Arcen Games

Games => AI War II - Gameplay Ideas => AI War II => AI War II - Resolved Ideas => Topic started by: kasnavada on September 12, 2016, 02:25:05 AM

Title: Review of AIP progress
Post by: kasnavada on September 12, 2016, 02:25:05 AM
Currently, the way AI progress is presented in game is very confusing.

First "mention" of it is in the game menu.
Second mention of it in game are the cryogenic pods & home human settlement. Which give +1 and +5.*
Then, there is the tooltip (at the top) which state the thresholds and current level. Threshold that go up to 1120 AI progress at diff 8, just to give an example.
Last, there are stuff in the AI planets that give AI progress. Most confusing for me is the command station, that states that it gives 15, but in practice gives 20 because the gate explodes automatically.

My opinion is that this presentation gives very wrong ideas to new players, like "oh, this can go up to more than 1000 ? Let's go nuts". And, invariably, they're going to screw up their first games because they don't get that AIP shouldn't go over 200. If not less at higher difficulty level. Then again, I think it's not written not conveyed anywhere. The second issue is that AIP progress consequences are not immediate. It takes a while for AI progress to have an effect, as waves, and other AI mechanics, kick in. That makes it even confusing to me.

What I'd prefer is a system where AIP progress ain't a "progress" that seems like it can go infinitely up. Rather... a count-down, to your immediate doom. With the new mechanics in AI War 2 anyway, I'm not sure that the AI progress possibilities as were in the first game will be able to apply (like the people that challenged themselves to conquer the entire galaxy, and the "low-AIP strategy" and so on).

So... what if, for example, AIP progress, in game, was made a more tangible timer ?
AIP progress would be at 300 at start (depending on diff level), and goes down from there instead of up. Once it reaches 0, the AI sends a large wave of whatever secret weapon it was building, and it's game over.

PS: I realize that it removes some playstyles.
Title: Re: Review of AIP progress
Post by: ewokonfire on September 12, 2016, 03:04:32 AM
I don't like it.  Removing playstyles has to be a bad thing, and I think it's too much of a tradeoff for a small benefit to the initial learning curve.  Seeing as it seems like some kind of tutorial will exist, it doesn't seem like a bad thing to just tell the player that they should aim for an AIP of around 200 by the end of a normal game (mention that advanced strategies can go higher or lower than that, but make it clear that 150-200 is normal for a first game).  The tech thresholds need reworking or removing anyway, as they were generally completely redundant.
Title: Re: Review of AIP progress
Post by: Cinth on September 12, 2016, 03:11:21 AM
In Classic, AIP has a fairly high hard cap.  Using the very start as a marker for that level.  The only way to reach that cap is to either kill all AIP structures (you can go nuts with warheads but whatever).  AIP in reality has a very low soft cap in comparison.  That soft cap leads to stalemate or loosing conditions.

Strategy revolving around AIP range from the super cheesy ultra-low AIP to conquer the galaxy.

And, invariably, they're going to screw up their first games because they don't get that AIP shouldn't go over 200.
1: Who says AIP has to stay under 200  :P
2: There is this thing called a learning curve.  If you go out and win your first game on a decent difficulty, did you learn anything or really have fun?

The second issue is that AIP progress consequences are not immediate.
AIP sets thresholds.  It shouldn't immediately cause anything to happen.  The AI is already doing stuff and AIP tells it how much and what types.

PS: I realize that it removes some playstyles.
Why do you continue to go there?
Title: Re: Review of AIP progress
Post by: kasnavada on September 12, 2016, 03:50:23 AM
In Classic, AIP has a fairly high hard cap.  Using the very start as a marker for that level.  The only way to reach that cap is to either kill all AIP structures (you can go nuts with warheads but whatever).  AIP in reality has a very low soft cap in comparison.  That soft cap leads to stalemate or loosing conditions.

Strategy revolving around AIP range from the super cheesy ultra-low AIP to conquer the galaxy.

That's about the issue I have. Once the player realize that the ultra-low AIP exists, which the game does not tell you explicitely, there is no challenge in the game. Only grinding time.

And, invariably, they're going to screw up their first games because they don't get that AIP shouldn't go over 200.
1: Who says AIP has to stay under 200  :P
2: There is this thing called a learning curve.  If you go out and win your first game on a decent difficulty, did you learn anything or really have fun?
[/quote]

1) the game at more than 9 difficulty ? Actually, at any difficulty level ? The "Ultra-low" AIP route is basically the safest way to win the game. But, it's annoyingly boring, and based on waiting, more waiting, even more waiting. Keeping AIP low neutralizes the AI nearly completely.

After that, some people like setting themselves challenges - but are those a majority ? A minority ?

2) Yes, I expect to win at a decent difficulty level, and have fun discovering the game itself. Yes, you can have a learning curve while winning, and, actually have fun, and learn stuff. I'm greedy, like that. In my particular case, I've played games for 25 years, mostly strategy ones, so if I actually lose at that level, it's because the game deliberatly held back information on me - and that does not makes a good impression. That's what AI war actually did, by the way.

No, I don't start games at "lower" difficulty level than "normal", because I think that toned-down games teach you bad strategies, which you'll have to forget when playing higher difficulties.

The second issue is that AIP progress consequences are not immediate.
AIP sets thresholds.  It shouldn't immediately cause anything to happen.  The AI is already doing stuff and AIP tells it how much and what types.

Errr, yes. Not my point. It's ok if it's delayed. I put a paragraph there, around that sentence, because I tried to convey a general idea. I'm wondering for options about making it less confusing.

PS: I realize that it removes some playstyles.
Why do you continue to go there?

Because I consider really stupid the opposite opinion. Not changing, not challenging everything to see what can be improved. Because I care about making AI War 2 a better game, and I do that this way. Trying to cut down stuff, remove stuff, see what people really like about this, or dislike about this or that feature, and / or add stuff unrelated, out of the box if possible. That's what made me successful at work, in my private life, by reviewing, critisizing everything about what I am (because I start with myself), and others, and things around me, whatever there is around to be questioned. I don't care for convention or following the leader(s), and question, if possible, everything I can, to keep an open mind.

If you cannot understand this, from all the answers and posts I made, and answers I made, they I wonder why you keep answering me. I'll always go where people don't want to. That's what I do. I don't care at all if it displeases people nor if it does not work. I care about thinking out of the box, and making others do the same.


PS: I'm actually not of the opinion that this will be accepted, yet I think it'll make a nice brainstorm. No need to attack "me" directly. I'll continue to do those kind of posts anyway.
Title: Re: Review of AIP progress
Post by: kasnavada on September 12, 2016, 03:51:00 AM
I don't like it.  Removing playstyles has to be a bad thing, and I think it's too much of a tradeoff for a small benefit to the initial learning curve.  Seeing as it seems like some kind of tutorial will exist, it doesn't seem like a bad thing to just tell the player that they should aim for an AIP of around 200 by the end of a normal game (mention that advanced strategies can go higher or lower than that, but make it clear that 150-200 is normal for a first game).  The tech thresholds need reworking or removing anyway, as they were generally completely redundant.

Good points.
Title: Re: Review of AIP progress
Post by: Cinth on September 12, 2016, 05:30:20 AM
That's about the issue I have. Once the player realize that the ultra-low AIP exists, which the game does not tell you explicitely, there is no challenge in the game. Only grinding time.

What good strategy game tells you how to cheese it?  And if you want to boil it down, is all grinding time, but most of us prefer to call it completing objectives.

1) the game at more than 9 difficulty ? Actually, at any difficulty level ? The "Ultra-low" AIP route is basically the safest way to win the game. But, it's annoyingly boring, and based on waiting, more waiting, even more waiting. Keeping AIP low neutralizes the AI nearly completely.

After that, some people like setting themselves challenges - but are those a majority ? A minority ?
Sounds like  you should play differently.  Play the game the way you enjoy it, not the way that's always easiest.

2) Yes, I expect to win at a decent difficulty level, and have fun discovering the game itself. Yes, you can have a learning curve while winning, and, actually have fun, and learn stuff. I'm greedy, like that. In my particular case, I've played games for 25 years, mostly strategy ones, so if I actually lose at that level, it's because the game deliberately held back information on me - and that does not makes a good impression. That's what AI war actually did, by the way.

No, I don't start games at "lower" difficulty level than "normal", because I think that toned-down games teach you bad strategies, which you'll have to forget when playing higher difficulties.
What didn't the game tell you?  How to easily beat it on 10/10?  AIP isn't a mystery and you should have it figured out relatively soon. 

Decent difficulty would be 7.  That's the full AI experience without it gaining extras.  Also, I would say there is a lot you can learn in 1/1 sandbox games that apply even in 10/10 games.

Errr, yes. Not my point. It's ok if it's delayed. I put a paragraph there, around that sentence, because I tried to convey a general idea. I'm wondering for options about making it less confusing.
It reads like you plain out don't understand how AIP works.

Because I consider really stupid the opposite opinion. Not changing, not challenging everything to see what can be improved. Because I care about making AI War 2 a better game, and I do that this way. Trying to cut down stuff, remove stuff, see what people really like about this, or dislike about this or that feature, and / or add stuff unrelated, out of the box if possible. That's what made me successful at work, in my private life, by reviewing, critisizing everything about what I am (because I start with myself), and others, and things around me, whatever there is around to be questioned. I don't care for convention or following the leader(s), and question, if possible, everything I can, to keep an open mind.

If you cannot understand this, from all the answers and posts I made, and answers I made, they I wonder why you keep answering me. I'll always go where people don't want to. That's what I do. I don't care at all if it displeases people nor if it does not work. I care about thinking out of the box, and making others do the same.


PS: I'm actually not of the opinion that this will be accepted, yet I think it'll make a nice brainstorm. No need to attack "me" directly. I'll continue to do those kind of posts anyway.
That's all fine and good, but don't act all surprised when you rub people the wrong way, repeatedly.
Title: Re: Review of AIP progress
Post by: kasnavada on September 12, 2016, 05:46:15 AM
Quote
(...) That's all fine and good, but don't act all surprised when you rub people the wrong way, repeatedly.

Errr... I don't stalk people. If I repeatedly rubbed you the wrong way, I apologize, but I really don't give a damn about who I'm answering to, as a whole. That said, you've been stalking me for a few posts, and, I got no clue what's wrong with you attacking me like that. With more and more personal attacks rather than being anywhere near being on subject. Just like that last post here, or the other one in the kickstarter thread. And, whatever issue you've got with me, private messages are here for a reason - no need to make it public.

(for reference) https://www.arcengames.com/forums/index.php/topic,19060.0.html

About your last post... I'm not answering it.
Title: Re: Review of AIP progress
Post by: Cinth on September 12, 2016, 06:05:07 AM
I found this suggestion absurd in its entirety and I feel like you're pointing the finger at AIP when there is something else that could be worked on.  AIP is core to AIWC.  It's part of what makes the game.  Changing that is taking away form the experience, not bettering it.  You not seeing that makes me wonder. 

Sorry if you feel it's a personal attack.  You posted something that I felt the need to respond to.  You and I happen to be diametrically opposed on a lot of topics you've posted about.
Title: Re: Review of AIP progress
Post by: kasnavada on September 12, 2016, 07:05:50 AM
I found this suggestion absurd in its entirety and I feel like you're pointing the finger at AIP when there is something else that could be worked on.  AIP is core to AIWC.  It's part of what makes the game.  Changing that is taking away form the experience, not bettering it.  You not seeing that makes me wonder. 

Sorry if you feel it's a personal attack.  You posted something that I felt the need to respond to.  You and I happen to be diametrically opposed on a lot of topics you've posted about.

Like, for real ? You feel the "need" to answer stuff like this, with the moderator badge attached to your name ?

What didn't the game tell you?  How to easily beat it on 10/10?  AIP isn't a mystery and you should have it figured out relatively soon.
Decent difficulty would be 7.  That's the full AI experience without it gaining extras.  Also, I would say there is a lot you can learn in 1/1 sandbox games that apply even in 10/10 games.

Never said anything about 10/10, nor about easy. I never stated that AIP is a mystery. I don't need to be insulted by stating that I don't know how AIP work, nor that I don't know that 7 would be a decent activity, nor that I could not figure it out.

I don't need to be insulted by stating that what I propose is "absurd".

Where is the matter of different opinions here ? You're just bashing what I'm not even saying by using strawman argumentation, then resorting to personal attacks. I'd go on about the rest of the post, but it's really useless, as it's more of the same.
Title: Re: Review of AIP progress
Post by: Cinth on September 12, 2016, 07:34:07 AM
Like, for real ? You feel the "need" to answer stuff like this, with the moderator badge attached to your name ?

What does me being a mod have to do with anything? 

I don't like your idea (or any idea that would destroy different play styles for that matter).  You propose a complete change in a core mechanic (something previously discussed at length on the forum here).  Not only is it a complete change, but you don't demonstrate a good understanding of the mechanic you want to change.

And yeah, this marks about the 4th time we've had words about something on the forum.  Usually about something I feel opposed to. 


Funny reporting my post btw.  Real funny. 
Title: Re: Review of AIP progress
Post by: Tridus on September 12, 2016, 08:15:03 AM
So... what if, for example, AIP progress, in game, was made a more tangible timer ?
AIP progress would be at 300 at start (depending on diff level), and goes down from there instead of up. Once it reaches 0, the AI sends a large wave of whatever secret weapon it was building, and it's game over.

In some of my FS games, I'd have AIP of around -500 in this proposal. That would probably be even more confusing. :)

Quote
PS: I realize that it removes some playstyles.

The problem with this is that it assumes low AIP is the only way to play the game. It may very well be, at high difficulty with no superweapons. Tons of people don't play high difficulty with no superweapons, and this doesn't work for them at all. If you're playing at 5/5, 300 AIP is not a crippling problem even without superweapons.


AIP can be presented and explained better in AIW2, to help ease any issues. At it's core, it works petty well as-is though. One thing I would change is making the tech level steps less cliff-like and more gradual. Sadly, we lost a great opportunity to do that when the newly proposed tech system got scaled back, but I think there's still some room to make it cleaner.
Title: Re: Review of AIP progress
Post by: kasnavada on September 12, 2016, 08:24:38 AM
So... what if, for example, AIP progress, in game, was made a more tangible timer ?
AIP progress would be at 300 at start (depending on diff level), and goes down from there instead of up. Once it reaches 0, the AI sends a large wave of whatever secret weapon it was building, and it's game over.

In some of my FS games, I'd have AIP of around -500 in this proposal. That would probably be even more confusing. :)

The idea there was also that there are possibly too many different playstyles to balance. I mean, I think that most people don't... bend... the gameplay to resist to that high an amount of stuff, as what I saw possible on some threads around here, especially those bent to capture 100+ planets (and winning). That it's even possible in the first place raise interrogations from me, because if that high difficulty games are viable, what does that mean of lower level games ? I'm fine with some difficulty increase from "normal difficulties" to "impossible ones", but I fear the gap could be a bit much.

The problem with this is that it assumes low AIP is the only way to play the game. It may very well be, at high difficulty with no superweapons. Tons of people don't play high difficulty with no superweapons, and this doesn't work for them at all. If you're playing at 5/5, 300 AIP is not a crippling problem even without superweapons.

Hum, yes. I realized that. About 5/5 games, I think I proposed the AI "progress" to be higher for lower diff, so it takes care of that.
That the AIP works differently depending on those options (not diff, but whether golems are in the game, for example) is one of the things I count as an issue. It must be a nightmare to balance properly, especially if all playstyles have to be taken into account.
Title: Re: Review of AIP progress
Post by: kasnavada on September 12, 2016, 08:31:44 AM
Removing playstyles has to be a bad thing, and I think it's too much of a tradeoff for a small benefit to the initial learning curve.

I thought again about your post, and about the first part... I'm not actually sure that more playstyles are better, by default. That one can set himself a challenge is something that's not really possible to remove anyway, but the thing is that one of those playstyles is mechanically "better" than others, at accomplishing one goal (generally, winning, or going as long as possible at least).

The main issue I have with that is that some of playstyles mean that "cheese" must be allowed to be kept in the game. I don't like this part much. And, AI War 2 changes a lot of mechanics, so some playstyles are going to die. Which do you think really makes AI War what it is ?
Title: Re: Review of AIP progress
Post by: Tridus on September 12, 2016, 08:48:21 AM
The idea there was also that there are possibly too many different playstyles to balance. I mean, I think that most people don't... bend... the gameplay to resist to that high an amount of stuff, as what I saw possible on some threads around here, especially those bent to capture 100+ planets (and winning). That it's even possible in the first place raise interrogations from me, because if that high difficulty games are viable, what does that mean of lower level games ? I'm fine with some difficulty increase from "normal difficulties" to "impossible ones", but I fear the gap could be a bit much.

Not only is it possible, there's Steam achievements for controlling 50/80/110 planets. I don't expect people are doing it a whole lot on 10/10 though. ;)

It's possible on low difficulties, and it's also possible with superweapons. A Spire City on a chokepoint is a very powerful defense. Throw in some core turrets and some trader goodies, and look out. (You can get really cheesy with it, too. One time I took a central point on an X map, put a city there, and then conquered *everything* behind it. That gave the AI only a single place it could possibly attack me, and some absurd AIP number. There was much CPU choking on the outcome of that.)

On a more open map it's still doable, but the defense is much more difficult to maintain. It seems to work okay, balance wise. Certainly better than just imposing a hard cap would, since superweapon options would have to change that cap due to how they change the game. Simply letting AIP continually go up works more easily.

Quote
Hum, yes. I realized that. About 5/5 games, I think I proposed the AI "progress" to be higher for lower diff, so it takes care of that.
That the AIP works differently depending on those options (not diff, but whether golems are in the game, for example) is one of the things I count as an issue. It must be a nightmare to balance properly, especially if all playstyles have to be taken into account.

It's not made any easier by having to come up with new AIP limits for every difficulty and every combination of options ,though. That's making it harder. It happens organically right now: push the number up higher, and the AI pushes back with more force.

I'm also not concerned about balance a whole lot, outside of the core gameplay experience. This isn't an eSport competitive game, it's single player/coop. Having options that totally rewrite the balance is fine. Having stuff that's flat out crazy is fine. There is no case where the current Botnet Golem on easy is balanced. Doesn't matter, because it's crazy fun to use. (People who want the "balanced" version can use the medium or hard options instead.)
Title: Re: Review of AIP progress
Post by: kasnavada on September 12, 2016, 08:59:12 AM
I'm also not concerned about balance a whole lot, outside of the core gameplay experience. This isn't an eSport competitive game, it's single player/coop. Having options that totally rewrite the balance is fine. Having stuff that's flat out crazy is fine. There is no case where the current Botnet Golem on easy is balanced. Doesn't matter, because it's crazy fun to use. (People who want the "balanced" version can use the medium or hard options instead.)

Nothing to reply to the rest of your post, but that part, I do.

I'm fine with some "let's have fun games" from time to time, but that's easily done via reducing difficulty.

To me a balanced game is one where whatever "opponent" you have matches your strength, and opposes you enough so that it pushes you out of your confort zone, but close enough so you'll expand your confort zone by what was learned in that game. Basically, if at all possible, I'd hope that it's possible, and easy, to predict when a game will be in that zone when activating options. That said, everyone has different expectations and "learning" possibilities.

Because, to me, it's disappointing when you get so crushed that nothing can be learned from your failure, or if you're "wasting your time" because the game has been made too easy with whatever new feature you've been trying. Ok, not really "wasting", but simply not having a challenge at all. Hence the concern about balance.
Title: Re: Review of AIP progress
Post by: Pumpkin on September 12, 2016, 09:11:43 AM
The main issue I have with that is that some of playstyles mean that "cheese" must be allowed to be kept in the game. I don't like this part much. And, AI War 2 changes a lot of mechanics, so some playstyles are going to die. Which do you think really makes AI War what it is ?
Just for the sake of discussion, I'll share a very funny game I won several times ago.

It was 7/7 on a 80/wheel map. My goal was to go damn-low AIP and take only the planets I was forced to: CSGs. It was a scarce game, with extremely few metal, energy and Knowledge (nearly all my Hacking points went into knowledge stealing, IIRC), but waves and CPAs were ridiculous. DGLairs and MkIV Subcommanders were nightmares, on the other hand. In the end, I assaulted the first homeworld while sitting on the AIP floor. I save-scummed a bit but I pulled the victory out.

Cool story bro. Why do I tell you that? Because different playstyles are important, and some are very odds and "cheezy". I don't want someone to step in and say "your playstyle is cheezy, we'll make it impracticable to balance the game." I also remember the Red.Queen's insane-auto-AIP and nuke-flying game. People here already talked about Fallen Spire, etc. That's also what made me understand that, despite I think the multi-homeworld start and high-cap option are harmful for the game "balance", they support a playstyle. A very different of mine (even of my "normal" playstyle), but who am I to remove that from the game?
Title: Re: Review of AIP progress
Post by: Tridus on September 12, 2016, 09:35:26 AM
I'm also not concerned about balance a whole lot, outside of the core gameplay experience. This isn't an eSport competitive game, it's single player/coop. Having options that totally rewrite the balance is fine. Having stuff that's flat out crazy is fine. There is no case where the current Botnet Golem on easy is balanced. Doesn't matter, because it's crazy fun to use. (People who want the "balanced" version can use the medium or hard options instead.)

Nothing to reply to the rest of your post, but that part, I do.

I'm fine with some "let's have fun games" from time to time, but that's easily done via reducing difficulty.

To me a balanced game is one where whatever "opponent" you have matches your strength, and opposes you enough so that it pushes you out of your confort zone, but close enough so you'll expand your confort zone by what was learned in that game. Basically, if at all possible, I'd hope that it's possible, and easy, to predict when a game will be in that zone when activating options. That said, everyone has different expectations and "learning" possibilities.

Because, to me, it's disappointing when you get so crushed that nothing can be learned from your failure, or if you're "wasting your time" because the game has been made too easy with whatever new feature you've been trying. Ok, not really "wasting", but simply not having a challenge at all. Hence the concern about balance.

Sure.

What if someone else wants to? They don't care about balance, or being challenged. They want to turn on every superweapon in the galaxy and watch 50,000 ships slug it out in an epic final confrontation. I don't see why that's any better or worse than what you want to do.

There is not one true way to play the game. You need to understand that.
Title: Re: Review of AIP progress
Post by: x4000 on September 12, 2016, 10:24:07 AM
I'm not responding to the main thread here at the moment, but to the rift between Cinth and kasnavada at the moment: https://www.arcengames.com/forums/index.php/topic,19108.0.html
Title: Re: Review of AIP progress
Post by: kasnavada on September 12, 2016, 10:51:45 AM
@Tridus & @Pumpkin.

I understand that, and just wanted to try out if making the "extreme" different playstyle that are currently in-game somewhat "closer".

The proposal I made here tried to reflect that. I had a larger post, but lost it because my computer rebooted.
Title: Re: Review of AIP progress
Post by: Tridus on September 12, 2016, 10:55:48 AM
The proposal I made here tried to reflect that. I had a larger post, but lost it because my computer rebooted.

Doh. :( Always sucks when that happens.
Title: Re: Review of AIP progress
Post by: kasnavada on September 12, 2016, 10:57:14 AM
The proposal I made here tried to reflect that. I had a larger post, but lost it because my computer rebooted.

Doh. :( Always sucks when that happens.

Thanks =).

Title: Re: Review of AIP progress
Post by: skrutsch on September 12, 2016, 11:44:04 AM
AIP can be presented and explained better in AIW2, to help ease any issues. At its core, it works pretty well as-is though.

Agree with both keeping the concept and rising structure of AIP and the importance of presenting and explaining it as best we can.

One thing I would change is making the tech level steps less cliff-like and more gradual.

For newer players, I think the cliffs are a good idea, to emphasize just how important the AIP mechanic is.  Newbies are not going to notice any subtle changes anyway.  The rest of this applies to the power players:

I understand the representation of AIP as a number, the returned value from some mysterious unknowable alien subroutine.  But must we humans always know exactly what the AIP value is and what the AI is most "worried" about?

Instead of +1 AIP every 10 minutes, what about 0, 1 or 2 AIP every 10 minutes, with an equal probability of each result?   (Or 0, 1 or -1 AIP every one minute for an unstable AI?)  The expected value of AIP remains the same, but the player's uncertainty increases a bit every time interval.  So the AIP might be 5 plus-or-minus 1 after 10 minutes, but 120 plus-or-minus 18 after three hours.  This really doesn't make the tech level steps more gradual, it just makes it easier for a player to trip on them in the dark. :)

Does this AI care more about losing ships or losing structures?  Is it more threatened by damage from player raids or large concentrations of player ships?  These kinds of things included in "AI Personalities" might give them more, um, personality.

(My mental model is from WWII board game World in Flames.  As the Japanese player, you want to Pearl Harbor the United States just before they decide to enter on their own, but as in real life you can't tell precisely when that will be, merely a range of results.)

Maybe hacking can "gather intelligence" (shown by narrowing the AIP range) or de-sensitize the AI to some particular player action?
Title: Re: Review of AIP progress
Post by: kasnavada on September 12, 2016, 11:51:07 AM
On and off topic at the same time, there are mentions in the design doc that Chris tries to "remove", or lessen, or disengage, at least, the impact of AIP on some of the AI responses. One topic about this is starting there :

https://www.arcengames.com/forums/index.php/topic,19107.0.html

An interesting idea is to tie "candy techs" to "capturing a system from the AI". Which also takes AIP, I suppose (unless it's going to change in AI war 2), but it adds a very visible "penalty" to the player, and possibly is game changing. Or not. So taking a new system is kind of a gamble, and at the same time not. This sytem is in the game called "infested planet", and works very well for all I can tell, and the praise that this game got.
Title: Re: Review of AIP progress
Post by: x4000 on September 12, 2016, 12:22:36 PM
Quick note: I was also unhappy about the giant increases in tech that AIP cause, but Keith addressed the fact that that is not stepwise at all anymore, actually.  There's more that can be done there, though.  Overall there are a variety of things that I intend to shift in the direction of more reactions to other things you do, versus just global AIP.  I see AIP as overall driving waves content/size/quality and reinforcement size and CPA size and a few other things like that.  Those are very important things, but not the whole picture; a lot of the other stuff can be decoupled to some lesser or greater degree.
Title: Re: Review of AIP progress
Post by: kasnavada on September 12, 2016, 12:31:07 PM
If you're talking about tech level threshold ? If I remember well, keith made it so the tech level are "mixed" in attacking waves, so if you're at the "middle" of the "tech tiers" tied to AIP, between mark 2 or 3 for example, you'd have 50% of each. So now the main impact is that waves get slightly bigger and slightly more "strong" and there is no gap when the step is reached.

About AIP, are minor factions going to have an AIP rating, determining how much the AI wants them dead ? That could be a cool idea.
Title: Re: Review of AIP progress
Post by: x4000 on September 12, 2016, 12:36:17 PM
Correct on the mixed waves.  On minor factions I'm not sure on that yet.  Could be a cool idea, had not thought of it before you just suggested it, and I'm not sure of the ramifications.
Title: Re: Review of AIP progress
Post by: PokerChen on September 12, 2016, 02:21:25 PM
Correct on the mixed waves.  On minor factions I'm not sure on that yet.  Could be a cool idea, had not thought of it before you just suggested it, and I'm not sure of the ramifications.
You'd need to set up a modified response curve, but it'd be a unified way to control the rate/order at which AI gobbles up particular factions. Have each faction start with randomised 0~100 AIP + Shark-AB, and give them slightly varying auto-AIP progress. The AI will then attack them in slowly increasing waves, but they should have the starting resources to fend them off for a while.

 Might be interesting to have the AI divide up its reinforcement pool against the various human factions. Gives you a visceral effect on the impending doomsday line, similar to the way funding countries leave X-COM to join the alien overlords. (Harder to balance.)
Title: Re: Review of AIP progress
Post by: kasnavada on September 12, 2016, 02:49:05 PM
Correct on the mixed waves.  On minor factions I'm not sure on that yet.  Could be a cool idea, had not thought of it before you just suggested it, and I'm not sure of the ramifications.
You'd need to set up a modified response curve, but it'd be a unified way to control the rate/order at which AI gobbles up particular factions. Have each faction start with randomised 0~100 AIP + Shark-AB, and give them slightly varying auto-AIP progress. The AI will then attack them in slowly increasing waves, but they should have the starting resources to fend them off for a while.

 Might be interesting to have the AI divide up its reinforcement pool against the various human factions. Gives you a visceral effect on the impending doomsday line, similar to the way funding countries leave X-COM to join the alien overlords. (Harder to balance.)

A lot different from what I was seeing. That might be better discussed in another thread. Could you ?
Title: Re: Review of AIP progress
Post by: Captain Jack on September 12, 2016, 02:55:37 PM
Correct on the mixed waves.  On minor factions I'm not sure on that yet.  Could be a cool idea, had not thought of it before you just suggested it, and I'm not sure of the ramifications.
You'd need to set up a modified response curve, but it'd be a unified way to control the rate/order at which AI gobbles up particular factions. Have each faction start with randomised 0~100 AIP + Shark-AB, and give them slightly varying auto-AIP progress. The AI will then attack them in slowly increasing waves, but they should have the starting resources to fend them off for a while.

 Might be interesting to have the AI divide up its reinforcement pool against the various human factions. Gives you a visceral effect on the impending doomsday line, similar to the way funding countries leave X-COM to join the alien overlords. (Harder to balance.)
Er, I might have dramatically misunderstood, but there aren't going to be human factions throughout. They're killed in the first hour and you play the rest of the game against the AI (but with the other minor factions).
Title: Re: Review of AIP progress
Post by: Tridus on September 12, 2016, 03:26:14 PM
Correct on the mixed waves.  On minor factions I'm not sure on that yet.  Could be a cool idea, had not thought of it before you just suggested it, and I'm not sure of the ramifications.
You'd need to set up a modified response curve, but it'd be a unified way to control the rate/order at which AI gobbles up particular factions. Have each faction start with randomised 0~100 AIP + Shark-AB, and give them slightly varying auto-AIP progress. The AI will then attack them in slowly increasing waves, but they should have the starting resources to fend them off for a while.

 Might be interesting to have the AI divide up its reinforcement pool against the various human factions. Gives you a visceral effect on the impending doomsday line, similar to the way funding countries leave X-COM to join the alien overlords. (Harder to balance.)
Er, I might have dramatically misunderstood, but there aren't going to be human factions throughout. They're killed in the first hour and you play the rest of the game against the AI (but with the other minor factions).

The way that I read it was that some of the bonus ships were coming from the other human factions, so you must be able to help some of them? Otherwise you'd never get those.
Title: Re: Review of AIP progress
Post by: x4000 on September 12, 2016, 03:29:29 PM
Correct on the mixed waves.  On minor factions I'm not sure on that yet.  Could be a cool idea, had not thought of it before you just suggested it, and I'm not sure of the ramifications.
You'd need to set up a modified response curve, but it'd be a unified way to control the rate/order at which AI gobbles up particular factions. Have each faction start with randomised 0~100 AIP + Shark-AB, and give them slightly varying auto-AIP progress. The AI will then attack them in slowly increasing waves, but they should have the starting resources to fend them off for a while.

 Might be interesting to have the AI divide up its reinforcement pool against the various human factions. Gives you a visceral effect on the impending doomsday line, similar to the way funding countries leave X-COM to join the alien overlords. (Harder to balance.)
Er, I might have dramatically misunderstood, but there aren't going to be human factions throughout. They're killed in the first hour and you play the rest of the game against the AI (but with the other minor factions).

The way that I read it was that some of the bonus ships were coming from the other human factions, so you must be able to help some of them? Otherwise you'd never get those.

I'm still trying to figure out some of the specifics on this in terms of what will feel ideal.  But my overall thinking at the moment is that they would start out as nicer civilizations that get obliterated early on, leaving you alone.  And then little pockets of survivors can still use your help and give you stuff (like those ships) in return for it.  Those last survivors probably wouldn't ever die completely out, but they'd be yet another point of back and forth contestation.  If you can help them grow back a little bit, that might be extra good in terms of what you can get out of them.