Engineers Shield + Booster Mod Calculator

[Herbrand]

If your shield is sufficiently resilient, for PvE there are zero reason to have any Heavy Duty ones. With 4 pips to sys, even if the added resistance is < 2%, that 2% is still around 10% less damage taken.

Since your shields are essentially never going to be broken, more resistance is the way to go.

 

[Arguendo]

Yes, back in Beta 2.1, when FDev first introduced the expression of resistances as positives, I took this up with the Developers. See my all weapon stats thread. I've been assured via PM (and the Devs have been consistent in saying, across several posts) that what the positive resistances actually mean is 100 - resistance = damage.

So a ship with +50% resistance will take 100 - 50 = 50% damage.

Thanks for clearing that up and confirming what the others were saying. That type of math seriously does my head in though! Incredible way to display the effect

Concerning what Cmdrs Frenotx and Sundae were talking about above with the Corvette guy, the point of course is that after three resistance boosters, resistances barely move, meaning that on almost all ships it is better thereafter to fit heavy duty.

From the calculator I have found that after 2 resist boosters the difference is negligible or drops for Thermal and Kinetic, but makes a difference for total strength. This is why I use 1 or 2 resist, and the rest heavy duty. Guess I am not too far off with the numbers, I just didn't understand the exact reason why

Thanks to all three of you for teaching me something new. Always useful.

 

[CMDR-AD]

Thanks to Frenotx for this tool. I dropped into this forum looking for just this kind of guidance.

 

[Frenotx]

Thanks to Frenotx for this tool. I dropped into this forum looking for just this kind of guidance.

No problem. Glad I could help.

 

[goemon]

coming back to this, after arriving back in the bubble:

question - in my understanding, reinforced shields increase "optimal strength" - doesn't play hullmass and optimal shield mass a role in this? looks to me like the calculate calculates that without taking hullmass into account?

I'd have to double-check, but I think due to the nature of how shield strength is calculated, a % increase to the optimal strength will result in an equivalent % increase in shielding.

now, here are screenshots of my cobra mkIII with a reinforced 4A shield generator (stating optimal strength +22,4%) and the values of a unmodified 4A shield generator (stating i would loose 18% shieldstrenght). the values shown (151,8 and 124) don't match any of both...

hullmass of a cobra mkIII is 180T, optimal mass of a 4A shieldgenerator is 285T. unfortunately i left my conda in beta at jaques - otherwise that would be the perfect ship to test whether hullmass/optimal mass/optimal strength plays a significant role in all of this.

or i'm simply getting the ingame stats totally wrong, i'm never sure!

 

[Arguendo]

now, here are screenshots of my cobra mkIII with a reinforced 4A shield generator (stating optimal strength +22,4%) and the values of a unmodified 4A shield generator (stating i would loose 18% shieldstrenght). the values shown (151,8 and 124) don't match any of both...

The loss and gain is a back and forth game with percentages; 120 x 122.4% ~ 147 and the opposite: 147 x 82% = 120.54.

So the numbers match.

 

[Cmdr Ultra]

Thanks Frenotx for this awesome tool. I tried to use it but it gave me a headache.

Still Cmdr's like you make this community better. I salute you O7.

 

[goemon]

The loss and gain is a back and forth game with percentages; 120 x 122.4% ~ 147 and the opposite: 147 x 82% = 120.54.

So the numbers match.

thanks for this!

makes sense.

 

[Cmdr. Numa]

Any chance for a similar sheet for armor ? The shields one is simply awesome.

 

[GW42]

snip...

Hi Frenotx - have you thought about combining your calcs with Coriolis? There's some helpful chaps who've taken up the challenge of updating Coriolis as apparently the original (fantastic) author is no longer able to support it. They've got a thread here: https://forums.frontier.co.uk/showthread.php/296769-Coriolis-EDCD-Edition

I've taken the liberty of posting a link to your thread over there as they have already said they'd like to expand Coriolis to deal with engineers stuff, and your calcs would help a lot with that. Hope you don't mind

Oh, and great work!

 

[taleden]

Very interesting! Thanks for making this -- would you mind if I try to incorporate these equations into edshipyard?

Also, regarding diminishing returns:

I believe hull resistances use the same formulas. The formula works as follows: First, convert all the resistance values to damage multipliers. -40 becomes 140% (1.4), +20 becomes 80% (.8), etc. In other words, 1 -
[listed value / 100]. You then take the base armour damage multiplier, and multiply it one-by-one with all the HRP multipliers. [Base armour resistance] * [HRP1 multiplier] * [HRP2 multiplier], etc. Once you've done that, you should have you pre-diminishing-returns value. Convert this back to the regular format for readability by doing this: (1 - [prediminishingreturns value]) * 100

If your answer to that is 50 or less, then you're done. If it's over 50, then you need to apply diminishing returns. First, subtract 50 from your total. Next, divide the result by 2. Finally, add the 50 back. Essentially, every point over 50 only counts for half a point. In other words, the 50-100 values are all remapped to 50-75.

Example: Calculating thermal resistance with regular armour (base thermal resistance 0), and three HRPs that have a thermal resistance of 25 (possible G3 thermal resistance mod).

Damage multipliers:
Armour: 1; HRPs: .75

1 * .75 * .75 * .75 = pre-diminishing-returns multiplier= ~ .42

Pre-diminishing-returns value = (1 - .42) * 100 = 58
That's over 50, so we have to apply diminishing returns. Each point over 50 only counts for half a point.

Final value = [(58 - 50) / 2] + 50 = 54

So starting with neither a resistance to, or weakness to, thermal damage, adding three "25 thermal resistance" HRPs bring the armour to 54 thermal resistance (not the 75 you'd get if all things were additive).

Can you point me to the source for this "half-credit past 50% resist" mechanic? Was that explained by FDev somewhere, or derived from experimental research?

 

[Codger]

Can you point me to the source for this "half-credit past 50% resist" mechanic? Was that explained by FDev somewhere, or derived from experimental research?

https://forums.frontier.co.uk/showt...-Calculation?p=4230114&viewfull=1#post4230114

 

[Ziljan]

Thanks Frenotx!

I've been recalculating this by hand each time I went to the engineers. The only bit I was missing was the post 50% diminishing returns calculation. Is it really that simple? 50 +(L3-50)/2?

 

[Frenotx]

Any chance for a similar sheet for armor ? The shields one is simply awesome.

If I get some free time, I'll see what I can do. As far as I can tell, the formulas are all the same, there's just a little more interface stuff to do, given the various class HRPs.

Hi Frenotx - have you thought about combining your calcs with Coriolis? There's some helpful chaps who've taken up the challenge of updating Coriolis as apparently the original (fantastic) author is no longer able to support it. They've got a thread here: https://forums.frontier.co.uk/showthread.php/296769-Coriolis-EDCD-Edition

I've taken the liberty of posting a link to your thread over there as they have already said they'd like to expand Coriolis to deal with engineers stuff, and your calcs would help a lot with that. Hope you don't mind
Oh, and great work!

I love coriolis and use it all the time, so I'd certainly enjoy having this stuff integrated. Let me know if you have any questions, or need any help.

Very interesting! Thanks for making this -- would you mind if I try to incorporate these equations into edshipyard?

Same answer as before- I don't mind at all. Let me know if you have any questions, or need any help.

Thanks Frenotx!

I've been recalculating this by hand each time I went to the engineers. The only bit I was missing was the post 50% diminishing returns calculation. Is it really that simple? 50 +(L3-50)/2?

Yep. As per the devs, the values 50 -> 100 are all remapped to 50 -> 75. The process for calculating that is as follows:

1) If the value is below 50, leave it alone. End here.
2) If it's above 50, subtract 50 from it. That first 50 isn't subject to diminishing returns, so can be just set aside.
3) Take the remainder, and divide it in half. This is the "remapping".
3) Add the remapped part back to the original, un-touched 50

That entire process simplifies down to:
If preDiminish > 50, then finalValue = [(preDiminish - 50) / 2] + 50

 

[taleden]

Yep. As per the devs, the values 50 -> 100 are all remapped to 50 -> 75.

From what I could see in the linked thread on that subject, the way you've stated it here is really all we know for certain: 50->100 is remapped to 50->75, but they didn't spell out exactly how. That makes me wonder if it's really a linear remapping or if it might be curved, which I think would make a whole lot more sense. That is, the next 1% resist past 50% might be worth nearly-but-not-quite 1%, but as you approach 100% raw resist, the value of each 1% added approaches zero.

That would be more intuitive to me given the number of other places FD has used various sorts of power curves (FSDs, shields, thrusters, heat). The counterpoint is that in the same thread they gave an example which appears to use the math as you've laid out (straight linear penalty), although it wouldn't be the first time an FD rep quoted incorrect/outdated information on the forums (it was famously claimed that thrusters used linear interpolation, which is plainly not true).

So a bit of experimentation may be called for, which I'll put on my list if I ever get the time. Regardless, thanks much for your research on this and other topics, it's very much appreciated!

 

[Frenotx]

From what I could see in the linked thread on that subject, the way you've stated it here is really all we know for certain: 50->100 is remapped to 50->75, but they didn't spell out exactly how. That makes me wonder if it's really a linear remapping or if it might be curved, which I think would make a whole lot more sense. That is, the next 1% resist past 50% might be worth nearly-but-not-quite 1%, but as you approach 100% raw resist, the value of each 1% added approaches zero.

That would be more intuitive to me given the number of other places FD has used various sorts of power curves (FSDs, shields, thrusters, heat). The counterpoint is that in the same thread they gave an example which appears to use the math as you've laid out (straight linear penalty), although it wouldn't be the first time an FD rep quoted incorrect/outdated information on the forums (it was famously claimed that thrusters used linear interpolation, which is plainly not true).

So a bit of experimentation may be called for, which I'll put on my list if I ever get the time. Regardless, thanks much for your research on this and other topics, it's very much appreciated!

Considering shield strength uses a linear function (actually two linear functions), it's not that much of a stretch that the remapping is linear. Usually, when I hear someone say X -> Y was remapped to X -> Z, I assume they mean linear unless otherwise specified. Furthermore, considering the nature of multiplicative boosts, the extra complication of another curve is somewhat redundant. I can attempt to get some confirmation though, either via dev comment or experimentation, but it may take a while.

 

[taleden]

Considering shield strength uses a linear function (actually two linear functions), it's not that much of a stretch that the remapping is linear. Usually, when I hear someone say X -> Y was remapped to X -> Z, I assume they mean linear unless otherwise specified. Furthermore, considering the nature of multiplicative boosts, the extra complication of another curve is somewhat redundant. I can attempt to get some confirmation though, either via dev comment or experimentation, but it may take a while.

That's my point actually -- FDev has said shields use a linear function, but they actually don't if you look at the data. I haven't quite worked out the right formula yet but it's definitely some kind of exponential curve and not a straight-line interpolation. It's easiest to see this discrepancy at mass points halfway between min-opt or opt-max.

So in this case, if we can't convince FDev to confirm the mathematical details, I guess we'd have to fit a ship with 75% raw resist to some type, hit it with a bunch of shots of that damage type (ideally from something whose damage stat is easily verified to be exact and not rounded in the in-game display), and see if it takes the expected number of shots per % shield/armor lost.

 

[Frenotx]

That's my point actually -- FDev has said shields use a linear function, but they actually don't if you look at the data. I haven't quite worked out the right formula yet but it's definitely some kind of exponential curve and not a straight-line interpolation. It's easiest to see this discrepancy at mass points halfway between min-opt or opt-max.

So in this case, if we can't convince FDev to confirm the mathematical details, I guess we'd have to fit a ship with 75% raw resist to some type, hit it with a bunch of shots of that damage type (ideally from something whose damage stat is easily verified to be exact and not rounded in the in-game display), and see if it takes the expected number of shots per % shield/armor lost.

If I'm not mistaken, shields actually use a piecewise linear. It's linear with one slope from max mass to optimal mass, then linear with another slope from optimal to minimum. Has someone done research that counters this? I'd certainly like to see it, as I have an unhealthy interest in figuring out / understanding the nitty-gritty details of the mechanics of this game.

 

[taleden]

If I'm not mistaken, shields actually use a piecewise linear. It's linear with one slope from max mass to optimal mass, then linear with another slope from optimal to minimum. Has someone done research that counters this?

Yes; here's my most recent attempt to refine the shield formula, which includes some graphs demonstrating that it is definitely not linear:
https://forums.frontier.co.uk/showt...-s-too-heavy?p=2770174&viewfull=1#post2770174

That's the formula currently in use by edshipyard and coriolis (as far as I'm aware), and it's pretty close but still not perfect, especially now that the in-game outfitter gives us the shield strength to the nearest 0.1.

 

[goemon]

i have another question (preparing my 2.2. modding...):

from my understanding, if you go for thermal-resistance shields and resistance augmented boosters, wouldn't it be better to start with a L4 (!) thermal resistant shield?

you would start with ~28% thermal resistance, and ~28% kinetic resistance, so that would be more balanced?

(L4 thermal shield: +13,2% - +48% thermal resistance, -3,6% - -12% kinetic resistance).