Progressively harder anom NPCs based on completion speed


(Rivr Luzade) #1

Gist of the idea: The faster you complete anomalies, the more intelligent the NPCs get. At some point, entire anomalies stop aggressing your super-/carrier and concentrate entirely on the fighters. In order to relieve the increased aggro on your fighters, you have to leave system and rat in another system. 3 anomalies need to be completed within a single tick until this behavior starts to show.

Under this scenario, only the more competent supers could stay in a single system for hours. More competent carriers would also have problems. In both cases, the strain on the pilots to manage dying fighters increases a lot so that they not only lose income on fighter replacement, they also lose tick value due to recalling of fighters and system changing if they want to keep ratting in a more relaxed way.
3 anoms per tick also keeps any solo sub-capital ratting ship to their normal dumb anom NPCs because no VNIshtar can complete a Haven in a single tick, no Rattlesnake can complete more than 2 in a tick. Smartbomb ratters don’t care about this because they don’t use drones or fighters, neither do Titan or Dreadnought ratters. It also means that many carrier pilots ratting in Havens and Sanctums are unaffected by this because, judging by my own experience, they cannot complete more than 2.5 either. Those are the ones that don’t belong to the 1% of the top 1% in terms of skills. If this turns out to be insufficient, you could reduce the limit to 2 anoms.

It essentially only impacts the biggest money makers without hard-capping their ticks. It has no impact on sub-capitals unlike increased resistances, lower sig-radius and the likes. Furthermore, it does not require the addition of more anomalies to accommodate the same number of people, they just disperse more frequently.


Simple Carrier Ratting Fix
(Cade Windstalker) #2

Couple of problems with this.

First off, you’re tailoring how NPCs react based on the player they face. This gets into some pretty messy state-based stuff on the AI Code side that’s pretty open to potential exploitation. For example if the state is set on warp-in then you just warp in an Alpha in a Corvette before your Carrier and the enemies never get any smarter. If it’s set based on who they’re shooting you can do something similar but with kiting ships that pull aggro first. If it just picked the highest available threat you run into issues with groups running or people trolling by warping into someone else’s site and spiking the difficulty.

Not insurmountable, but enough of a pain to be a mark against something like this.

The bigger problem though is that you’re basically punishing success. If the AI starts ramping up, and especially if they start going for Fighters, then the correct answer here becomes to artificially limit your completion speed to just below the point where you start taking losses. This is neither fun nor rewarding as a mechanic, it’s not even particularly challenging. Things like this feel like arbitrary punishments for good performance, something which a game should reward not punish.

Also, generally speaking, from the place you seem to be looking to target this it doesn’t actually solve the problem with Carriers, since the problem point isn’t the highest end players it’s the majority of middling ones suddenly seeing a spike in income from swapping from whatever they were doing to Carriers.


(Rivr Luzade) #3

I would do it on a per player basis: If the rats for this particular player are at an elevated threat level, they don’t care if there is someone else in the anom first. They just go for the fighters/drones of this high threat player. Even the dumb AI of old anoms/missions can distinguish between threats as many people complain about regularly when they switch aggro instantly to the new player tackling an existing player on grid.
The focus point of the elevated aggro are mainly the fighters/drones of the high threat-level player. Kiting the NPCs with another player should not help if this particular player enters grid.
If it’s kept to a per player base, trolling shouldn’t be possible. I may not be an expert in programming and databases, but I cannot imagine an issue with assigning players a certain state that the NPCs respond to accordingly. This is not meant to increase the alert level of an entire system, but only towards a certain player. If there are several people with the same high threat level running an anom in carriers, the NPCs just turn back to their normal, erratic behavior of aggroing random fighters of both ongrid people.

Yes, it is punishing success if you stay in a certain system. You can just move around to a different system, which is something many people ask for regularly. It may not sound fun, but there are many things that are un-fun in EVE. It may be a slight punishment for good performance but it’s one that you can easily mitigate and it does not affect other people just so that you don’t perform as well any longer.

Depending on what they run, they easily hit the 3 anoms per tick rate very quickly. I recently started pushing a system’s index with my carrier and hit up to 8 anoms per tick. Even the worst of the new carrier pilots can hit 3 very quickly, which in turn discourages them and makes them go back to their subcaps if they find that the massively increased attention strain is too much for them to endure for prolonged periods of time.

It’s not rewarding, for sure; however, it’s a better approach than the other things I mentioned which impact every other ratter more than the actual targets for the nerf. As far as I remember CCP is already working on some kind of change to NPCs that makes it harder for carriers to rat, which is why I see this suggestion as a, to my knowledge, different thinking point for the problem.

I am aware this will sound a bit defying, but people ask for harder NPCs on a regular basis. While this does not make this wish come true for everyone, it does make ratting in fighter based ships a lot more challenging.


(zluq zabaa) #4

I’m pretty sure it is not something we will see tomorrow, but I’d generally like NPCs that are smarter, react to ships players bring and overall don’t just behave like absolute morons who take shot after shot without finding any way to counter. At the very least they should always tackle, use warp-ins/MJD and some form of range dictation mechanism.

As for Supers/Carriers, a first step could be to include a random chance of NPC HICs and/or NPC Anoms dropping Caps on Caps as a normal mechanic.


(Cade Windstalker) #5

That’s not really them distinguishing between threats, it’s just the AI following a pretty simple behavior that’s biased towards new people on grid. If I recall right it was added ages back to stop people from being able to warp in a tanked Battleship, establish agro, and then warp in a bunch of people in paper ships and have them never get scratched.

Also this could still be abused. Whoever has the highest thread either just doesn’t put out drones/Fighters or just kites away so no one else ever gets shot.

I personally haven’t seen any threads asking for more moving around to do PvE. Moving is boring, it’s risky, and in Null a one-system move doesn’t even really lead to anything interesting.

What this would lead to is a buff to renting income, since you now need to rent two systems instead of one…

Anyways, the problem with the system you’re describing is that this sort of sudden switch in agro creates a sweet spot that you want to stay below because the cost of moving systems is always going to be greater than the income lost from simply playing a little worse and not having to do so. This is known as a perverse incentive, since the game should always incentivize the player to play better with increased rewards, even if the increase isn’t linear in relation to the effort expended.

The problem here isn’t the drop in rewards, it’s the way you’re choosing to implement your system. If you simply changed the interactions between the Anoms and Carriers that would be a better change than what you’re proposing, which creates the above mentioned perverse incentive.

It’s very likely that what CCP is considering doing is changing how the anoms behave in general when a Carrier, Super, or Titan is added to the mix so that you still see a direct increase in rewards compared to a subcap but the increase is lower and is more based in something like loot or salvage instead of raw ISK.


(Rivr Luzade) #6

And that is a problem. Adding more and bigger rats with more bounty just keeps increasing the income because you just need a carrier or super from the available pool in system to help blap it. Adding bigger rats to increase risk for carriers is not going to solve the income problem at all. And if they do not have bounties, you just kill off the anom around it and ignore it, like you do already if you cannot destroy a dreadnought in your haven. This approach is not going to solve anything.

I see your issue with the kiting exploit. However, if they do this, this kiter is not going to earn anything, which is fine with me. The people using that guy as aggro trigger complete anoms faster, which increases the danger of hitting the 3 anom wall and get more aggro, too. If the carriers stay below the limit, that is fine, too. The main point of any nerf is a reduction of income for the highest level earners, not for anyone else. The increased aggro also does not force someone to stop ratting or artificially reduces bounties or loot on NPCs just because you are a carrier. It simply makes it more challenging to earn the money you want.


(Cade Windstalker) #7

I think I may have been unclear somewhere, it’s only a problem if it results in a net increase in ISK per hour. If you make it so that a Cap Ship spawns consistently when a Carrier warps into the site, and that Capital takes long enough to kill relative to its bounty or has a relatively low bounty but valuable drops then it’s a good change.

If it takes longer to kill that slows down the player’s income.

If it drops loot instead of rewarding a big bounty payout then that both scales downwards the more people are Carrier Ratting and doesn’t inject raw ISK into the economy.

You can prevent just destroying the Anom around it by either restricting the respawn of the Anom until it’s killed, giving it a longer point range, or making it a trigger for a wave to proceed.

You can make it uneconomical to group up to kill it by giving it a large HP pool rather than a large active tank, so the increase in benefit from bringing in a second person is negligible since that can only do slightly better than halve the kill time as opposed to right now where doubling your DPS significantly more than halves the kill time because you overwhelm the active tank on the spawn.

The point I’m trying to make here isn’t about any specific exploit and how to counter it, it’s about how many problems this sort of AI mechanic creates. For every one issue you and I can come up with in this thread the actual implementation will have 3-4 more that will have to get fixed and at that point you’re starting to look at an absurdly complicated mechanic, with a stupid number of exceptions and rules just to make it work, and someone will probably still find a new way to break it every few months.

That’s why CCP and most other game devs have stopped trying to institute weird AI behaviors like this to deal with certain player behaviors. They either just crank up the numbers or rework the base behavior of the encounter because once you get into line-item stuff like this it’s turtles all the way down.


(Rivr Luzade) #8

You see, these examples are exactly the kind of things harming smaller fish more than they harm pro ratters. Supers can still wear them down in no time, whereas carriers like Archons and Chimeras (poor souls who rat on those) already have massive issues with the existing dreads. Solutions like these make it easier for the extreme pros to keep their massive income flowing (be it items or direct ISK) but make it a lot harder for the smaller ratters.

Not to mention that this kind of approach is easy to game, for instance, by warping supers around the anomalies, spawning the big NPC capitals and thus effectively reserving lots of anomalies for yourself without any sub-capital or carrier being able to run them instead. Sure, the NPC capital can despawn after a while, but that does not help the capital pilot who warps in and gets tackled at 80 km by that NPC.

I find these approaches a lot easier to game than a limited-to-a-player NPC behavior adaptation. Either way, it looks like any changes open just more unenjoyable cans of worms.


(Cade Windstalker) #9

So give a different spawn for Supers and Titans than Carriers. That’s part of why I tried to make it clear that I’m divorcing this entirely from existing Capital Spawn mechanics. The main reason a Carrier currently has issues with a Dread spawn is because the Carrier has trouble dealing with the Dread’s strong active tank.

If you’re warping a Capital around an enemy system to troll them then they don’t care about the anomalies they’re going to be trying to kill you. If you’re doing that to your allies then you’re probably going to get kicked or otherwise punished pretty quickly.

Also as you point out that could be solved by having the ship despawn. If you make it so the ship despawns fairly quickly when there’s no Capital in the site then you can’t use that to reserve sites. Alternatively have it point at long range and quite aggressively (or both). You then can’t easily warp around to multiple sites like that.

The difference between these sorts of changes and a change in AI behavior is one of complexity and ease of tuning.

AI requires complex rules to function even for simple behaviors and because you’re defining an indefinite behavior pattern rather than something definite like a set spawn or the Rat’s HP or bounty it becomes very hard to remove all of the bad cases from your system.

In comparison something like a simple flag that says “If Carrier, spawn: doom” is very simple and easy to adjust, as are the stats of whatever doom that flag spawns (doom here should be taken as tongue in cheek, I don’t intend anything to spawn stuff that can’t be beaten by the ship that spawns it here).

On top of that something that targets fighters is naturally quite spiky in terms of effect because Fighters cost quite a bit relative to what’s being earned and losing even one T2 Fighter an hour has a significant impact on net income. This means that performing at below the level of fighter losses will almost always be better than risking losing a fighter at all.

In comparison installing a “speed bump” of HP for the player to shoot through is quite easy to tune and has a very clear effect on income that’s clear to both the player and the devs. It also scales cleanly with player skills and skill level, which is a nice bonus.


I’d also like to elaborate that it doesn’t have to be a Capital spawn type of mechanic, it just needs to be something that adjust the site difficulty for the Carrier in a more numerical way as opposed to behavioral. They could just as easily spam a bunch of higher HP cruisers to soak up time for less bounty provided. The problem with that is that it doesn’t feel as good to have little ships spammed at you in response to your big ship showing up. Having a big ship spawns is generally more fun because it creates a big aspirational target that has a big payout next to its name, even if the math shows that you’re actually losing money compared to if the big target wasn’t there.


(Felyx Ravencroft) #10

An intriguing notion in principle, and though I’ve read most of what has been discussed above, I haven’t thought this through in detail (plus I lack first-hand experience in those specific situations.) Personally, I like the notion of “diminishing returns” being implemented in some way, as this applies in EVE in many other contexts: consider, for instance, skill training - many skills give you steady increase steps but at increasingly longer intervals; still worth it to train them, and of course many open up other options at higher ranks, but the basic point stands.

As for making those anomalies harder (perhaps after a certain point with a hot-drop reinforcement, or whatever), how about this for a rather simple notion that also has logical flavour: those high-bounty (pi)rat(e)s getting hit on by an egger in a carrier call in bountyless mercenary backup, perhaps even aligned with Thukkers or what-not and thus carrying standings implications. Then the player has to decide whether to kill off the current target and perhaps another, then GTFO (depending on tank endurance et al), or to deal with the mercs (for no reward other than wreck salvage and loot, and perhaps even a standings hit) before being able to complete the anomaly.

Alternatively, instead of mercs, the reinforcements could be allies from some other pirate faction with radically different resist and damage profiles, to disrupt the player’s overspecialised fit. harder to “justify” lack of bounty in this case, though, although I quite like the notion of that spanner in the works/curve ball of unanticipated damage+resists (which could be randomised to some extent, but always be different enough to be… “inconvenient”.)

As I said, I lack experience specific to this context, so maybe this is a dumb idea - or perhaps it can be modified in some way to make it feasible. What appeals to me is that it comes with a logical rationale suggestive of the NPC enemies being more than just ATMs with shields and armour, but rather entities with connections and resources befitting the hefty price tags on their heads.