Fab hacking insanity

[Nibelung44]

Any article on hacking? Except knowledge hacking I don't know how to do the others.

Plus it seems suicidal to hack most of the time...

[Shrugging Khan]

I disagree most fervently with any changes that would make the HaP prices of the hacks vary. I'm too tired to coherently explain why, and I hope someone smarter than me can explain why that would be a very bad idea, but the price of an action must stay the same regardless of the stage the game is in. Making AI response the flexible variable (as it is, although the exact formula certainly needs some work) is the right way.

[chemical_art]

I disagree most fervently with any changes that would make the HaP prices of the hacks vary. I'm too tired to coherently explain why, and I hope someone smarter than me can explain why that would be a very bad idea, but the price of an action must stay the same regardless of the stage the game is in. Making AI response the flexible variable (as it is, although the exact formula certainly needs some work) is the right way.

From what I understand, the.cost of.hacking remains constant. What.is.changimg.is the.reaction if the.AI in positive terroritory.

In effect, these changes make it so.that the player doesnt feel.compelled to ride.ultra.low aip simply to minimize hack response.

[Tridus]

I disagree most fervently with any changes that would make the HaP prices of the hacks vary. I'm too tired to coherently explain why, and I hope someone smarter than me can explain why that would be a very bad idea, but the price of an action must stay the same regardless of the stage the game is in. Making AI response the flexible variable (as it is, although the exact formula certainly needs some work) is the right way.

From what I understand, the.cost of.hacking remains constant. What.is.changimg.is the.reaction if the.AI in positive terroritory.

In effect, these changes make it so.that the player doesnt feel.compelled to ride.ultra.low aip simply to minimize hack response.

Yes, exactly. Hacking costs stay the same, but the hacking response is based on AIP - HaP instead of just AIP, so while you have HaP available, hacks get easier. As you use it, hacks get harder. The cost of the hacks themselves won't become variable due to this change.

[Tridus]

Hmm, ok, fair enough.  For next release we can try base response = AIP - HaP.  This will actually make negatives hurt more, but I don't mind

Cool.

Negatives should hurt, really. Using the explanation that HaP is your knowledge of exploits and such of the AI code, being out of it and trying to brute force instead is likely to provoke a huge response.

Also if you're concerned about FS games where people take 40 systems making too much hacking viable, you could lower the amount of HaP you get per planet after you take X planets. That would still let taking more planets give you more hacking strictly speaking, but would slow it down so that you can't just hack all the things.

Another option would be to add a FS mod to hacking response, as the exo-galactic AI network starts paying more attention due to your Spire alliance it would make hacking more difficult.

(Also, I would really love a high HaP cost action to disable an AI bonus ship. No Tackle Drone Launchers for you, AI!)

[LaughingThesaurus]

In FS games, you already have a super overpowered FS fleet that rips everything apart and your fleet ships comparatively do next to nothing. I don't think it's a problem if you can do a lot of hacking for a lot of perma-fleet ships and stuff. The superterminal's probably a problem, sure, but not fab/ARS/etc hacking. That kind of hacking is really just going to get you more of something that's already fairly weak in comparison to what you have.

[Martyn van Buren]

Would it make sense to have shard retrieval cost HaP? If it's going to be a base mechanic it seems like it might as well get a role in some older parts of the game, and shards do have kind of a similar vibe.

[GP_Trixie]

I don't think there is a balance problem even for superterminal as the cost of everything with the new hacking system is exponential. Reducing AIP by 60 costs 140 HaP, but after that it's 160 for the next 20 and then 320. I even think it's a bad move to spend that must HaP for such a low reduction. With the new system it's far better to diversify your hacking.

[LaughingThesaurus]

I don't think shard recoveries involve much in the way of hacking, or of compromising the AI security, and I don't know that the AI bolsters its security network when you do successfully complete a shard recovery. Been a while since I've done anything fallen spire, but don't you just 'detect' the shard, rather than hacking the AI for it? Sensors take time like hacking, sure, but they don't do anything but sense stuff right?

[LintMan]

@LaughingThesaurus - no, the FS stuff doesn't currently use any hacking.  I think MvB was just suggesting that could be changed, to address keith's concern that FS players will have too much HaP.

Personally, I think you're right that the main FS fleet stuff makes any hacking bonuses almost irrelevant, but at the same time, some variations in the "detect shard, retrieve shard, fight off onslaught, build city" lather rinse repeat cycle of FS would be welcome.

[chemical_art]

As others have pointed out, the effect of FS on hacking is limitted due to escalating costs.

It is true if one diversifies their hacking they can do quite a bit..but in the big picture of FS, does in relatively change things so much that base game hacking must be nerfed? I doubt it.

[LaughingThesaurus]

Oh, I absolutely want variety there too. That's probably going to be coming in some kind of later update after hacking stabilizes. I think this is more of a complete hacking focused 'does it need nerfed' discussion. I actually totally forgot about the escalating costs per hack as well. So, you can't just pile up a massive amount of fab hacks with a massive amount of AIP. Not to mention, again, the benefits of those extra ships probably won't contribute in too huge of a way in FS (Unless it's like, maybe a protector starship fab, which would be totally awesome).

[Toranth]

Warning:  Wall-o-text incoming.
I was preparing a discussion post on the new Hacking that we've seen so far, and what it might mean.  I'm hesitant to draw any conclusions yet, because this is such a change from the way it used to be, but I think putting numbers out there will make it easier for eveyone to talk to the same points.
Please correct me if I've got anything in here wrong - I've tried to double check everything, but, well, history.


All in all, once you look at the numbers, this is a MAJOR rebalance of the existing features, without even considering all the awesome new stuff.
Let's look at what used to be first.


K-raiding used to be about 80,000 antagonism per planet - meaning 2-3 was the limit on casual k-raiding most games.
ARS hacking went 30,000; 90,000; 270,000; 810,000; etc.  Meaning you COULD hack 3 ARS without much difficult.  Hacking 4 was difficult without beachheading.  Hacking 5 was basically impossible without MAJOR perparation to handle the huge spawns.
SuperTerminal hacking was 1500 antagonism per tick/per AIP reduced.  A 200 AIP reduction ride was about 300,000 antagonism, which like the 4th ARS, required either a major beachhead or was difficult with a fleet-only effort.


Now, K-raiding is 30*2^(n-1) Hacking Progress cost, for 30, 60, 120, 240 as the per-K-raid cost.  Total costs are 30, 90, etc.

K-Raid #	Unit cost	Total cost	Old Antag.
1	30	30	80,000
2	60	90	160,000
3	120	210	240,000
4	240	450	320,000
5	480	930	400,000
6	960	1890	480,000
7	1920	3910	560,000

That's AIP equivalents, remember, for the new HP.
So, it used to be possible to casually K-raid 2-3 worlds, and 4-5 with beachheading (and more with extreme preparation - think multiple fortress cap beachheads when hacking).
Now, to keep the HP positive, you need to capture 1.5 systems to k-raid once, 4.5 systems to k-raid twice, etc, up to capture 100 systems to k-raid 6 systems.
As you can see, the cost used to be linear, but is now exponential.  This puts a much harder cap on how much k-raiding can be done.  About 3 would be my estimate.  That 4th K-raid gets nasty fast as you go negative.


Now, ARS hacking was 10,000 * 3^(n) antagonism, now it's 50 * 2^(n-1).

ARS Hack #	Unit cost	Total cost	Old Antag.
1	50	50	30,000
2	100	150	90,000
3	200	350	270,000
4	400	750	810,000
5	800	1550	2,430,000

This was exponential both before and in the new system.  Whereas before, the first ARS hack or two was easier than K-raiding but the later ones were much harder, it is now the reverse:  the 5th ARS hack is about as difficult as the 6th K-raid; it used to be the 5th ARS hack was as difficult as the 30th K-raid.
However, where doing 2-3 ARS hacks was easy before, it now is much more difficult.  The player will need to think long and hard over whether or not that 3rd or 4th ARS *really* needs hacking.


The SuperTerminal hacking has also gone from linear to exponential.  However, because it was usually triggered so many more times, this has gotten a bit out of control.

Planet Equiv.	HP Cost	AIP Reduction	AIP Increase	Net AIP Reduction	HP Change
1	20	40	20	20	0
2	40	80	40	40	-20
3	80	120	60	60	-80
4	160	160	80	80	-220
5	320	200	100	100	-520
6	640	240	120	120	-1140
7	1280	280	140	140	-2400
8	2560	320	160	160	-4940
9	5120	360	180	180	-10040
10	10240	400	200	200	-20260
11	20480	440	220	220	-40720
12	40960	480	240	240	-81660
13	81920	520	260	260	-163560
14	163840	560	280	280	-327380
15	327680	600	300	300	-655040
16	655360	640	320	320	-1310380
17	1310720	680	340	340	-2621080
18	2621440	720	360	360	-5242500
19	5242880	760	380	380	-10485360
20	10485760	800	400	400	-20971100

As you can see, anything beyond net 180 AIP reduction puts you beyond Hacking Progress -9,000, which is death.  That's a pretty hard cap, alright.
Previously, a normal game would be AIP total 450, AIP Reduction 250-300 (100-ish from Data centers, rest from ST), for an Effective AIP of around 150-200.  This is no longer possible.
As it stands in 7.005, 100 points of net AIP reduction is possible, 150 pushing it.  200 you are seriously risking Game Over no matter how good your preparations are.


The FacIV/ASC hacking is HP Cost 100 for the first, 200 more (300 total) for the second.  Seems reasonable to me.

Fabricator hacking is 20 * 2^(n-1).

Fab #	Unit cost	Total cost
1	20	20
2	40	60
3	80	140
4	160	300
5	320	620
6	640	1260
7	1280	2540

Again, the exponential increase makes doing more than a few utterly impractical.


Why is going negative so impractical?
Here's the Doom Chart - What the AI response looks like based on your Hacking progress level:

Net HP	AI Response Multipler
0	Effective AIP
-50	300
-100	1100
-150	2400
-200	4200
-250	6500
-300	9300
-350	12600
-400	16400
-450	20700
-500	25500
-550	30800
-600	36600
-650	42900
-700	49700
-750	57000
-800	64800
-850	73100
-900	81900
-950	91200
-1000	101000

Note that this is the 'ideal' case: if your effective AIP is higher than the absolute value of your negative Hacking Progress, the response will be even higher.
So, at -100 hacking Progress, it's like you were hacking with an Effective AIP of 1100.  At -1,000, it's like you had an AIP of 101,000.
The Hacking Progress response is basically a polynomial equation y = .1*x^2 - x, where the dominant factor is x^2. 
This means that for a doubling (x2) in the Hacking Progress negative, the AI response quadruples (x4).  For a tripling (x3), the AI gets a nonupling (x9).
So, the more hacking you do, as the current HP goes negative, the AI response goes up crazy fast.  And this is all compounded by the fact that HP prices increase exponentially.




Summary
K-Raiding has taken a nerf, going from a linear cost increase to an exponential one.
ARS hacking is mostly unchanged, a bit harder than it used to be.  Call it a 1 ARS hack reduction in the practical case.
SuperTerminal has taken a super nerf, nearly gutting the old multi-hundred point runs that were a mainstay of keeping AIP low.
Fab Hacks and FacIV/ASC hacks are wonderful, meaning you can finally get some use out of those fabs on/adjacent to Core worlds.  Also cuts down on the need to protect all your stuff.

Strategic implications?
Because cost increases are all exponential, it's far better to scatter your Hacking points around into the various categories.
For example, 200 Hacking Progress lets you get approximately:
1 of each (K-Raid, ARS, Fab, ASC/FacIV)
or 3 K-Raids
or 4 Fabs
or 2 ASC/FacIV
or 3 ARS
or 75 AIP Reduction.

Second, because of the SuperTerminal nerf, you can't reduce AIP nearly as far as you could before.  This means you'll need to avoid raising the AIP as much - by using Hacking to avoid capturing stuff.
Don't capture that Fab or FacIV, hack it, and save the AIP and HP you would have spent for less return at the ST.

[TechSY730]

I sort of like the fact the new system strongly encourages you to spread out what you hack, instead of dumping it all into 1 or 2 things like the old system did.

On the other hand, I am very wary of the exponentials. Exponential formulas are VERY hard to balance (look at Borderlands 2 for example. They went exponential, and because of that, the gameplay sort of falls apart into extreme grind territory filled with near instadeaths at around level 45). Even if you do balance it, there is the lack of granularity. Like hack N is difficult, but not overwhelming or terrifying, and hack N+1 is pretty much insta-death, with no warning that there would be such a jump.

Maybe repeated hacking of the same "thing" should be polynomial instead of exponential? That would still accomplish the goal of encouraging diversity, but be less likely to have "worst case cliffs", and not be quite a "out of the blue" punishing.

[madcow]

This is a pretty insight last couple of posts. And I think does a good job summarizing what I saw (and why) in my posts in the 7.002-7.005, the exponential growth really makes it go from easy to death ball in a snap.