Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Goumindong]

How did you calculate armour survival time? I'm attempting to recreate the results you received and I'm getting slightly different values...
Another thing worth mentioning is that while relative increase gets smaller, absolute number of hits needed to get through armour increases.

Recursively by calculating the damage each shot does to armor, subtracting that damage from the armor, counting the number of shots to take armor to zero and then multiplying the number of shots minus one* by the number of shots per second.

*This is because the first shot comes at time== zero.

So i was fitting out a conquest the other day and was trying to make it good and was failing. So i went and changed tactics. Instead of even half attempting to get its shield to a reasonable level i slapped heavy armor on it. And it worked a lot better and i felt dumb for not having seen it before.

The reason for this is obvious, the less armor you have the more valuable that heavy armor is. And similarly the worse your shield the less valuable that attempting to shield tank is. And, similarly, better the higher armor damage being done(to a point) I did some basic armor value calculations and came up with some handy tables

Did you mean the more armor you have the more valuable that heavy armor is?

No. The less armor you have the more valuable heavy armor is as a percentage of survival time

http://fractalsoftworks.com/forum/index.php?topic=12268.0

Armor here is outdated not the least of which because iirc, one of the calculations for skills isn't working as it should (IIRC) and because the weapon values have changed.

This is interesting, thank you for posting it. Did your % increase calculations include the effects of minimum reduction? It would also be interesting to see the effects of Heavy Armor on effective hull strength (as more base armor means more hull damage reduction, means more effective hull points).

This depends largely on the type of weapon hitting armor. Heavy armor on a conquest being hit by Mark IX autocannons only reduces damage by about 12.5%. But it reduces damage on Storm Needler by 23%.

However, I think that "% increase in survival" is a bit of a misleading statistic - I would be more interested in 'shots to break through' before and after, or 'added shots', the difference between the two. As an example of why I want this, consider the case of heavy armor taking a frigate's armor from 100 to 200: the percentage increase in toughness is huge, but the absolute toughness is still pitiful and the absolute added toughness is small. This is a general rule: the higher the base armor, the higher the absolute toughness added by Heavy Armor. In the example despite the large % increase, taking any but the weakest hits on the armor is a bad idea because the absolute toughness is still too low. The Conquest isn't such an extreme example, as 1200 armor is pretty good against common threats, but in absolute terms the increase is going to be much less for it than for an Onslaught.

I didn't do that because the numbers are kinda messed up by the inability to perfectly model accuracy. That is, we only really have a good way to simulate a hit hitting the exact same spot each time. And when that happens the numbers are pretty correct. But armor doesn't work like that because you never hit the exact same spot right after each other. So saying "it takes 8 seconds for an HAG to kill 1200 armor and 12 seconds to kill 1600 armor" doesn't give you a full accounting if you think you're only going to get "4 extra seconds". Because in effect the HAG spreads damage out and really takes 40 seconds to take down armor because it hits 5 different armor cells(or misses some). And so you go from 40 seconds to 60 seconds. Which is a significant increase

The main thing that makes me a bit hesitant on relying on pure armor tanking for the Conquest  (for player piloting) is the distribution of damage sources in live play though: HE damage is delivered in large spikes that really needs to be stopped by the shields, and Hardened Shields is the only way to help stop these bursts (well, that and IPDAI). The effective HP increase of HS is enough to block 9 additional Atropos (without skills on either side). Of course the ship is going to be hit while having flux, but even being able to stop 3 more Atropos is the difference between a completely failed strike and one that does heavy armor damage. Not saying Heavy Armor is bad, but the Conquest's shield is weak enough that it really needs Hardened Shields in my experience.

This would be true if you regularly had to tank >23 Atropos at a time onto your shields before you could vent. But well, its more efficient to shoot them down and its more efficient to only block the atropos on shields than also block kinetic damage (which would do significant armor damage without Heavy Armor)

The main reason to put Hardened Shields on Conquest is for AI protection.  With very bad shields, the ship in vulnerable to overloading from big spikes of damage to the shield.  At least in previous versions (not sure if AI still behaves this way), AI will try to prevent flux from getting too high, but with terrible shields, it can jump straight from what AI thought was high yet safe to outright overload.  Hardened Shields helped fix that AI fault.

Yea, this was for personal use. The AI used to be better at protecting itself with shields and now it only is good at letting shots through if its close to overfluxing.

I was confused about the part where it was specifically stated that 600 armor Conquest with Heavy armor was more effective than normal armor conquest....

It increased the value of heavy armor because 600->1000 armor is a larger % increase than 1200->1600.

Something doesn't seem right. I have a hard time believing, for instance, that a Frigate with 200 armour, now with heavy armour so now having 300 armour, suddenly has +100% survivability i.e. twice the time to kill it with a Hephaestus Assault Gun. It only takes about 3 shots to deplete the armour in either case and even a kite requires 9 shots to kill, so +100% survivability cannot possibly be true.

Generally speaking, the higher the armour value, the higher the damage the shots the ship can take before armour is depleted. So the higher the armour, the better heavy armour mod should be.

Goes from 2 shot to strip armor to 3 shots to strip armor. 3 shots has a total duration of .5 seconds of firing because the first shot comes at time=0. 2 shots has a total duration of .25 seconds firing. Note that these values don't actually mean that much unless those shots hit the same armor square.

[intrinsic_parity]

http://fractalsoftworks.com/forum/index.php?topic=12268.0

Armor here is outdated not the least of which because iirc, one of the calculations for skills isn't working as it should (IIRC) and because the weapon values have changed.

This was made during 0.8 which is only one major update ago. Nothing about the armor mechanics has changed. It's possible that some weapon damage values have changed but that has nothing to do with how the mechanic works generally. I'm also not aware of any issues with skills not working in this analysis. Even if that were true, the armor skills just modify values in the armor calculation (which results in stuff like scaling and shifting of graphs, not major changes in trends/behavior), they don't change how the calculation is done, so that would not change the general mechanic. There was a major thread that led to the creation of this guide (which I started). I've done the calculations myself and I can confirm that it is at least generally accurate, even if some numbers might be outdated.

The TLDR is that weapons that do high damage per shot strip armor much faster and armor effectiveness increases non-linearly as damage per shot decreases (or as armor increases). There's some other interactions with the damage reduction limits and stuff, but that general rule of thumb is always true regardless of the details. The effectiveness of increased armor will be very different vs. different weapons. I can guarantee you that is true regardless of minor inaccuracies in numbers.

[Goumindong]

If you don't believe me just takes FOOF's word on it. His spreadsheet is outdated and defunct by his own estimation. He says so at the top of the sheet.

The rest of it is not relevant to the purpose of this tread, which is to compare the effect of heavy armor not discuss how armor works in general.

[SapphireSage]

I think the main reason his spreadsheet is outdated is because it can be used to calculate very many things such as shots, time, flux to break armor and had to do with a lot of raw, hard numbers of the weapons and ships which are constantly shifting.

But, the underlying mechanics and theories should definitely be very true with regards to how armor works and is improved with respect to importance of damage per shot and such.

@intrinsic_party I believe Goumindong is referring to the fact that in 0.8 Impact mitigation didn't increase armor by 150 flat amount(for calculations) but erroneously increased armor by 150% more than doubling the armor of ships(for damage calculations).

[intrinsic_parity]

Thank you SapphireSage for clearing that up. My point was that heavy armor impacts different weapons very differently, and you've only used a very narrow subset of weapons (with pretty average damage per shot) in your comparisons.

Your metrics are also not very good representations of how armor impacts general survivability, increase in survival time (in seconds) would probably be much more useful. Percentage increases are generally pretty misleading. Doubling a half second TTK to a 1 second TTK is a 100% increase while being almost entirely irrelevant (.5 seconds is barely more than human reaction times) while a 20% increase in a 100 second ttk is a full 20 seconds which is very significant.

I agree with your conclusion about armor tanking as much as possible, but it's right regardless of how much armor you have. Heavy armor just lets you tank for more/longer, it doesn't change the fact that armor tanking low damage per shot weapons (particularly kinetics) helps win the flux battle. I utilize that strategy with every ship I fly. The actual time in seconds that you buy with heavy armor would be a much better indicator of how effective the armor is at achieving that.

[Goumindong]

Your metrics are also not very good representations of how armor impacts general survivability, increase in survival time (in seconds) would probably be much more useful.

No. It would not, because no such data can reasonably be compiled. As already explained

As an example. the "Survival time" for the armor on a 1200 armor ship against HAG's is "4.5 seconds". Have you ever seen a conquest have its armor stripped in 4.5 seconds from HAGs? No. Because even though HAG's are reasonably accurate they're not that accurate. There are about 100 armor cells* on a battleship. And if you damage these at the same point you will strip 20 cells of their armor roughly at the same pace you will strip the theoretical single cell we're talking about. And if you spread it around you will damage the armor at up to 1/5th the speed (as you damage entirely separate parts of the ships rectangle). So the HAG survival time is anywhere between 4.5 seconds and 23.5 seconds** and the Heavy Armor survival time is anywhere from 7.25 seconds to 34.75 seconds. Assuming everything hits of course. And you don't attempt to tank the HAG on shield. For the Mark IX the duration is anywhere from 15.57 seconds to 77.85 seconds. That is a LOT of time and a lot of value you get out of not having to have your shield raised. For a conquest that mark IX number would be about 30,000 flux saved on raising the shields(not including flux saved from tanking those shots) which is 30,000 (plus) extra flux worth of damage you would be putting out.

As for the raw time increases over specific values... well the majority of weapons that do not penetrate ALL of the armor in a duration that makes more armor inconsequential tend to hit initial armor values at minimum armor damage. The Mark IX hits minimum armor damage above 560 and its pretty high all things considered. The HAG minimum armor above 1360. Only the Heavy Mauler (2266) and the Hellbore (8500) are significantly above the minimum armor point on capital ships for ballistic weapons. And right around those values the margin on the increase isn't enough to break into extra shots***. As a result the time to survive on more or less all weapons where it matters is fixed. A conquest gains ~400 armor at minimum armor damage and an onslaught that gains 400 armor gains 400 armor at minimum armor damage.

*guesstimate: i don't know the actual value. Its probably between 70 and 100. I counted but the cells shown on a ship are not quite indicative of the actual total cells a ship has and should underestimate it.

**4.75 normalized seconds * 5 - shot cycle.

***As an example a HAG does 4 extra damage to armor at 1200 vs 1360 armor. In order to get from 1600 to 1200 it takes 11 shots(like.. 10.99 or summat). In order to get from 1800 to 1400 it takes... 11.111 shots. The extra shots are a rounding error that you will notice only when you get over 2200 armor.

[TaLaR]

Conquest has fairly vulnerable guns - they are roughly at centers of it's sides. Exactly where AI auto aims.
Odyssey is even more fragile, with only 1 gun side. A few good hits can incapacitate it. TLs, HVDs, Ion beams - all of these can disable guns quickly and from afar, even with RFC.

TLs are particularly important - they are easy to shield-tank (and Hard shields are very important to make doing so more bearable), but no armor can stop them in significant way (full Onslaught gets armor-stripped and badly damaged from single unmitigated salvo of 4x TL paragon). Capitals need to be build with countering TLs specifically in mind.

Then there is AI problem. AI will use shields, you can't force it to armor tank before it already builds up too much flux (wasting what could go to weapons).

[Sinosauropteryx]

Really cool thread, and I find myself using it often ever since you posted it.

There is one thing slightly off about the calculations in regards to D-mods, though. The armor-reducing D-mods also reduce the bonus armor granted by Heavy Armor. So the 600-armor Conquest with -30% armor D-mod would only be getting +280 armor from HA. It's not the same boost as having a capital with a base armor of 600 and taking HA.

[Goumindong]

Really cool thread, and I find myself using it often ever since you posted it.

There is one thing slightly off about the calculations in regards to D-mods, though. The armor-reducing D-mods also reduce the bonus armor granted by Heavy Armor. So the 600-armor Conquest with -30% armor D-mod would only be getting +280 armor from HA. It's not the same boost as having a capital with a base armor of 600 and taking HA.

Thanks for letting me know i didn't even notice that

[Goumindong]

@Alex

IF you're reading this; how many armor cells do ships tend to have anyway?

[intrinsic_parity]

Your metrics are also not very good representations of how armor impacts general survivability, increase in survival time (in seconds) would probably be much more useful.

No. It would not, because no such data can reasonably be compiled. As already explained

As an example. the "Survival time" for the armor on a 1200 armor ship against HAG's is "4.5 seconds". Have you ever seen a conquest have its armor stripped in 4.5 seconds from HAGs? No. Because even though HAG's are reasonably accurate they're not that accurate. There are about 100 armor cells* on a battleship. And if you damage these at the same point you will strip 20 cells of their armor roughly at the same pace you will strip the theoretical single cell we're talking about. And if you spread it around you will damage the armor at up to 1/5th the speed (as you damage entirely separate parts of the ships rectangle). So the HAG survival time is anywhere between 4.5 seconds and 23.5 seconds** and the Heavy Armor survival time is anywhere from 7.25 seconds to 34.75 seconds. Assuming everything hits of course. And you don't attempt to tank the HAG on shield. For the Mark IX the duration is anywhere from 15.57 seconds to 77.85 seconds. That is a LOT of time and a lot of value you get out of not having to have your shield raised. For a conquest that mark IX number would be about 30,000 flux saved on raising the shields(not including flux saved from tanking those shots) which is 30,000 (plus) extra flux worth of damage you would be putting out.

No one is arguing that the situation isn't more complicated than a single armor cell model, but using percentage increase rather than survival time doesn't help that problem, it just obfuscates the information further. Those time ranges you just cited are much more useful information than some percentage increase (I assume you got them from multiplying by 5, but you appear to have slightly miscalculated a few). The percentage increases are all relative to different initial values so direct comparison is not very useful. Survival times (or number of shots to penetrate) are all the same metric so the comparison is much more informative. Even if the numbers are not perfectly indicative of the true situation, they still give you better information about how much value you are getting.

I already gave examples of how you can see a very high % increase that correspond to a mostly irrelevant amount of time (i.e. small amount of value). Percentage increases aren't good metrics when they are relative to different values, the results don't reflect the amount of value you are actually getting.

As for the raw time increases over specific values... well the majority of weapons that do not penetrate ALL of the armor in a duration that makes more armor inconsequential tend to hit initial armor values at minimum armor damage.

It is a good point that increasing armor past the damage reduction limit becomes a linear increase in value, but that means that means that increase of the same amount give the same amount of value (the same increase in survival time approximately), which is further evidence that the percent increase is a bad metric. For instance going from 1400-->1600 and 1600-->1800 armor vs HAG has (within rounding error) the same increase in number of shots (and survival time) but the %increase will be different because the survival times at 1400 and 1600 are different. I think shots to destroy one armor cell is the clearest metric (and the one that got used in past threads). The effects of inaccuracy can be considered separately.

[intrinsic_parity]

I'm actually also very interested in how many armor cells things have (and how big the cells are). If the cells are different sizes, that could make a really big difference.

[Goumindong]

Those time ranges you just cited are much more useful information than some percentage increase (I assume you got them from multiplying by 5, but you appear to have slightly miscalculated a few)

I corrected for the zero time fire rate assumption and then multiplied by 5 and then refixed the zero time fire rate assumption. It was explained in the footnote/comment

It is a good point that increasing armor past the damage reduction limit becomes a linear increase in value, but that means that means that increase of the same amount give the same amount of value

No. Assume for a second that you have 10,000,000 armor. Going from that to 10,000,400 does not produce the same amount of value as going from 1200 to 1600. The early seconds you add on survival are more valuable than the latter seconds. Its true that it may flip at some point. But its unlikely you're at that point with almost any ship that actually gets hit.

[intrinsic_parity]

It is a good point that increasing armor past the damage reduction limit becomes a linear increase in value, but that means that means that increase of the same amount give the same amount of value

No. Assume for a second that you have 10,000,000 armor. Going from that to 10,000,400 does not produce the same amount of value as going from 1200 to 1600. The early seconds you add on survival are more valuable than the latter seconds. Its true that it may flip at some point. But its unlikely you're at that point with almost any ship that actually gets hit.

That is because the battle is a finite amount of time, so increasing survival time beyond the length of the battle has no value. If the battle were infinitely long, then going from 10000000 to 10000400 would have the same value as 1200 to 1600. Clearly the ship with 10000000 armor would be far better at surviving, but the increase in capability would be the same. Another example, going from .001 seconds of  survival time to 1 second is an increase of 1000% but it is clearly not 10 times more valuable than going from 10-->20 seconds, an increase of 100%. I would also argue that going from 10-->20 seconds is clearly far more valuable than goin from 1-->2 seconds even though they are the same percent increase. It's not a good metric, and it doesn't give a valuable measure of how much survival capability you are gaining.

If you have survival times on the order of 1 second vs average damage values, you basically can't tank any damage and your armor is just giving you a bit of a margin for error in dodging and shield tanking damage which is your actual defensive strategy. If you are perfect at dodging, then your armor gives you no value at all. Going up to 2 seconds does not change that, your margin for error is just slightly bigger. You didn't gain some huge amount of value like the 100% increase from 1-->2 seconds suggests. Going from 50 seconds to 75 seconds is by your metric half as valuable as going from 1-->2 seconds (or 5--> 10 seconds), but clearly gaining 25 seconds of time armor tanking damage is going to save you A LOT of flux and probably win you multiple duels. The amount of benefit you are gaining is very significant in comparison.

[Goumindong]

But there are not survival times that low in the relevant range. That is, againt similar weapon types which you might be expected to brawl against. And extra armor increases hull survival times as well

Either way the percentage estimation is valuable