Chokepoint Balance

[TechSY730]

To counter those runners wouldn't be any different from exo-s, just drop 2 or so forcefields to prevent any stragglers from leaving. Even better if with your military command center can cover your wormholes.

It would still encourage chokepoints, except now you want chokepoints with not a crazy amount of nodes.

True, my idea alone would not be enough to encourage defense in depth or give the AI sufficient tools to deal with/bypass chokepoints.

However.

The forcefields counter thing would still take extra forcefields (now for your wormholes to your "inner planets" as well), cutting into your already tight ff cap.

It wouldn't do jack against FF immune units, which would have a higher chance of doing this.

If a bunch of bombers decide to join in with runners, or wave sizes start getting large, those forcefields won't last long.

So it still would be a tactic by the AI to reward defense in depth, and give the AI a way to bypass chokepoints.

But your observation that it alone would not be enough for the AI's "chokepoint bypassing" "bag of tricks" (some other new mechanics may be needed as well), as well as this not discouraging chokepoints (despite encouraging defense in depth) are true.

[Diazo]

Okay, having vented my spleen in my ranting posts yesterday in this thread, then starting a game on the realistic map type, and after a night to mull it over I'm not sure there is actually something worth changing here.

I think I was probably the primary force pushing for this in other threads before this one started, but I had forgotten a very important fact.

The lattice map type is as much as an outlier as the snake map type, it is not supposed to be balanced.

Because I've been playing lattice pretty much exclusively for the past year I had forgotten this and so when I enabled exo-waves for the first time in forever, I did it on a lattice map and because I had no experience otherwise I thought "exos are overpowered!" as I had no chokepoint to stop them.

The question here needs to be changed, on the average maps (so do not include the 3 high-connection maps of lattice, and cross-hatch A/B or the 3 low-connection of snake, vine and {i'm not sure}), is the game balance with chokepoints the way it is?

I think the answer is pretty close to yes.

When I started the game yesterday on the realistic map type and I started looking at my defensive options, there were chokepoints everywhere. Being used to lattice which essentially does not have a chokepoint anywhere, even the realistic map type had a plethora of defensive options where I could chokepoint.

So, lattice (and cross-hatch) maps should be classified as outliers, they are maps in which you are deliberately giving up your ability to chokepoint for that game just as a snake map is an outlier in that you give up your ability to pick your direction to expand.

Having said that, I would still like to see the AI get more abilities to warp forces past chokepoints, but they need to stay both small enough and rare enough that they are interesting diversions, not a full wave getting dumped past your defenses.

D.

[eRe4s3r]

Thinking about it I find that I am absolutely not agreeing to local turret caps either way, that is the worst idea so far ;P Might as well entirely remove turrets then unless the cap is well above 50 of each per planet.

70% of my fun is creating the bestest defense chokepoint ever. When you make some obtuse rules like "Go 1 further" I make the chokepoint 1 or 2 worlds larger. Doesn't really fix anything ;p

Any such weird AI cheat will always be exploitable. Teach the AI patience and Sun Tzu instead. Cheese is never going to solve the AI's lack of intelligence, it just makes the game less fair. Chokepoints should have a tactical trade-off. Maybe have the AI construct defense towers with free threat near high value buildings. And every 5000 ships it goes up a level. Up to 50 levels.

Obviously, I am would always for a proper gameplay solution to problems rather than cheese.

[Hearteater]

If they are going to siege us out when we choke them, they need to build "siege weapons".  Maybe if they identify a choke, after an hour they start building a Warp Jump Gate that lets them skip the choke pointed work.  Specially, it is a normal Worm Hole that just leads to one of the systems connected to your choke point.  Threat fleets, waves, Exos, whatever can use it.  So ok, maybe you make your choke point 2 systems deep.  Too bad for you, that's just the Mark I Warp Jump Gate.  They upgrade it every so often.  At Mark II it skips 2 worlds.  By Mark V I'm pretty sure nothing but a snake map will be able to stop it.  The purpose is to make us push out past their blockade and kill the thing.

Also, is there any way to make Beach heads workable?  Maybe make them mobile but disable supply only in an area around them (think the size of a Gravity Drain fleet ship).  That means multiple Beach Heads are meaningful and they aren't so powerful that no one uses them.  Oh, and rename of course (*cough* Nullifier *cough*).

[LordSloth]

I am mostly content with the nature of chokepoints in the current game, at around difficulty 7. I can avoid setting up any chokepoint at the beginning of a vanilla game, if I so desire and avoid warp gate raiding. It obviously gets more difficult later on, but I've been able to do various things on a simple map type with the saved AIP from warp gate raiding, riot starships, spacetime manipulators, spider turrets, etc.

The concern I have comes to co-op. Pretty much any large (5+) co-op game resorts to setting up a single whipping boy. In fact, the ability for any of our games to get to the thirty minute mark without being horribly crippled by AIP revolves around the leadership coordination.

The coordination isn't a problem, though it can be tiresome occasionally. What is an issue is that a whipping boy is so essential and that it is so inflexible. It NEVER moves, and people invest ridiculous amounts of resources into setting it up. I'd like it if units that leave remains (turrets, fortresses?) returned just as much resources from scrapping as it costs to rebuild them. Apparently, the 10% rule has been in place since at least 3.0, so I can't find the specific reasons for that number. I feel it might be about ion cannons or avoiding 'banking up' for big projects- or forcing a window of vulnerability when you do move chokepoints, reasonably enough.

The importance of a whipping boy on even a 7/7 game is a bit troubling in the big co-op games (either no warp gate raids till the fifteen minute mark, or everybody but a couple forgetful people warp gate raid, racking up AIP, and stuff comes into a homeworld anyways), but that's more a matter of teamwork than a game design issue. I'd like to see viable alternate strats in an 8 player game, but I'm not sure if that is reasonable despite/because of teamwork or not. Certain map styles really do help with that.

I'm not convinced any changes to chokepoints with regards to co-op need to be made or even should be made, any thoughts on that?

 I would love some incentives to set up temporary whipping boys, such as better scrapping returns as I'm a bit tired of knowing that any choice I make with regards to that in a big game is permanent - but that's partially a matter of personality, fortresses, and zenith traders.

[eRe4s3r]

If they are going to siege us out when we choke them, they need to build "siege weapons".  Maybe if they identify a choke, after an hour they start building a Warp Jump Gate that lets them skip the choke pointed work.  Specially, it is a normal Worm Hole that just leads to one of the systems connected to your choke point.  Threat fleets, waves, Exos, whatever can use it.  So ok, maybe you make your choke point 2 systems deep.  Too bad for you, that's just the Mark I Warp Jump Gate.  They upgrade it every so often.  At Mark II it skips 2 worlds.  By Mark V I'm pretty sure nothing but a snake map will be able to stop it.  The purpose is to make us push out past their blockade and kill the thing.

Yeah, or something along these lines. I would even prefer that solution. It would have to be somewhat close to our territory, not on their homeworlds though ;p But aside that, sure. As long as it doesn't cheese past defenses without any way to stop it I could live with that.

I think such upgradable threat dependent events are the best solution by far ^^

Speaking of events... I also think more event based quests should pop up. Like an elite attack squadron when you kill 10000 AI ships or something. More dynamic events would do the game good.

[chemical_art]

I dont feel like ai weapons thatst be built will really change chokepoint importance. in practice chokes would still be too good it only means i have the building.

[keith.lamothe]

Chokepoints are one of the 2 things that AI war is designed around. You can't tell me it isn't. 1 to 1 connections are by design in this game, if you wanted us not to have and completely abuse chokepoints, you should not have made jumping the way it is now ^^

I'm not worried about chokepoints being too effective.  I'm not even all that concerned that chokepointing is basically required in some of the harder scenarios (FS, notably, but generally with any exo source and sometimes not even that). 

I'm concerned that the current balance of "offensive options available to the AI" and "defensive options available to the player" result in the player either outclassing the AI overall (on lower difficulties) or being unable to defend more than a single chokepointed-region (on higher diffiiculties), and thus that the rest of the galaxy basically has to be a no-mans-land where the player cannot be expected to hold anything in or the player will lose.

That cuts pretty heavily into an interesting part of the design space, and I don't think that's good for the game.

But it's possible that this perception of mine is flawed.  I'm certainly not dead-set (or even moderately-set) on changing anything here, but when I say "hey, guys, let's make crystal into a more unique resource where you have to take-and-hold to get more than a basic income of it" and the response from a substantial portion of the player-base is some variation on a central theme of "No! Take-and-hold is impossible!" ... well, I think that indicates a problem.  Not one that could be solved by just ramming through the crystal change anyway, and not one that could be solved by nerfing chokepoints into oblivion or whatever.  But something worth finding a solution to.

Thinking about it I find that I am absolutely not agreeing to local turret caps either way, that is the worst idea so far ;P Might as well entirely remove turrets then unless the cap is well above 50 of each per planet.

I think your rationality levels are dangerously low   But I do take the point, as has come up every time we've discussed this in the forums, that a substantial portion of the player-base does not want per-planet turret caps.  Generally speaking, that's good enough for me.

70% of my fun is creating the bestest defense chokepoint ever.

And you're not alone in that.  There's something deeply satisfying about making a 100-cubic-meter volume in which absolutely anything dies

Any such weird AI cheat will always be exploitable. Teach the AI patience and Sun Tzu instead.

One of the most common complaints for a while was that AI threat would not commit to attacking a chokepoint until it thought it had a massively overwhelming force.  I replied that it was smart for it to wait rather than just throw its units in the chipper shredder, but most people didn't seem satisfied by this

Cheese is never going to solve the AI's lack of intelligence, it just makes the game less fair. Chokepoints should have a tactical trade-off. Maybe have the AI construct defense towers with free threat near high value buildings. And every 5000 ships it goes up a level. Up to 50 levels.

The Hunter/Killer basically began as this: if the AI detected that you were hammering its attacks on one specific planet (a chokepoint) it would create an H/K and send it after said chokepoint.  Now, the thing could be killed but generally speaking it would just kill you.  People didn't like this.  Why would they like it if I made a more complicated version of the same?

Of course, "solution X is unsuitable" in no way proves that "solution Y is suitable", despite that assumption's common place in public rhetoric.


Perhaps the next step, as I think we found previously in this discussion (around and around we go...), is to resuscitate the Beachhead AI Plot as something not as auto-death, and see if the people who want the AI to kill chokepoints like playing with that option. 

Of course, there's already the resuscitated Train plot which is also pretty anti-chokepoint, though perhaps that's just as hard on distributed-defense, and there my understanding is that people basically went from "I never play with it because it doesn't do anything" to "I never play with it because it always tears me into tiny pieces".  This last version made the nastiest parts of it easier to handle, but I don't know if it's enough for people to step into that end of the pool again



One quick things comes to mind, though: what if the logic that increases the time between waves when the AI detects a small number of ingress points were extended such that if it only sees one or two entry points it actually diverts a part of its wave strength directly into threatfleet, and on higher difficulties specifically picks some of the units to be nasty siege breakers (EMP guardians, high-mark plasma siege starships, etc).  That way over time it would pile up a pretty substantial threatfleet that wouldn't commit against you unless there was something it thought it could kill.

Though that doesn't really deal with the take-and-hold problem, it's just another thing for the AI to get mad at your chokepoints with

To actually solve the take-and-hold problem probably the best approach is to relax what it means to "take and hold", like with the rebuildable-after-an-hour (and when you temporarily lose it the AI gets some benefit to hit you harder) thing.  Though I'm getting a lot of negative feedback on that too.  Not that negative feedback means "never do it", but probably best to try to refine the idea first.

[orzelek]

From what I understand keith's approach we would like to broaden current game options.


So per planet turret cap mechanics would still need to have option (via control buildings or something similar) to allow construction of choke point as it is right now. Whole mechanics would mean to expand on options available for players not remove existing ones.

For me current approach means that there are no meaningful alternatives - you choke it or die. If there is only one winning strategy and any alternative is clearly not up to the task... I don't see why I would play if I know I need to do exactly that one strategy from start and I can't experiment. Only game I actually finished was FS in which I had 3(or 4) defense lines. It's quite simple why it worked - it was while FS was still in beta. There were some epic fights there with in depth defense to whittle down exos - something that has been effectively removed from game after FS got finished.

As for the "I don't want the feature even if it doesn't affect my playing" mentality - it still surprises me and I don't understand it. And I'm not planning to try - everyone has their own opinion and we discuss it here to decide if changes are worth developer time spent on them.

[Diazo]

Actually, how do minor factions scale?

I'm specifically thinking FS and golem exo-waves here, but are they tied to the AI's difficulty at all?

Part of what I think happened is that exo-waves were originally introduced and were quite a challenge at the time. However, as players figured them out they got buffed to still represent a challenge to those with experience in that minor faction.

This leads to a harder introduction to the minor faction to someone coming along later, but as they can ask for advice that is not a huge problem in my eyes.

Then the scaling to the minor factions got added and the current behavior at the time was added at a rather low 4 of 10 instead of 6 or 7 of 10.

I'm wondering if the minor faction scaling is too steep, I don't think I've heard of anyone finishing a game with a minor faction that makes the game harder set in the 8-10 range.

The only one I tried personally was the new Astrotrains when their new form first came out (so before their nerfs in the last patch) and I'd set it at 8 of 10. I got roflstomped at about the 40 minute mark easily. Even with the nerfs, trains 8 is going to be hard, I'm not sure 10 is even possible. (I know it technically is, it would just require a lot of cheese.)

This is drifting onto yet another topic, but of the 'penalty' minor factions, specifically golems-hard, spirecraft-hard and fallen spire as they spawn the exo-waves, what's the highest minor faction intensity (and AI difficulty) you have won a game with those enabled? (Or know of someone winning?)

I'm thinking the high end can be toned down to allow a smaller difficulty jump per intensity increment. I'm playing a game with Fallen Spire 2 of 10 at the moment and it's really easy as compared to fallen spire 4 of 10 my last attempt was.

D.

[LordSloth]

However, I don't have enough hands on experience with >2 chokepoint or no chokepoint style strategies to really give an informed opinion on their current balance.

Personally, I don't have enough latergame experience with multiple or no chokepoints. I've done a few early game builds with no chokes - I'm going to try seeing how far I can take that with a few of the easier AIs on 7/7, vanilla, no AIP.

I'm also going to try to gather from a relatively small group of people willing to try a no choke game from my AI War steam group of SA goons. It'd be impossible to get a full five-eight people willing to try that out, but I should be able to pull together two or three other people to see what happens when we skip the choke and turn off the zenith trader. It should be fun, and if it goes poorly, we're used to that.

I was contemplating an idea that wouldn't really help with the multi-choke build... an AI Siege Carrier. The AI Siege Carrier would piggyback on a wave, but start out empty like a Carrier Guardian. Unlike a regular carrier, the Siege Carrier would hang around the wormhole, not trying to push past the defenses. As AI ships died on the planet, roughly a quarter of them would contribute to the Siege Carrier. If the AI calculated it was losing the fight, the siege carrier would pull back through the wormhole to AI territory, to piggyback on the next wave in addition to any new Siege Carriers. Possibly the siege carrier would have ludicrous health or a small forcefield or rapidly decaying armor in human territory, if it needed to be more durable initially.

Problem: the Siege Carrier would be handled the same way as any escaped threat, except less exciting- you send something after it, but it wouldn't be moving around and attacking you from different angles, or using any of the interesting abilities. However, due to it's inability to head deeper into player territory, you could safely give it immunity to force fields. In other words, the AI could retreat a portion of it's strength from your deathtrap, effectively adding onto later waves unless you go leave the comfort of your little stronghold.

Anyways, lunch is about to be over, so I can only commit to the horrible sacrifice of playing more AI War and trying something different. Woe is me.

P.S. I can certainly say whipping boy is an extremely prominent strategy for big co-op games. I'm not sure how viable a no/low chokepoint strategy is when you have eight players and giving everybody a second planet's worth of income is an additional 120 AIP. The additional resource points certainly help, but I'd like to see some shared resource gain from new command stations in co-op games if possible, for more of a resource curve rather than stepladder.

[Hearteater]

I'd take rebuilding capturable objectives if we got +1 AIP each time one of our CS died .

Otherwise I feel rebuildable objectives trivialize them completely.  I can always just wait an hour and watch a movie.  Unless I have AIP/hour on (and most people don't play with that) there is no penalty.  Any given CPA or Exo that comes my way isn't going to kill me.  I feel this just ends up being a game-delaying change.

Some people like to "collect" things and don't want to lose any of there "collection".  Which is why they don't want to lose the AF.  But I'll say this: I do not care about the AF or any other capturable.  They are nice perks, but in no way do I need any of them to win.  If you are one of the people who hates that you can lose an AF, try this: don't build any Mark IV ships or units from Fabs and win the game.  You'll find it isn't as hard as you think, especially when you keep this in mind when spending K.

[Diazo]

On 7/7, no chokepoint is quite doable, it just requires changing your unlock ratio of mobile ships to static defenses so that you unlock more mobile ships then you usually would, and then as you get into late game the Mk I stuff stays home for defense as it can move around.

It also requires paying more attention to your defenses, you can't ignore them like you can a chokepoint world, you need to move ships around to intercept as the AI waves arrive.

However, offense should also be easier. With more connections it is easier to get around so you'll find the AI homeworlds fewer hops from your own and you'll have multiple ways to attack them so you can go around the really nasty Mk IV worlds that are in your way.

It is my setup and I like it for the multiple offense options it gives me, just don't enable exo-waves, those will just walk all over you without the chokepoint to stop them.

D.

[_K_]

Heh, havent visited the forums in a while.

a) Convert all very-important capturables (AdvFact, ASC, Fabs... probably leave ZPGs and Ions and such as-is) to use this approach I mentioned in the other thread:
"Have the unit be invincible, but it shuts down for an hour after a human command station dies on the same planet.  So it's not fully gone, but you lose use of it for a time long enough (pending tuning) that it hurts bad enough that you genuinely care about not losing the planet."

Please, no. Mortality is one of the core traits of all capturables. Once its gone, its gone forever.  Loss of a capturable is a very significant event and there should be no way to completely undo the damage. That's a very important part of the game progression.

AdvFacts are the only ones that are so valuable that it might be a ok to make them rebuildable at some great permanent cost (AIP, i guess?). And even for those, we have MKIV enclaves as a substitute, for fleetships at least, so i dunno.

b) Make most turrets per-planet-cap.

Unelegant. but it might work, especially if energy generation becomes much more restrictive for turret building.

c) Whenever you lose a system with such a capturable, the AI gets a bit of "blood in the water" temporary boost to its aggressiveness.  Like throwing an extra wave at you or adding 10% or whatever to the next 30 minutes' worth of waves, etc.

That will result in higher snowballing, where you cant recover after a significant (but not yet fatal) loss. Struggling for survival after 50% of your territory is annihilated due to an Exo/CPA is fun and a great challenge.

One quick things comes to mind, though: what if the logic that increases the time between waves when the AI detects a small number of ingress points were extended such that if it only sees one or two entry points it actually diverts a part of its wave strength directly into threatfleet, and on higher difficulties specifically picks some of the units to be nasty siege breakers (EMP guardians, high-mark plasma siege starships, etc).

Eh, beat me to it. EMP guardians are a huge pain to deal with when defending. They are like mobile mini-beachheads.
I dont know how threatfleet works, though, is it able to travel between systems? If yes, why doesnt the AI send it against weaker systems that are wave-safe?

Also, how is the AI's current turret power evaluation? It used to underestimate them alot, IIRC, is that still true?

And if we can increase the threat build up with that wave leech, how about also decreasing the CPA timer for each such wave?

70% of my fun is creating the bestest defense chokepoint ever.

Yeah, but how about a multi-layered distributed system, where losing one point just means retreating to the next one and making a stand there, again and again, until you finally make a last stand at the Home Command station? That's the part of defence i love the most. Defence of Helm's Deep doesnt feel epic if the Orcs dont make it past the outer wall.


I havent played alot recently, so how long does it take to rebuild a completely annihilated chokepoint fortress system? Just wondering, because if its less than 10-12 minutes, then this whole thing probably doesnt have much of a point, unless a very large chunk of the enemy fleet survives. The player resource gains have been buffed too much, especially for games with golems and asteroids off. The time to rebuild is an important game aspect and decreasing it means a much easier way out of bad situations.

[keith.lamothe]

For me current approach means that there are no meaningful alternatives - you choke it or die. If there is only one winning strategy and any alternative is clearly not up to the task... I don't see why I would play if I know I need to do exactly that one strategy from start and I can't experiment.

I don't think it's nearly that bad.  Plenty of the AARs I've seen, even on higher difficulties, don't rely on a single choke or even (in some cases) any real chokepointing at all.  Granted, those aren't heavy-exo games, but not all games have those.

Only game I actually finished was FS in which I had 3(or 4) defense lines. It's quite simple why it worked - it was while FS was still in beta. There were some epic fights there with in depth defense to whittle down exos - something that has been effectively removed from game after FS got finished.

How was it removed from the game?  I remember doing multi-layer exo defense on two fronts around that time and it worked fine (well, "fine" as in "I sometimes almost died", but that's kind of the point).

Actually, how do minor factions scale?

I'm specifically thinking FS and golem exo-waves here, but are they tied to the AI's difficulty at all?

Exos scale with difficulty, faction intensity, and human homeworld count.  The FS ones (the preannounced ones) also scale with how many city structures you have, and the Golem/Spirecraft ones scale with AIP.

Exos on 10 are supposed to kill you; they're like 5x what they would be on diff 7.  Diff 9 is a lot less than that but still harsh.  Diff 8 less so.


To some extent I think people who like to play on as-high-a-difficulty-as-they-can will be pushed heavily towards choke defense, since it will always (aside from some very strange new mechanics) be at least somewhat better than non-choke.  On Diff 7.6 or whatever you have a lot more options... but on that difficulty a lot of you guys just kind of roll over it absent-mindedly, so it doesn't really matter

I'd take rebuilding capturable objectives if we got +1 AIP each time one of our CS died .

They'd lynch me!   More seriously, I'd like to try that as part of an AI plot.  The "the game is really being a bastard now" plot