AI War Beta 6.011 "Nemesis of the Dubious Honor" Released!

[Dichotomy]

I would like to point out that the starshipI fleet is where most of our fleet strength comes from early game. Even mkI starships rarely die while in the fleet, and are a noticeable part of our fleets even lategame with no new unlocks. I would like to see more incentive to unlock higher marks, but the mkIs are about right balance-wise.

[Diazo]

In my opinion, Zenith and Spire won out because their Mk I costs knowledge while the other starships you start with Mk I unlocked.

Having said that, I need to revise my earlier statement about their balance point.

Their are a bunch of comparison scales that exist in the game, but at the core it all boils down to "how much damage can I do with this chunk of resources?".

In the early game, before you reach the AI's higher mark worlds, and before the AIs worlds have had time to get serious reinforcement, the high health of the starships shines through as that high health means you never (or very rarely) lose a starship and so have a better return for your resource expenditure with them as they will survive for a very long time.

As you move into mid and late game, the AI's defences are high enough that it can kill your starships pretty easily and so it does not matter that the starships have better health then fleet ships, the higher DPS output of the fleet ships means they are preferred as they will output more damage before dying.

Also, as you move to mid and late game, your economy is usually big enough you can absorb the loss rate of fleet ships to replace them relatively easily as compared to early game where replacing a cap of ships is a bigger deal.

So, combining these two things as you progress through the game, lighter AI defenses earlier and a tighter economy that moves towards heavier AI defenses and a more powerful economy changes the balance between starships and fleet ships as the game progresses.

For me, this means that (excluding the current zentih and spire starships that are outliers) in the early game starships are a better investement for my resource cost but as I hit mid game (or early-mid game really) fleet ships catch up and then by late game fleet ships are a significantly better investment.

I'm actually toying with the idea of making starships "survive at all costs". By this, I mean change their balance point from the (roughly) current 50% DPS for 200% HP to something like 35% DPS for 300% HP of the equivalent fleet ship.

I'm not sure how practical that would be in game though, the starships surviving to come back would be nice but if they are not killing anything, they are not worth the resource investment.

D.

[Faulty Logic]

I think the mkI starship balance point is near perfect. But there isn't much incentive to unlock higher marks, thus the advanced constructor proposal.

As you move into mid and late game, the AI's defences are high enough that it can kill your starships pretty easily

I have never observed this. If I lose more than a couple starships, it usually means I made a mistake.

I'm actually toying with the idea of making starships "survive at all costs". By this, I mean change their balance point from the (roughly) current 50% DPS for 200% HP to something like 35% DPS for 300% HP of the equivalent fleet ship.

I really don't like that idea. A lot of their utility is "kill x then get out," and this would make it harder.

then by late game fleet ships are a significantly better investment.

Yeah, around when we acquire an advanced factory.

[chemical_art]

The tricky thing is that I'm getting the strong impression that further increasing (much less tripling) post health is going to really slow the game down. (edit: specifically in a way folks really don't enjoy)

You do? Right now if a bomber or frigate cap of mk (X-1) attacks with its bonus agaisnt guard post mk (X), it dies in one volley. For fighters, it takes two volleys. Doesn't seem to really slow the game down.

In fact, MK II bombers and frigates against mk IV guard posts, take down over 90% of their health in one volley. Even against MK V guard posts, it takes only two volleys for MK II units to take one guard post.

[Diazo]

The tricky thing is that I'm getting the strong impression that further increasing (much less tripling) post health is going to really slow the game down. (edit: specifically in a way folks really don't enjoy)

You do? Right now if a bomber or frigate cap of mk X attacks with its bonus agaisnt guard post (X-1), it dies in one volley. For fighters, it takes two volleys. Doesn't seem to really slow the game down.

In fact, MK II bombers and frigates against mk IV guard posts, take down over 90% of their health in one volley.

That's nice in a reasonably ideal world.

Now toss in any of the myriad things that prevent that ideal world from happening. (Eye, gravity guardian, tractor guardian, and many more.)

The problem is that to make a post a challenge against it's counter in an ideal setup would turn it into an absolute stonewall when you could not bring your ideal counter along.

Yes, in most cases you can arrange for the ideal counter to come along, but I find it is not exactly uncommon that I need to kill a guard post but conditions around it prevent me from bringing its counter along to kill it.

D.

[chemical_art]

That's nice in a reasonably ideal world.

Now toss in any of the myriad things that prevent that ideal world from happening. (Eye, gravity guardian, tractor guardian, and many more.)

The problem is that to make a post a challenge against it's counter in an ideal setup would turn it into an absolute stonewall when you could not bring your ideal counter along.

Yes, in most cases you can arrange for the ideal counter to come along, but I find it is not exactly uncommon that I need to kill a guard post but conditions around it prevent me from bringing its counter along to kill it.

D.

While true, the same could have been said before the posts were buffed. There were still plenty of sticky situations that could develop. I guess they develop more often now, but it isn't because they are tougher, it is because they hit harder.

[Jonz0rz]

As others have pointed out, starships are weak right now since they have nothing to compete with the advanced factory. That thing makes it go from spending 6k knowledge for 3 marks (assuming an ARS design) to spending 6k for 7 marks. Starships are in a decent place early on, spending 6.5k for 5 marks, but they fall off lategame with no advanced factory equivalent. Again, you don't have to do anything about this right now (and in fact, you probably shouldn't, seeing as it wasn't on your "patch goals"), but it's something to consider down the road.

The guard posts are difficult to comment on right now because Spire/Zenith starships are dominating games so completely (last game had a wave 1 MK.II spire ship to homeworld on 7/7, the one I loaded up earlier squished me on the first exo wave with 12 MK.IV spires, almost all of my new games have had a MK.II or higher spire/zenith ship guarding a planet adjacent to my homeworld, and so on).

The whole consolidated caps thing may be causing more problems than it's worth. I'd like to try the Spire/Zenith starships with just the buffs you gave them, and no consolidated caps or forced higher marks.

[Winge]

Welcome to the forums, Jonz0rz.

I would try increasing the caps or decreasing the marks, but not both--at least at first.  A big part of the problem is that the AI gets to choose a very strong Mark II ship in a Mark I slot...potentially multiple times per wave.  That said, I think they should either keep the higher mark or the lower cap benefit to try and avoid falling right back into their previous spot (almost never used).

[nitpik]

Enjoying the new more noticable guard posts already. Just started my first game with the new patch, and ran into a Mk IV Missile guard post under a forcefield on a world next to my Homeworld. It gets a bonus against both bombers and Raid Starships.

I much prefer the AI War definition of "fun" to the boring old one everyone else uses

[Faulty Logic]

We adopted it from Dwarf Fortress. Also, the proper spellings are Fun or FUN.

Yeah, the guard posts feel right, now.

[keith.lamothe]

We adopted it from Dwarf Fortress. Also, the proper spellings are Fun or FUN.

Or, the first time you run into a Wrath Lance or whatever, !!FUN!!

[Faulty Logic]

I focused on spelling. Punctuation is a whole other issue.

[keith.lamothe]

I focused on spelling. Punctuation is a whole other issue.

In DF, the exclamation mark isn't punctuation, it's part of the alphabet

[Toranth]

I have some questions about the behavior in the Counter-Champion Dirty Tricks logic.

Current game is a 9/9 8-Champion game.  Current AIP is 10, Champion Hull size is still Frigate.

Here's what I see in the log file:

3/9/2013 12:29:29 PM (6.011)
-----------------------------------
Doing Counter-Champion-Dirty-Tricks check; Game Time: 1:05:09
Checking for adding points to dirty tricks bank.
baseValue = 100
numberOfChampionsInGame = 8
humanChampionHullSizeFactor = 1
difficultyFactor = (standard non-linear difficulty factor for special threats) = 2.5
lowAIPFactor = (if AIP < 30, then 1, else if AIP < 50, then 0.9, and so on until 0.3 for AIP >= 150) = 1
dirtyTricksPointsToAdd = baseValue * numberOfChampionsInGame * humanChampionHullSizeFactor * difficultyFactor * lowAIPFactor = 2000
CurrentAccumulatedPoints = CurrentAccumulatedPoints + dirtyTricksPointsToAdd = 6000


3/9/2013 12:29:37 PM (6.011)
-----------------------------------
Doing Counter-Champion-Dirty-Tricks check; Game Time: 1:05:17
Spent 6000 points to spawn a threatfleet nemesis unit; remaining total = 0

Every minute, it gains 2000 points.  Every 5 minutes, it spends 6000 points to spawn a Nemesis frigate.  But it never buys more than one at a time, and it never buys more often than once every 5 minutes.  So, it's accumulating a large buffer of points it isn't spending (20,000 points at time 1:30:00).  It also overspent at time 1:00:00, going to -4000.

I know you intend to add more stuff for the AI to spend points on later.  But is there a reason it isn't spending all the points available to it?
Is this roughly the accumulation rate you intend?  9/9, AIP 10, 8 Champions = 2,000 points = A Nemesis every 3 minutes?
Finally, is this Nemesis spawn for a single AI, or for both?  Right now, it looks like BOTH of the AIs are getting a new Nemesis unit every time the Counter-Champion-Dirty-Tricks triggers (effectively twice the rate the logs lists).

[keith.lamothe]

I have some questions about the behavior in the Counter-Champion Dirty Tricks logic.

(...)

Every minute, it gains 2000 points.  Every 5 minutes, it spends 6000 points to spawn a Nemesis frigate.

Here are the three primarily-important lines from the log:

(...)
numberOfChampionsInGame = 8
(...)
difficultyFactor = (standard non-linear difficulty factor for special threats) = 2.5
lowAIPFactor = (if AIP < 30, then 1, else if AIP < 50, then 0.9, and so on until 0.3 for AIP >= 150) = 1
(...)

In other words:
1) If you were playing with 1 champion instead of 8, the "bank" would accumulate enough points for 1 nemesis every 24 minutes, instead of every 3.
2) If you were playing diff 7 instead of diff 9, the interval would be 7.5 instead of 3.
3) If your AIP were >= 150, the interval would be 10 instead of 3 (reduction is 10% per 20 AIP over 10).

It's just the combination of you playing with a _ton_ of champions, on a high difficulty, while remaining very-low-AIP (thus the AI has more to respond to your champions because it doesn't see the rest of your stuff as a threat), that you're getting hammered that badly.  The main thing is the champion count.

But it never buys more than one at a time, and it never buys more often than once every 5 minutes.  So, it's accumulating a large buffer of points it isn't spending (20,000 points at time 1:30:00).

Yea, I should change it to spawn multiple nemesis units if it has that much backlog, I just hadn't anticipated the case; thanks for noting that.

It also overspent at time 1:00:00, going to -4000.

Yea, the overspending is intentional for when I add other cases where it can spend more (or less) than 6000 on something.  The current rule is that as long as the bank is >= zero it can pick any available option regardless of cost.  But naturally if that puts the bank in the negative it can't do anything at all until it's back to >= zero.  So the costs function as more of a "cooldown" than a cost per se.  If I had it work as a cost I'd have to build in special mechanisms to prevent it from always buying the cheap stuff and never saving for the more expensive stuff.

Is this roughly the accumulation rate you intend?  9/9, AIP 10, 8 Champions = 2,000 points = A Nemesis every 3 minutes?

That's faster than I figured it would be, but it doesn't sound out-of-the-ballpark.  You're basically asking for a beating with that; I imagine you can still win but it gives you some kind of run for your money.  If player experience shows that it's just overwhelmingly hard with no reasonable counter, then I can tone it down.

Finally, is this Nemesis spawn for a single AI, or for both?  Right now, it looks like BOTH of the AIs are getting a new Nemesis unit every time the Counter-Champion-Dirty-Tricks triggers (effectively twice the rate the logs lists).

It only triggers for the first AI.  Are you seeing solid evidence that the second AI is doing it?