What is *ACTUALLY* going on with Armor

Xinxspuz
Line Manager
Line Manager
Posts: 30
Joined: Sun Jun 07, 2015 11:06 pm

What is *ACTUALLY* going on with Armor

Postby Xinxspuz » Sun Aug 09, 2015 3:59 am

Bear with me, this will be a long post, with lots of images (as they say, a picture is worth a thousand words.)

According to the pop-up text in the Shipbuilder, if the Average Armor Value (AAV) of a ship "is greater than the armor penetration value of a weapon, it will reflect 97% of that weapon's fire without any damage." This seems pretty straightforward. The AAV is continuously updated as you build your ship, and handily displayed in the lower right corner of the screen for your reference. Adding additional armor plating seems to have diminishing returns. I can't put more than 6 or 7 armor plates on a Dreadnought before the AAV starts going DOWN with additional plates, rather than up.

Now, take this ship design (not my design; credit goes to alix965 for posting the challenge that I noticed this in):
2015-08-08_00007.jpg
Scrap Hauler
2015-08-08_00007.jpg (356.26 KiB) Viewed 4323 times

It lists an AAV of 10.24, which seems pretty lackluster. Most weapons should have no trouble penetrating that armor value.

Wait, so what's happening here?
2015-08-08_00002.jpg
HBL_fail
2015-08-08_00002.jpg (353.38 KiB) Viewed 4323 times
A Heavy Beam Laser, with an Armor Penetration of 25, has hit the ship, but failed to penetrate the armor. Hmm. That's weird.

Let's try an even heavier weapon:
2015-08-08_00004.jpg
PB_fail
2015-08-08_00004.jpg (347.3 KiB) Viewed 4323 times
That's a Particle Beam, with an Armor Penetration of 30 (the highest currently in the game), also bouncing off the ship's armor.

This was the point at which I started to get curious. Obviously, if the ship had a higher in-game AAV, then it was being calculated differently than the stat the Shipbuilder gives you. I suspected that the game was still using the GSB1 formula: Total armor points, multiplied by hull bonuses, multiplied by stacking modifiers, divided by number of module slots actually used on the ship design.

So I designed a hull, calculated the armor value using the GSB1 formula above, and labeled it. As an added bonus, I wanted to test whether the in-game formula was applying the stacking penalties to the first armor plate or not, so I picked a value that would give me a 31 AAV without stacking penalty, or a 26 if the stacking penalty was applied (even though there is only one armor module on the ship hull). Here's the hull in Shipbuilder, for reference. Note that the listed AAV isn't even CLOSE to what the in-game AAV is:
2015-08-08_00009.jpg
D31
2015-08-08_00009.jpg (354.04 KiB) Viewed 4323 times

And here's the hull getting blasted by a Particle Beam, with 30 Armor Penetration (sorry about all the purple nebula clouds):
2015-08-08_00008.jpg
PBvsD31
2015-08-08_00008.jpg (278.1 KiB) Viewed 4323 times
As you can see, the in-game formula is applying the stacking penalty to even the very first armor module, which means you're really not getting your money's worth when you add armor to your ship.

For funsies, here's a second iteration of the Particle Beam test, with a target dreadnought configuration utilizing two armor modules, to check whether the the armor stacking penalty is being applied linearly or geometrically:
2015-08-08_00010.jpg
PBvsD3129
2015-08-08_00010.jpg (282.87 KiB) Viewed 4323 times
Answer: It's geometric, at least in-game.
In-game armor formula.PNG
ingame
In-game armor formula.PNG (3.6 KiB) Viewed 4323 times
(Where H is the hull armor modifier, M1p is the first armor module multiplied by its stacking penalty, M2p^2 is the second armor module multiplied by the square of its stacking penalty, M3p^3 is the third armor module multiplied by the CUBE of its stacking penalty, etc, and n sub M is the total number of modules actually installed on the ship hull.)

This is basically the formula used in GSB1, except that in GSB1, the stacking penalties for armor were generally around 0.97. This meant that even at your 10th armor module, that module was still providing nearly 75% of its rated armor protection. However, with a stacking penalty of 0.85, by the time you get to your 10th armor module in GSB2, the in-game performance is less than 20% of the rated armor value for that module alone.

So, where is that number in the Shipbuilder coming from?
Answer: Here's the formula:
Shipbuilder armor formula.PNG
shipbuilder formula
Shipbuilder armor formula.PNG (3.27 KiB) Viewed 4323 times
(H sub n in this formula is the total number of module slots on the ship hull, filled or not. For some reason, the Average Armor value adds one to this "number of modules" value before using at as the divisor.)

Now, normally, I'd be a little put out by the second formula. You see, it penalizes ALL of your armor modules every time you add another armor module to the ship. That's why you can't seem to get an AAV higher than the 18-20 range in Shipbuilder - adding additional plating after that will actually make your AAV go DOWN because the new plate adds a penalty to every plate before it, not just a penalty to itself. To be fair, the Shipbuilder formula is correctly ignoring the stacking penalty for the first armor module installed, but it snowballs pretty quickly after that.

But, as I showed in the first part of the post, the game isn't even USING that AAV that's being calculated and shown in Shipbuilder, so... why get mad?

tl;dr
The Average Armor shown in Shipbuilder is a complete garbage number - it doesn't reflect in any way, shape, or form, what the actual in-game armor value of your ship is. Cliff, why???

neverless
Line Manager
Line Manager
Posts: 33
Joined: Mon Oct 15, 2012 10:21 am

Re: What is *ACTUALLY* going on with Armor

Postby neverless » Sun Aug 09, 2015 6:55 am

Hmm...

curious... could this explain an observation I made....?

I have a Dread with probably around 4 Powered Armor modules, exact number doesn't matter. Lots and lots of fighters, no other ships shoot their weapons and always "Armor too thick". I get a few lucky hits (single digit damage values). If I do not ship a repair module for the armor, after about 5 to 10 of those lucky hits, suddenly they are getting through the armor?! I tested this in both directions, it happens with my ships and also with the opponent's ships. It also does not happen with fighters, also with beam weapons and missiles of larger ships.
Don't know if I am still waiting for a Linux update newer than V1.35, or have given up... :-(

Alekan
Type I Robot
Type I Robot
Posts: 47
Joined: Sat May 02, 2015 3:38 am

Re: What is *ACTUALLY* going on with Armor

Postby Alekan » Sun Aug 09, 2015 11:22 am

Heh, good catch. In GSB1, the actual effectiveness of a ships armor was dependent in part on how many modules were installed. A Cruiser with 2 modules and an armor plate had extremely high Resistance, because there were only 2 modules to armor. As you added more modules, the armor was spread thin and the Resistance went down.

A a general rule, in GSB1 any Armoured ship was built with the Minimum of modules, as each additional module drastically reduced Resistance.

In GSB2, the Armor Resistance in the shipbuilder displays resistance as if a ship had ALL slots filled. Regardless of just how many slots are actually filled. I had always thought it was just one more change, but it sounds like the display is bugged and the full armor system was actually carried over.

If true, it would explain why those test ships – which have very few equipment modules – could stand up to heavy firepower. In GSB1, a Light Carrier (Hangar, Engine 1, Micro Crew) had an armor rating of 29.00 with a single Ultraheavy Armor Plate (145.00 armor). Remove one of the 3 modules and the armor rating jumps to 36.25.


I’m going to start using Plated DNs with 3 modules (Gun, Engine, ArRep). If the theory holds, nothing in the game will be able to overcome the ACTUAL Armor Rating.

…After some testing, the theory is true… Even a Particle Accelerator (ArPen 30) just bounces. WTF. At least for Cruisers and DNs armor is stupidly durable if you use Gun/Engine/ArRep. As in a squadron of Cruisers will bounce off an armor tank until the repair system runs out of supplies. Resistance really is a fundamentally broken mechanic…
Armor Durability.jpg
Armor Durability.jpg (942.47 KiB) Viewed 4312 times

Again, WTF. Well, I guess this is the new FOTM (it's news to me anyway). My 3 Battlecruisers pounded away at that DN until its SINGLE Armor Repair system ran out of supplies. 18 heavy guns - and 3 of those were Heavy Beams... nothing. That DN is 2625 cred, and its easily worth 5x that as a tank. ArPen 30 weapons don't work either. Note that the DN is the best platform due to its Antimatter engine, good base fitting, and specialized Armor Repair Array. Cruisers are a distant second. Frigates and DD's don't have the fitting to take advantage of this (and Graphite is crazy expensive).

I tried a dual-rep version of the DN... it tanked 5 of those BCs until it ran out of supplies. I always thought the Resistance system was fundamentally bad, but this really proves it. Left as is, Armor is on Par with Shields... and its a broken mechanic just like shields. If Armor Resistance is 'fixed' to match what the display shows, it will be completely useless.

Xinxspuz
Line Manager
Line Manager
Posts: 30
Joined: Sun Jun 07, 2015 11:06 pm

Re: What is *ACTUALLY* going on with Armor

Postby Xinxspuz » Sun Aug 09, 2015 1:08 pm

My "guess" as to what happened was that Cliff really did intend to "fix" the armor system, to prevent ships with 55+ armor ratings (like you would get sometimes in GSB1). He modified the stacking values, and wrote (intentionally or otherwise) a new stacking formula for EVERYTHING, which makes it counterproductive to add more than 4 or 5 of a single type of defense module to a ship.
Then, as programmers have done since the dawn of time_t, a block of code got copied over from GSB1, still containing the old formula. Now we have a mismatch between what the Shipbuilder tells you, and what your ship actually does in-game.

Personally, I prefer the old method of calculating armor. It gives rise to a greater diversity of ship designs. That has a caveat, though: there HAVE to be weapons with armor penetration values higher than 30, if the armor values themselves are capable of going higher than 30.

Cliff, I know you've sweated bullets over some of the balance issues currently in the game. This is a prime opportunity to fix them. If you revert the armor system fully back to the old method (including in the Shipbuilder), and boost the armor penetration on some of the armor-cutting weapons (Beam Lasers, Fusion Beams, Particle Accelerators, etc.) to match, you give people a way of dealing with the currently ubiquitous Pulse Cannon problem. Best part? We already had that way, we just didn't realize it.

User avatar
cliffski
Positech Staff
Positech Staff
Posts: 7968
Joined: Sat Feb 11, 2006 10:27 am
Location: UK
Contact:

Re: What is *ACTUALLY* going on with Armor

Postby cliffski » Sun Aug 09, 2015 5:52 pm

Yikes. well done, I'm amazed nobody has spotted this before. I have screwed up! Possibly because I almost never leave slots empty in my own designs...
basically the ship designer is correct and the actual battle sim is WRONG, but only in cases where you have empty slots. The code that was counting the number of slots was stupidly counting the number of filled ones, when i was sure it wasn't.
Obviously its trivially fixed...
But it does have balance implications, in that currently armor resistance for some designs is way, way too high. It should be vastly lower, but for 99% of ships it makes no difference, only designs where slots are empty.
Any real good reason not to just fix that for the next patch?

Alekan
Type I Robot
Type I Robot
Posts: 47
Joined: Sat May 02, 2015 3:38 am

Re: What is *ACTUALLY* going on with Armor

Postby Alekan » Sun Aug 09, 2015 10:02 pm

Cliffski, the takeaway from this situation is that Armor is SO BAD that no-one even bothered to see if this still worked. After Release, I have never seriously fielded an Armor Tank, simply because under ‘intended’ game mechanics Armor is a ‘fail cascade’ system. A single Heavy Beam Laser hit (65dmg) on a max-armor 6-Plate 80% Bonused DN will open the door for all 19 ArPen Heavy Pulse Cannons. Once the beam lands it only takes one instant follow-up shot (from any heavy weapon that can pen) to guarantee that even 3 Armor Repairers wont be able to recover the resistance rating. That is the best armor can do – and it’s a very expensive suicide ship.

All non-bonused Armoured ships have an additional weakness: with 6-Plates operating at only 44.37%, each Heavy Armor plate has only 62 Armor… which means a clean shot from a Hvy Beam Laser (65dmg) will One-Shot any armor plate. You cant repair destroyed plates/modules. Again, it’s an expenisve suicide ship. Armoured Frigates and Fighters are jokes.

The other main issue with armor is that since it is based on the total number of modules on a ship, ships with more module slots are severely penalized.
---The Eisenhower DN (80% Bonus Armor) has a Max Average Armor of 20.33
---The Yamamoto DN (85% Bonus Armor) has a Max Average Armor of 19.70
-----the ‘Better’ armor tank is actually worse because it has more total modules



Thoughts on Armor:
Under the current system, the fatal flaw is the lack of granularity. Its that simple. Resistances are so compressed that there is no tolerance for damage – a 20.33 Armor tank is dead if its rating drops to 19.

In GSB1 a 70 Armor Tank could survive a few shots until missiles could pen… and then survive a few more shots before Cruiser Pulse could pen. The main problem in GSB1 was the edge-cases of Armor being affected by the amount of equipment modules on the ship. That was the only way you could get 70+ Armor Tanks.


As such there are 4 action points to address the situation:
---Cliffski you should pick a fixed module value for each size class (DN, CC, FF, DD, etc.) and use that for all ships of that size class. This prevents unfair Armor penalties to larger ships and eliminates hull-based edge-cases. This should be implemented regardless of any other changes to Armor.

---Armor should be fixed so it is not affected by the amount of equipped modules – it is only affected by the fixed Module Value of the Size Class (see above).

---Stacking penalties for both Armor and Armor Repair modules should be reduced to 0.90. This is consistent with Shield Stacking and will make Armor more useful (better on HP/Cost, more buffer for Repair systems to work with).

---With the elimination of Hull and Module Edge-Cases, the potential of Armor will be far easier to plan for. Anti-Armor weapons should have their ArPen rebalanced to account for the new resistance ratings.


What this does:
-Armor Plates have greatly increased value due to reduced stacking penalties.
-Far greater granularity and gap between Beam Lasers and Pulse Cannons (making Specialist Anti-Armor weapons more attractive)
-Increase in buffer increases time for Repair systems to work (making them actually useful)
-Reduction in edge-cases makes Armor more intuitive and easier to balance
-ArPen is rebalanced to account for vastly higher Armor Resistances
-Increases Player need to diversify weapon selections (Pulse is no longer the answer to everything)
-Armor is still inferior to Shields (no Remote-Repair systems…yet)

If you are truly paranoid about Armor becoming too good (…you are Ultra-Conservative in balancing…) then there is the perfect solution:
Radiation Guns!!!

You just reworked the radiation effects. How about Radiation weapons are stopped by Shields, but Bypass Armor. The counter-part to Shield Destablizers. Yeah it doesn’t make 100% sense, but whatever. The point is that it would be a way to stop Armor Tanks. There are two ways to do it:
-Radiation Bypasses Armor and Damages the Hull, meaning the ship will eventually be destroyed
-Radiation disrupts the Repair Nanites, making a ship suffering from Radiation unable to Repair itself (or Repair Systems operate at a penalty)
-If you are worried about Radiation getting out of hand… we already have a Radiation Shield Module

Xinxspuz
Line Manager
Line Manager
Posts: 30
Joined: Sun Jun 07, 2015 11:06 pm

Re: What is *ACTUALLY* going on with Armor

Postby Xinxspuz » Mon Aug 10, 2015 1:48 am

Cliff, before you go adjusting the in-game armor calculations to match the stat displayed in Shipbuilder, consider the following points regarding the Armor Value and Armor Penetration as it stands in Shipbuilder:
  • Right now, it's impossible to get an Armor Value higher than 21, on ANY ship hull. This is a major contributor to why Pulse Cannons are so broken - with SP of 23 and AP of 19, they can chew through any shield module and any amount of armor in the game, all while already having one of the highest tracking speeds and highest DPS vs. Hull of any weapon in the game.
  • As Alekan pointed out, Armor Value is too constrained. I can get an AV of ~10 with two modules, but it takes me 5-6 modules to get that AV to 19 or 20, and even then that's only possible on certain hulls. In game terms, it really doesn't matter whether your AV is 5, or 10, or 19 - a Pulse Cannon can chew through all of it. And everyone has to take Pulse Cannons because they're so powerful.
  • Right now, high armor is the only hard counter to Pulse Cannons. If you "fix" armor, you're just perpetuating the "Pulse Cannons Are Broken" problem.

If you want Armor Value to function the way it does in Shipbuilder, that's fine, but you need to open up the range so it's actually possible to get Armor Values above 20. This retains armor as a counter to Pulse Cannons, while also increasing the relevancy of Beam Lasers and Particle Cannons. If you raise the Stacking Penalty for armor, across the board, by 0.05 (in other words, modules with 0.85 will go to 0.90, and modules with 0.80 will go to 0.85), and fix the in-game to match the Shipbuilder formula, it will enable Armor Values up to 32 on certain Dreadnought hulls, and 35+ on certain Cruiser hulls (this is assuming you completely max out the armor, to the point where the inflection of the armor function flips - i.e. 9-10 modules devoted to armor). This preserves high armor values as a barrier to medium-penetration weapons. You can then tweak the Armor Penetration of the beam-type weapons a little higher, so they can penetrate the 30+ AVs, and thereby encourage people to have a little more variety in their weapon selection.

IIRC, in GSB1, the weapons with the highest AP were... Proton Beams, Beam Lasers, and Fusion Beams? Missiles and Plasma Torpedos had "moderate" AP, while the rapid-fire weapons had low AP. Right now, pretty much every weapon in GSB2 has "moderate" or "high" AP, if we go by the Shipbuilder values. That means armor is a waste of a module most of the time, and also that you've backed yourself into a corner in terms of 'countering' certain classes of weapons. Fixing the formula, reducing the stacking penalty, and increasing the AP values of some weapons that already have high AP will improve the overall balance of the game, while also fixing the "bug" between the Armor Value displayed in Shipbuilder vs. what's happening in game.

tl;dr

Fix the formula in-game, but ONLY if you also reduce the Stacking Penalty for all armor modules. This will enable Armor Values up to ~32 for Dreadnoughts, and ~38 for Cruisers.
Increase the AP values for some of the beam-type weapons to handle the higher Armor Values that will result from the above changes.
This will fix your bug as well as improving weapon balance overall, without "breaking" existing challenges (which you seem to worry about a lot).

User avatar
yurch
Supreme Robot
Supreme Robot
Posts: 637
Joined: Mon Nov 23, 2009 3:25 am

Re: What is *ACTUALLY* going on with Armor

Postby yurch » Mon Aug 10, 2015 2:10 am

The stats on pulse cannons are pretty much out of control.

I've maintained throughout GSB1 that tying resistance to armor health is just balancing madness. It doesn't even make intuitive sense - a penetrating hit to the front of the ship makes the rear weak enough to be hit by fighters? That's a limitation of shields, not hunks of metal.

I'd very much like to see what a percentile block chance determined by armor health would look like, instead.

User avatar
cliffski
Positech Staff
Positech Staff
Posts: 7968
Joined: Sat Feb 11, 2006 10:27 am
Location: UK
Contact:

Re: What is *ACTUALLY* going on with Armor

Postby cliffski » Mon Aug 10, 2015 6:18 pm

All very interesting points. I agree armor will need some proper changes now thats fixed... And I think changing thew stacking penalty is a no-brainer. Its definitely a weak choice right now.
I've spent all day crunching down to find some graphics/performance bugs I suddenly discovered so I'll be back on this sort of stuff soon.
The next patch will be a good one :D.

User avatar
cliffski
Positech Staff
Positech Staff
Posts: 7968
Joined: Sat Feb 11, 2006 10:27 am
Location: UK
Contact:

Re: What is *ACTUALLY* going on with Armor

Postby cliffski » Wed Aug 12, 2015 4:59 pm

Ok, I've fixed it and also changed the armor stacking penalties so they are reduced by 5% in each case (so 85% is now 90% etc).
In my test case of a yootan capone cruiser, I can get an average armor of 27.12 with 9 heavy armor modules, meaning that only pulverizer beams and particle accelerators can hurt me, making me invulnerable to everything but a dreadnought.
I think at the very least heavy beam lasers and fusion beams/sledgehammer pulse cannons should be capable of doing some damage here so I propose a general 50% boost to all armor penetration values as a result.

So the multiple warhead launcher would do 22.5 armor penetration, for example.

As well as this, I think the actual damage absorbable values of armor modules are generally too low and need raising, possibly just by 25%, so as to not to completely destabilize everything.
Thoughts on these two proposals?

1) 50% boost to all armor penetration values
2) 25% boost to all armor damage absorbables.

Cheers

Xinxspuz
Line Manager
Line Manager
Posts: 30
Joined: Sun Jun 07, 2015 11:06 pm

Re: What is *ACTUALLY* going on with Armor

Postby Xinxspuz » Wed Aug 12, 2015 6:20 pm

Cliff under no circumstances do Pulse Cannons, Sledgehammer Pulse Cannons, or any derivative thereof need ANY sort of "boost" right now.

User avatar
yurch
Supreme Robot
Supreme Robot
Posts: 637
Joined: Mon Nov 23, 2009 3:25 am

Re: What is *ACTUALLY* going on with Armor

Postby yurch » Wed Aug 12, 2015 6:47 pm

cliffski wrote:1) 50% boost to all armor penetration values
2) 25% boost to all armor damage absorbables.

Since armor absorption indirectly determines resistance, at this point we are just confusingly pushing the thresholds around.

Wouldn't a more straightforward buff to absorption be cutting weapon armor damage percentages instead?

User avatar
cliffski
Positech Staff
Positech Staff
Posts: 7968
Joined: Sat Feb 11, 2006 10:27 am
Location: UK
Contact:

Re: What is *ACTUALLY* going on with Armor

Postby cliffski » Wed Aug 12, 2015 6:49 pm

you mean any pulse weapons? I'm only talking about changing armor penetration, and as those weapons already penetrate all armor, it shouldn't matter., My concern is that there are not *enough* weapons to penetrate armor tanks at the moment.

yurch: Yup this is a very good point!. So the solution would be to boost weapons armor penetration but reduce weapons armor damage percentages agreed?

Xinxspuz
Line Manager
Line Manager
Posts: 30
Joined: Sun Jun 07, 2015 11:06 pm

Re: What is *ACTUALLY* going on with Armor

Postby Xinxspuz » Wed Aug 12, 2015 7:16 pm

Cliff, the problem with Pulse Cannons is that they already do everything.
  • They penetrate shields.
  • They penetrate any realistic level of armor you can put on a ship.
  • The have a high tracking speed, so they can hit ANYTHING, even fighters.
  • They have a decent range.
  • They have a high DPS.

By raising armor values and keeping Pulse Cannons where they are now, you're helping to fix that problem with Pulse Cannons, which I think pretty much everyone on here agrees is one of the most glaring problems with GSB2 right now.
If you want weapons to be able to deal with high armor values, great, raise the armor penetration on SOME weapons: Beam lasers, Particle Accelerators, Pulverizer Beams, Fusion Beams, and possibly Torpedoes.
Not Pulse Cannons. Definitely not Sledgehammer Pulse Cannons.

Xinxspuz
Line Manager
Line Manager
Posts: 30
Joined: Sun Jun 07, 2015 11:06 pm

Re: What is *ACTUALLY* going on with Armor

Postby Xinxspuz » Wed Aug 12, 2015 7:22 pm

Cliff, a good analogy here is a scientific experiment:
If you're trying to determine what effect something has, you change one variable at a time, and run a new experiment. You don't change 15 variables at the same time, and you definitely don't change a bunch of variables that are already deeply interrelated. All that does is muddle the results.

You like to publish comprehensive updates to your program, I get it. It takes time and resources to push through an update, across multiple platforms, for different OSes. YOUR time. YOUR resources. But when you change a bunch of related things all at the same time, people can't tell what helped fix a problem, and what just made it worse.

I pointed out a bug in the way armor was being calculated. Great! You should fix it. Fix the formula. Change the stacking percentages so the new formula works a little better. Don't suddenly change armor values and armor hit points, too. Definitely don't go muddling around with the Armor Penetration of weapons, as long as you have at least one or two weapons that will penetrate the highest Armor Value you can POSSIBLY generate on any hull with the new formula.


Return to “GSB2 Gratuitous Discussion”

Who is online

Users browsing this forum: No registered users and 2 guests