |
| Dec 17, 2019 9:54 am | RocketAssistedPuffin | Note Edited: 0055019 | |
| Dec 17, 2019 10:53 am | NRSirLimbo | Note Added: 0055021 | |
| Dec 17, 2019 11:15 am | RocketAssistedPuffin | Note Added: 0055023 | |
| Dec 17, 2019 11:16 am | RocketAssistedPuffin | Note Edited: 0055023 | |
| Dec 17, 2019 11:20 am | RocketAssistedPuffin | Note Edited: 0055023 | |
| Dec 17, 2019 12:16 pm | NRSirLimbo | Note Added: 0055024 | |
| Dec 17, 2019 12:18 pm | RocketAssistedPuffin | Note Added: 0055025 | |
| Dec 17, 2019 2:45 pm | NRSirLimbo | Note Added: 0055026 | |
| Dec 17, 2019 2:57 pm | BadgerBadger | Note Added: 0055027 | |
| Dec 17, 2019 7:27 pm | RocketAssistedPuffin | Description Updated | |
| Dec 18, 2019 10:47 am | Chris_McElligottPark | Note Added: 0055045 | |
| Dec 25, 2019 12:30 pm | RocketAssistedPuffin | Description Updated | |
| Dec 25, 2019 1:05 pm | RocketAssistedPuffin | Description Updated | |
| Dec 25, 2019 1:06 pm | RocketAssistedPuffin | Status | assigned => resolved |
| Dec 25, 2019 1:06 pm | RocketAssistedPuffin | Resolution | open => fixed |
