Discussion: AI hitting you when you're down

[Kahuna]

This wave should probably trigger the same way the "Why do you throw your live away" AI taunt does to avoid 2000 ship waves @ 5 minutes in-game time. If I spam Neinzul Youngling ships at the beginning or lose 97% of my fleet to the first 10 diff Starfleet Commander wave (much harder than Spire Hammer).

[LaughingThesaurus]

Seems similar to that "short term AI progress"/"human threat meter" of another recent thread, except scaled with recent human losses rather than recently seen human firepower on AI planets.

Both ideas seem intruiging.

Unlike that idea, I think this shouldn't be a meter of some kind. The AI can get mad about you having firepower gradually, but this seems more like a spontaneous reaction to "Man we just saw like 400 human-flavored explosions, lets go for the finishing blow!"
I should probably also mantis that human threat-o-meter at some point, but I'm usually too lazy to actually do that. =/

That would still be a meter (over recent human casualties), but one with a much, much faster "decay", such that would wouldn't really see it build up over multiple attacks or anything like that.

Technically, if either the conditions or the response are scaled somehow based on game state data and not just a boolean or a constant over the game (like faction intensities), then a "meter" (visible or not) is technically involved somehow, even if it is reset at the end of each "encounter".

Oh yeah, in technical terms, it would definitely be a meter. I just meant something that absolutely should be invisible. Seeing the AI's reaction to you dying in real time would be a bit odd. Seeing the AI gradually get more concerned the more heavily you're attacking would be kind of like how the humans can just see the AI Progress indicator.

[chemical_art]

On the one hand interesting. I don't like the idea of the result of a nasty fleetball fight that you still win but wounds you results in a strike back out of the ether. It would narrow tactics, and increase the desire to blob even more and nerf high cap / self destructive units. That it sounds eeriely like reclamation for the AI without regard to all sorts of things. I really don't like the idea of the fight to clear out threat over a wormhole has the potential to wound the player greatly yet spawn something that will contribute to yet more wormhole threat.

But I still like the idea. What if the units drawn upon from the AI were from the strategic reserve? That's a nice cap of units that is not needed that often, and yet is well suited to enable bursts of threat.

[keith.lamothe]

@chemical_art: I'm aiming for this as an AI plot, rather than a base-game mechanic, mainly because it's like hybrids in the sense of striking at the player holding the tempo all the time.  But also because it's the sort of thing that won't sit well for a lot of players but (hopefully) be really cool to others.

That said, hopefully it will be cool for many

[chemical_art]

@chemical_art: I'm aiming for this as an AI plot, rather than a base-game mechanic, mainly because it's like hybrids in the sense of striking at the player holding the tempo all the time.  But also because it's the sort of thing that won't sit well for a lot of players but (hopefully) be really cool to others.

That said, hopefully it will be cool for many

I'm not disliking the idea (in fact I love it) I just don't want it to come out of the ether.

So if the ships were pulled from the strategic reserve, or from special forces, then I like it. Because when those ships are knocked down, you hinder the AI's ability in costing it reinforcements. Even a hybrid where half the AI response is drawn from reserves / special forces and the other half were from the ether I would really like.

It allows the AI to more frequently lob sledgehammers rather then stingers at you without relying on an offensive tactic from no where.

It gives the player a lever against repeated counterattacks.

[keith.lamothe]

I sympathize, but there are issues the other way too; a couple:

1) SF and the SR come from the ether.  This pulling from those would just be one step removed.  It would basically be a mini-CPA (or not-so-mini) coming in response.

2) If you have a "press here to drain SF and SR ships into my chipper shredder" button, then you could probably find a way to effectively use this new mechanic to neutralize both of those.

[chemical_art]

I sympathize, but there are issues the other way too; a couple:

1) SF and the SR come from the ether.  This pulling from those would just be one step removed.  It would basically be a mini-CPA (or not-so-mini) coming in response.

However, they don't simply not exist before they are not needed, unlike this counter attack. If you knock down the SF then you have a window of freedom to more freely move in AI turf or make attacking them more manageable. It's not the same because those SF would exist anyway.  Unless you are saying the SF regrows to full strength in less then an hour or two then even a skeleton human fleet can react without its full force, but the desire to blob too much will result in the "window of oppertunity" being lost. It would make a strategic blunder hurt, but at least enable some tactical raiding. The strategic reserve I thought was supposed to recharge over time as well.

2) If you have a "press here to drain SF and SR ships into my chipper shredder" button, then you could probably find a way to effectively use this new mechanic to neutralize both of those.

If you make so the lost to the human is hard to the relative to the cost to the AI, then the player should be behind to start. The nuke is a "chipper shredder button" but the cost to enable said button is high enough to make it desired rarely. This button, however, you can only avoid by not blundering.
 

[chemical_art]

The core of my fears I come to realize is I don't want to incur greatly blobbing I've decided.

It's well established blobbing causes the least casualties for a situation. So a mechanic that triggers due to losses encourages it even more.

I thought counter lever of some sort to provide a silver lining to a disaster would make it not so feared, but I guess in practice it would not.

[TechSY730]

It's well established blobbing causes the least casualties for a situation.

Unless the attack fails. In that case, the losses are much greater.
Sure, it is less likely to fail when blobbing, but if it does, ouch.

[KDR_11k]

Depends, blobbing isn't perfect, you often take ships with you that are quite vulnerable but not so expensive that you care.

But yeah, the fact that blobbing basically amounts to smashing the AI with superior force in a game where you're supposed to cower in fear of the AI's arsenal makes it seem just wrong. The Eyes are inconsistent and apparently don't care that you're waltzing around with planet-sized death machines as long as you don't have more death machines than it has targets for them. If you actually try to use fleetships against Eyes you run into the problem that no matter how small and specialized your force is, if you win against the defenders you will eventually pass the threshold. Plus they negate high-cap ships, as if those needed any further reasons to be avoided.

[chemical_art]

It's well established blobbing causes the least casualties for a situation.

Unless the attack/defense/whatever fails. In that case, the losses are much greater.
Sure, it is less likely to fail when blobbing, but if it does, ouch.

But...the plot this is meant to target, the most experienced players, know with a degree of certainly which side will win in a fight 80% of the time.

I cannot remember the last time I was wiped when I didn't expect it and on offense I can run easily.

It does raise a good point though, if the AI launches an attack that requires you to use both your fleet and turrets to hold the line, the AI can be rewarded with a second attack?

In my situation, this means 98% of my fleet wipes are on defense, so it really just serves to box me in since my wipes are only when an ai attack in the right location means either I lose a strategic irreplaceable object or lose the fleet. Given the AI can then quickly launch a second attack on that strategic object I lose it anyay if I fight!

That...for me at least, doesn't cause me to reconsider wiping my fleet, it means the RNG determines more then ever whether I lose my factory IV.

[TechSY730]

But yeah, the fact that blobbing basically amounts to smashing the AI with superior force in a game where you're supposed to cower in fear of the AI's arsenal makes it seem just wrong.

Just a slight technicallity, actually, you are cowering in fear of the attracting the AI's attention to the point where it uses its arsenal at you. As such, as one reviewer put it, making the AI think you are weak, and then suprising it with a huge force is a core "theme" in the game.
That said, it would be cool if the AI had a short term "boost" in responciveness/reaction/whatever if it sees that you are blobbing, and thus you are not weak even though you haven't caused it much long term pain yet. Like a short term boost in AIP or something. (There are some other threads about this sort of thing)

This also plays into that old discussion for ways to reward keeping planets out of alert state beyond just lessened reinforcements. (As mentioned, specific case buildings like AI eyes and raid engines are too few and far between to curb the general tendency)

The Eyes are inconsistent and apparently don't care that you're waltzing around with planet-sized death machines as long as you don't have more death machines than it has targets for them. If you actually try to use fleetships against Eyes you run into the problem that no matter how small and specialized your force is, if you win against the defenders you will eventually pass the threshold. Plus they negate high-cap ships, as if those needed any further reasons to be avoided.

Yea, I think having the AI eye scale its count with caps of the ships in combination with counting "mega-ships" (like golems) as multiple ships would greatly fix AI Eye oddities.
However, this does come at the cost of making the determination of whether your force will trigger the eye or not much harder, and makes it much harder to tell how much you need to retreat to stop an activated eye from staying activated.

So, do we want accurate modeling, or simplicity, or a compromise between the two?

[keith.lamothe]

It's well established blobbing causes the least casualties for a situation. So a mechanic that triggers due to losses encourages it even more.

It varies.  If by blobbing you mean "send in everything to the shortest range of any ship in the blob" then you're going to take awful casualties when an enemy fort is involved.

Anyway, yes, my suspicion is that blobbing (in the sense of "circling the wagons") will actually be encouraged by this rather than discouraged.  To some extent that's fine with me, because the result is more "play smart" (to reduce casualties) rather than "play while using or not using specific-tactic-X" (even if X is "blobbing").  And often that kind of play involves taking in a subset of the blob first or only.

But it's definitely an "AI Plot for people who want it" thing, because I don't want folks to have to care all that much about watching their casualties in the normal game.  That'd drastically change the average micro required to win.

In a lot of ways this is an experiment to show people what it's like when you have to nanny   But there's enough masochists here to like it, too.

[KDR_11k]

It should probably only count casualties in AI territory. Losses on the defense are expected, the price for a complete wipe is that the AI gets free reign to wreck your territory (and likely end the game), only attacks need a cost.

[TechSY730]

It should probably only count casualties in AI territory. Losses on the defense are expected, the price for a complete wipe is that the AI gets free reign to wreck your territory (and likely end the game), only attacks need a cost.

Agreed on this. Losses on defense are not a sign of what this idea was intended to target.