Author Topic: So, this whole crystal thing (wait I think I used that title already...)  (Read 23211 times)

Offline Cinth

  • Core Member Mark II
  • *****
  • Posts: 2,527
  • Resident Zombie
Also realize that everyone plays differently.  Diazo (low aip player) and I (CPU destroyer) play completely different games.

This is the difficulty that Keith has to wrestle with every time he wants to make a change to the game.  Where do I set the balancing point?

It has to be low enough to give low AIP players a challenge (170 AIP minimum here) and have some impact on high AIP players.

while I might not get much use from it (said many times why already), hacking is available to everyone.  Low AIP players do it and well as everyone up through high AIP players.  If you expand the options you should also expand the opportunities to take advantage of those new opportunities. 

In the end it comes down to what kind of challenge the player wants. 
Quote from: keith.lamothe
Opened your save. My computer wept. Switched to the ST planet and ship icons filled my screen, so I zoomed out. Game told me that it _was_ totally zoomed out. You could seriously walk from one end of the inner grav well to the other without getting your feet cold.

Offline Toranth

  • Hero Member Mark III
  • *****
  • Posts: 1,244
Knowledge-hacking generates 3.33~ antagonism per point of K raided.  So 10,000 per full planet's worth of K.  In multiplayer this is divided by the number of players in the game who use knowledge (so excluding AI and champ-only players), since everyone K raids separately there.
- 10k sounds like a typo on my part; it started at 80,000 and while that was probably too much 10k sounds abusably low (quick, go exploit before it's fixed!).  Anyway, will look at that more.
You sure about K-raiding being only 10,000?  According to the logs, it's
Code: [Select]
kRaidAntagonism = TotalAmountOfKnowledgeRaided * (10000/3000) = 9840
kRaidAntagonism *= (8/TotalHumanPlayersThatGetKnowledgeCount) = 78720
and when I've raided only 1 planet, I see "Hacking Response: Moderate" displayed.


My current thought implementation-wise for the scaling-with-AIP is to leave the raw antagonism number as-is (balancing aside) and to make "effective antagonism" (which is what would actually be used for computing responses) equal to this raw number times a multiplier determined by AIP.
Hmm, this sounds different from what I thought you were trying to achieve.
What I thought you wanted was more of a Hacking Ceiling type scale, which would be based on AIP.  When your hacking antagonism was below that ceiling, the AI response would be more muted.  When above, the response would be enhanced.  Then, the percentage displayed on screen would be the Current_Antagonism / Ceiling_Antagonism.

If you use a Hacking_Ceiling approach, the AI response multiplier could be something like 2^(Current_Antagonism - Ceiling_Antagonism).  This gives a slowly increasing response as you approach the cap, but if you surpass the cap, it starts to increase dramatically.  Increasing the cap flattens out the curve again.  A power curve (a * x^b) may be a better fit than an exponential (a * b^x).  The easiest way to fit the curve is to decide what the response multiplier should be at various ratios of the cap, and then just curve-fit.

The balancing concern is then: How does the Hacking_Ceiling go up as AIP goes up?

Linear doesn't seem too bad.  Sure, the FS player with 900 AIP could hack three times as much as the normal player with 300 total_AIP, but compared to the benefits of FS, would that be even noticable?


Offline Hearteater

  • Core Member
  • *****
  • Posts: 2,334
I think this may be a good opportunity to buff the mid-range AIP game (250-400 before first HW).  Right now the only reason low AIP is dominant is there isn't a compelling advantage to more systems.  If the hacking mechanic was ideally suited for mid-range AIP, and harsh for low AIP, that might give enough benefit to mid-range that we could come off the floor.  Super high AIP is generally FS dominated, so I'm not sure what hacking could possibly offer that playstyle.

Offline KDR_11k

  • Hero Member Mark II
  • *****
  • Posts: 904
Does the AI need a special anti-Cinth logic?

Offline Diazo

  • Master Member Mark II
  • *****
  • Posts: 1,717
  • I love/hate Diff 10
Hmmm.

That would be interesting actually. Make an AI-Cinth and then put it up against AI-Faulty Logic.

Watch your CPU give up in frustration, or go up in flames.

Anyways, actually on topic.

This is now getting elaborate enough that is could stand on it's own as a minor faction, or even an alternate win condition. You don't destroy AI command stations, you have to hack them. Then there is ways of snowballing the hacking so you are able to hack the AI Home Command stations after certain pre-requisites are met.

D.

Offline Oralordos

  • Sr. Member Mark III
  • ****
  • Posts: 434
  • Suffering from Chronic Backstabbing Disorder
Next expansion! Centered around gaining control over the AI networks through hacking, you will reclaim the AI back into humanity's control!

Offline KDR_11k

  • Hero Member Mark II
  • *****
  • Posts: 904
If you are Cinth the AI enters anti-Cinth mode and the only way to keep it from spamming HK Vs like candy is to continually hack to make the AI forget that you're Cinth!

Offline keith.lamothe

  • Arcen Games Staff
  • Arcen Staff
  • Zenith Council Member Mark III
  • *****
  • Posts: 19,505
I think there's some confusion on my last post.  A few points to clarify before I have to run:

The multiplier I was talking about is the sort that "the lower it is, the better for the player", so it would start at a relatively high value (representing that the humans can only get away with a limited amount of hacking before distracting the AI with some overt actions of some kind) and decline logarithmically as AIP increases, until it reached some minimal value.

I clarify that because it looks like some folks were thinking that the hacking response would get _worse_ at higher AIP, rather than easier.  I'm not looking for another way to make AIP hurt.  I have other ways for doing that ;)


@Toranth: yes, you're right, the number of players factor was actually making it 80k per planet's K hacked, thanks.


@_K_: yes, this would actually be a pretty substantial buff to coprocessors.  And I think that's great, actually :)  They're currently a very backhanded bonus to the player, which is fine in the whole "the whole game is trying to kill you" sense but I won't be sad to see them get more of a use.
Have ideas or bug reports for one of our games? Mantis for Suggestions and Bug Reports. Thanks for helping to make our games better!

Offline Fluffiest

  • Full Member
  • ***
  • Posts: 140
I'm nervous about the hacking response modifier being based on total AIP rather than effective AIP. That feels like even more of a buff to (already dominant) low-AIP strategies.

A low-AIP game has a large window after killing all the data centres and co-processors where the total AIP might be quite high, but the effective AIP is glued to the floor. During this time, the normal AI response is suppressed, and it seems that the hacking response would also be suppressed.

A high-AIP game generally won't have this window, because the AIP reducers generally won't take you all the way down to the floor.

Also, I'm generally opposed to anything that forces the player to think about the difference between total and effective AIP.

Offline Radiant Phoenix

  • Sr. Member Mark II
  • ****
  • Posts: 362
I think this calls for different AI Command Centers that give bonuses:

1) Remove 'Data Centers' and add AI Data Station (a command center type) that reduces AIP by x/y minutes when hacked; increase AIP floor by z per minute. The command center would respond locally by increasing its spawn without affecting other areas.
You mean, the Superterminal? How about we just make the "Data Station" give a 20 AIP reduction when hacked, even after it's blown up? (which cancels out the 20AIP if you do blow it up)

2) Add AI Assault Station that reduces spawning in the immediately surrounding systems; this increases AIP by x/y minutes when hacked.
Better idea:
- assuming we do go with "individualized AI command stations" like I mentioned elsewhere recently, there could be a variety that buffs reinforcements but if you hack it you gain permanent (as long as that station stands, at least) metal income from the resource spots on that planet
Because spending AIP to temporarily have less enemy ships to deal with wasn't very appealing when warheads offered it to me.

3) Add AI Fabrication Station that allows for the ability to build Mk IV ships when hacked (takes x minutes); this increases AIP but the ability is permanent.
- hacking an advanced factory before capturing it to permanently unlock mkIV production capability (on a mkIV, or I guess mkIV&mkV tab on the normal space dock; whatever), though this would need to have a huge impact on the hacking response level to be at-all balanced.

4) Add AI Subroutine Station that increases knowledge by x but costs y AIP per minute when hacked; increase AIP floor by z per minute.
So, standard knowledge hacking made more complicated.

5) Add AI Energy Station that increases energy pool by x when hacked; this effect last only as long as the station is being hacked and increases enemy spawns in-system.
- in the same vein, there could be an AI command station variety that gives a planet-wide munitions boost (or whatever) but if you hack it you gain permanent (as long as that station stands, at least) energy production from that planet

Offline keith.lamothe

  • Arcen Games Staff
  • Arcen Staff
  • Zenith Council Member Mark III
  • *****
  • Posts: 19,505
On the total vs effective AIP thing, I'm actually more in favor of basing it on effective AIP.  But that then makes it so you want to time hacks to come after increases (which is fine) and before popping DCs/CoProcs (which I'm not sure is a good thing) and could make superterminal truly crazy because it would be increasing antagonism and decreasing "antagonism tolerance" at the same time.

FWIW, this would not be the first thing to care about total instead of effective; the AI's extra bonus type unlocks are done according to total.
Have ideas or bug reports for one of our games? Mantis for Suggestions and Bug Reports. Thanks for helping to make our games better!

Offline Winge

  • Hero Member
  • *****
  • Posts: 601
Yeah, I don't think it would be bad for hacking to be based on effective AIP.  Then, the player would have to chose whether to take out DCs and CPs early (thus making hacking very difficult), or getting AIP and then hacking (risking reinforcements, waves, or even a CPA in exchange for an easier hack).  That said, I'm not sure how I would set up the Superterminal hack.  Maybe make it work off of the AIP you have to start with (a 'constant' variable), and ignore AIP changes during the hack?  Come to think of it, the floor changing would affect the ST hack anyways, as right now a ST tick is +1 AIP, -2 AIP...
My other bonus ship is a TARDIS.

Offline Fluffiest

  • Full Member
  • ***
  • Posts: 140
On the total vs effective AIP thing, I'm actually more in favor of basing it on effective AIP.  But that then makes it so you want to time hacks to come after increases (which is fine) and before popping DCs/CoProcs (which I'm not sure is a good thing) and could make superterminal truly crazy because it would be increasing antagonism and decreasing "antagonism tolerance" at the same time.

FWIW, this would not be the first thing to care about total instead of effective; the AI's extra bonus type unlocks are done according to total.

Can the Superterminal base its retaliation solely on the AIP level before you started hacking it? That might make it a little bit less berserk.

Offline keith.lamothe

  • Arcen Games Staff
  • Arcen Staff
  • Zenith Council Member Mark III
  • *****
  • Posts: 19,505
Can the Superterminal base its retaliation solely on the AIP level before you started hacking it? That might make it a little bit less berserk.
That would be fiddly, but I suppose it's possible since there's only ever one "the superterminal hack has started" moment in any game.  It only ends when the thing is destroyed.

And it wouldn't be solely based on the earlier AIP level, it has to be based on current hacking antagonism too or it never escalates and someone can just ride it all the way to the floor pretty easily.
Have ideas or bug reports for one of our games? Mantis for Suggestions and Bug Reports. Thanks for helping to make our games better!

Offline contingencyplan

  • Full Member
  • ***
  • Posts: 147
Just finished reading the 6-thread page, so apologies for not quoting / linking. Also, breaking into multiple posts for easier reading.

Put me down for "HELLS YES." Like I said in one of the prior threads on the topic, a new mechanic is an opportunity to cut some Gordian Knots in playstyles, and this fits the bill perfectly.

If nothing else, this would make me shut up about indestructible capturables, so everybody wins! ;)

I do think that the hacking should be based on total, rather than effective, AIP in order to avoid having to optimally order capture / destruction of AIP-reducing things, which just increases micro for no reason. In particular, using effective AIP would nerf Spire CivLeaders (which I use in place of auto-increasing AIP), since my hacking options would decrease as time goes on (assuming I free all of them, which I tend to do).

Regardless, I think that the SuperTerminal mechanic should probably be changed to a straight decrease, rather than the current +1 - 2 mechanic, to avoid the "You got your hacking in my AIP!" conflict.