CMaster
11-10-12, 21:57
I've posted already about how the current armour system is broken, a combination of old items and new ideas in an amorphous, unjoined up lump. I received critcism there that I didn't really do enough in demonstrating how the current system is broken rather than just silly and there was a lack of suggestions as to what would be better.
This posts details a variety of failings of the current armour and resist system, how these failings let down the class structure and particuarly hurt PEs. There is then a proposal for a new resistance and armour strcuture, that without dramatically departing from the current state of the game would solve a lot of these issues, and remove a lot of the nonsense.
Introduction to some important concepts around damage mitigation
This post is not going to shy away from the mathematics of the situation, because understanding those is key to understanding what will and won't work. It also seems to be something that whoever designed the current system didn't really understand.
An important but not entirley inutitive element of this is that resists do not scale linearly. Going from 10% to 20% resistance (or more importantly 90-80% damage) is a small change. Going from 90% to 95% is a huge change, which will result in a character being able to soak up twice the damage (ignoring healing factors). In practicality, it's the latter issue we want to look at, to understand the situation appropriatley. The important parameter to look at from a balancing point of view is lifetime - how much longer does the mitigation let a character live than they would without any. From this we can see that 90% mitigation would lead to a 10 times longer life, and 95% to one that is 20 times longer.
Why this all matters
Well, the amount of punishment that a character can take is a big part of how effective that character can be in both PvE and PvP. It's a big part of what makes a class viable, and also what can make them overpowered.
Also, with every class having a total of 300 levels, spread across the five main skills, an implicit part of this system is that a class's strengths in one area lead to weakness in another. For this to hold true, all mainskll have to matter to all classes - this includes strength and constitution. Right now, there is a case to be made that stregth really doesn't matter to anyone except tanks.
How things are at the moment - numbers
When calculating damage, each different form of damage mitigation (sheild, psi resist, resist, armour) is applied seperatley. If the total mitigation is greater than the mitigation limit for the class, then damage is brought back up to that limit. Total damage remaining can be easily calculed by multiplying the damage factors (ie 100% - mitigation %) of each mitigation togheter
The numbers used here for existing values for armour and resist mitigation are taken from Neoskiller, which uses a reverse engineered resist calculation. I can't promise it is entirely accurate, but the shape of the curves certainly fits with what I saw in testing long ago. Below are the resistances given per point for armour and con skills at the moment in NC2.2:
http://www.cmaster.linxsoft.co.uk/images/forumstuff/NCCurrentResist.png
Not exactly a pretty picture, with weird bumps and rather abrupt changes in slope. To understand a bit more what this means, let's look at the lifetimes graph (y axis is the lifetime muliplier, ie 2 means last twice as long), below:
http://www.cmaster.linxsoft.co.uk/images/forumstuff/NCCurrentLife.png
So what we can see is that despite the changes in slope, more resist points actually get increasingly effective as you go along. Equally, the "bumps" between slope changes lead to some weird behaviour in the resist section. Armour actually works out relativly flat and reasonable, eventually giving about 50% life extension.
It seems that improving the effectivness of resists with increasing points spend is not entirley desirable, and the link between armour and resists, and what these points values mean, is pretty obscure to the player actually in the game.
How this could be better - numbers
While a little mystery can be exciting, it strikes me that players should be able to understand what a "point" of armour or resistance really means. It would also be good to have a nice, stable increase in lifetime throughout the range. To this end, I suggest introducing a new armour/resist calculation:
http://www.cmaster.linxsoft.co.uk/images/forumstuff/ncresistcalc.png
Where D is the factor that incoming raw damage is multiplied by to give final damage, A is armour or resist points value, C is an adjustable factor, and R is the resistance value that players tend to talk about (D = 1 - R).
The prescence of C as a tuneable constant is important, as it lets us decide just how influential we want armour and resists to be. The reason for picking this function is quite simple - it gives a linear increase in lifetime with respect to investment of points. Using a factor of 0.01 (as is used for later examples), every point in armour/resist gives a 1% increase in liftime. So for 20 points, you'd last a fifth longer than somebody without any. With 100 points, twice as long (and 2/1.2 = two thirds longer than the guy with 20 points).
Graphically, this looks something like this for resistances:
http://www.cmaster.linxsoft.co.uk/images/forumstuff/NCNewResist.png
and this for lifetimes:
http://www.cmaster.linxsoft.co.uk/images/forumstuff/NCNewLife.png
Notice how lifetime now increases by a nice, consistent amount in return for investing the points/ getting the armour.
Practically, I'd say 1% per point is too much (it's used here because of similar values to existing situations). 0.5% per point produces a more reasonable picture that might be approriate to start off for both armour and resist - they can be diversified if needs be.
"But" some may say "Shouldn't overinvesting be discouraged? This linearity looks like the opposite". Except that the game already punishes overinvestment - in resist, those first 50 points to get your lifetime up to one and a half of what it was without takes only a single skill point each point. To raise it then up to 75% will cost you more. In armour, the same thing happens - the higher up the TLs you go, the less additonal points you see per TL (and this can be refined of course).
How things are at the moment - armour
It shouldn't really be necessary to provide a recap of the current armour system - for those who need it, take a look at the item DB (http://www.techhaven.org/db/). Some key issues to draw out: belts still have a whole set of int-based belts, letting Spies and monks boost their armour in one stat heavily for the torso only. Spies (and monks) can use high tl int-bsaed underwear, which gives better stats than the mid-tier str or int based that PEs can use. Monks have their complete own set of armour, that gives good resists for pretty much every damage type except poison. The fact that there are only two armour slots per target zone, combined with the current armour-resistance curve means that armour that protects against multiple damage types is rather perferable to those that don't. Energy armour is only available from all-over pieces, and the tank-only Duranium.
So, the conclusions from this: carbon armour is a joke at the moment, delivering less posion than the more versatile kevlar - the boost in piercing and force is largley irrelevant considering how much other protection players will get against this. Duranium is tank only. Kevlar at the moment is pretty fantasic armour. One of the key things to note is that PEs gain very little in this whole space from their strength. The kevlar armour that spies wear is more useful to the than anything in the 40-65 category. They lack the int to use the really effective belts. Getting even bad duranium requires some serious gimpage, and the overalls available to PEs are wearker than those for spies or monks. Practically, PEs wind up using either spy armour or newbie monk armour. Meanwhile, monks with their cripplingly low strength get away scot free (other than low force resist/transport_). Spies get to dance around near enough as tough as a PE. Tanks remain happily built of bricks of course, with the best of almost everything available to them. Spies do have to flail a little for any reasonable energy protection, althoug in theory Holovests serve that very well.
How things could be better - armour
Here, there's a need to tear down a lot of elements that are there for old or bad reasons. To start, I'd see an end to all int-based armour - let class's strength be a genuine indication of their survivability. Remove the locational armour system - it's a neat idea, but the reality is that having to cove everything in a few slots really hampers setup viability, and makes one-offs he titan and viper king chest weird. Sort carbon armour out - the poison component needs to be competitive with kevlar - or perhaps better it could have x-ray added to the mix. Bring in a new line of energy armour, (call it Duranit like the old stuff maybe), in the same strength range of inq/carbon. So Spies could spec for the bottom tier as a choice, and PEs would have default access to middle tier and option to spec for top tier.
More dramatically, I'd like to see psi armour gone as well. It's stats remain absolutely fantastic (holy spirit has more energy than duranium, then adds xray and fire too!). Monks are meant to be weaklings, let's see them feel it. As a compensation, perhaps make some low tier (10-30 ish) "vaporisation" armour available - small amounts of energy/fire, maybe x-ray - at the cost of no pierce force whatsoever. This step needs to be considered a lot more cautiously than the others, but it would be nice to see monks who actually consider con and armour setups like other characters do.
The idea over all of this is that characters lacking in strength have to make harder choices about what they are doing with their strength, and at the same time would see benefits from imping/drugging to increase it. That PEs would actually have something to differentiate them from spies, other than being able to use worse weapons.
One thing that is also important to keep, is that the highest TL armour is 90 or so. Tanks shouldn't be able to bump up strength and see damage AND defense go up as a direct consequence.
Some notes on Power Armour
One of the general principles in Neocron is that setups trade damage for defense (in theory, they should trade off mainskill, weaponskill and defence, but for the moment the first two almost always go togehter). Power armour right now spits in the face of that idea. Higher levels of PA, for every class, give more armour, more mainskill and more weaponskill. This is wrong. The simple fix that works for this is simply to fix all PAs for a class at the same armour values (still allowing for camo to be different). It would be interesting to see PA do different things, but that's changing the game, rather than tweaking it.
Conclusion
Set out above is a plan to make the resistance more comprehensible to players, and more supportive of the class balance and differentiation. The two stages, adjusting the resistance calculation and adjusting the armour sets are related, but the implementation of one needn't require the other - I'd place much higher priority on adjusting the armour set than on tweaking the way things are calculated.
For those interested in the numbers part above, you can find the spreadsheet I did calculations on, with some factors that can be adjusted, here. (http://cmaster.linxsoft.co.uk/images/forumstuff/nc%20armour%20toying.xlsx) - You also see what a linear and an exponential resistance system would be like.
Anyway, thanks for those who bothered reading the whole way through, and open to comments we go!
This posts details a variety of failings of the current armour and resist system, how these failings let down the class structure and particuarly hurt PEs. There is then a proposal for a new resistance and armour strcuture, that without dramatically departing from the current state of the game would solve a lot of these issues, and remove a lot of the nonsense.
Introduction to some important concepts around damage mitigation
This post is not going to shy away from the mathematics of the situation, because understanding those is key to understanding what will and won't work. It also seems to be something that whoever designed the current system didn't really understand.
An important but not entirley inutitive element of this is that resists do not scale linearly. Going from 10% to 20% resistance (or more importantly 90-80% damage) is a small change. Going from 90% to 95% is a huge change, which will result in a character being able to soak up twice the damage (ignoring healing factors). In practicality, it's the latter issue we want to look at, to understand the situation appropriatley. The important parameter to look at from a balancing point of view is lifetime - how much longer does the mitigation let a character live than they would without any. From this we can see that 90% mitigation would lead to a 10 times longer life, and 95% to one that is 20 times longer.
Why this all matters
Well, the amount of punishment that a character can take is a big part of how effective that character can be in both PvE and PvP. It's a big part of what makes a class viable, and also what can make them overpowered.
Also, with every class having a total of 300 levels, spread across the five main skills, an implicit part of this system is that a class's strengths in one area lead to weakness in another. For this to hold true, all mainskll have to matter to all classes - this includes strength and constitution. Right now, there is a case to be made that stregth really doesn't matter to anyone except tanks.
How things are at the moment - numbers
When calculating damage, each different form of damage mitigation (sheild, psi resist, resist, armour) is applied seperatley. If the total mitigation is greater than the mitigation limit for the class, then damage is brought back up to that limit. Total damage remaining can be easily calculed by multiplying the damage factors (ie 100% - mitigation %) of each mitigation togheter
The numbers used here for existing values for armour and resist mitigation are taken from Neoskiller, which uses a reverse engineered resist calculation. I can't promise it is entirely accurate, but the shape of the curves certainly fits with what I saw in testing long ago. Below are the resistances given per point for armour and con skills at the moment in NC2.2:
http://www.cmaster.linxsoft.co.uk/images/forumstuff/NCCurrentResist.png
Not exactly a pretty picture, with weird bumps and rather abrupt changes in slope. To understand a bit more what this means, let's look at the lifetimes graph (y axis is the lifetime muliplier, ie 2 means last twice as long), below:
http://www.cmaster.linxsoft.co.uk/images/forumstuff/NCCurrentLife.png
So what we can see is that despite the changes in slope, more resist points actually get increasingly effective as you go along. Equally, the "bumps" between slope changes lead to some weird behaviour in the resist section. Armour actually works out relativly flat and reasonable, eventually giving about 50% life extension.
It seems that improving the effectivness of resists with increasing points spend is not entirley desirable, and the link between armour and resists, and what these points values mean, is pretty obscure to the player actually in the game.
How this could be better - numbers
While a little mystery can be exciting, it strikes me that players should be able to understand what a "point" of armour or resistance really means. It would also be good to have a nice, stable increase in lifetime throughout the range. To this end, I suggest introducing a new armour/resist calculation:
http://www.cmaster.linxsoft.co.uk/images/forumstuff/ncresistcalc.png
Where D is the factor that incoming raw damage is multiplied by to give final damage, A is armour or resist points value, C is an adjustable factor, and R is the resistance value that players tend to talk about (D = 1 - R).
The prescence of C as a tuneable constant is important, as it lets us decide just how influential we want armour and resists to be. The reason for picking this function is quite simple - it gives a linear increase in lifetime with respect to investment of points. Using a factor of 0.01 (as is used for later examples), every point in armour/resist gives a 1% increase in liftime. So for 20 points, you'd last a fifth longer than somebody without any. With 100 points, twice as long (and 2/1.2 = two thirds longer than the guy with 20 points).
Graphically, this looks something like this for resistances:
http://www.cmaster.linxsoft.co.uk/images/forumstuff/NCNewResist.png
and this for lifetimes:
http://www.cmaster.linxsoft.co.uk/images/forumstuff/NCNewLife.png
Notice how lifetime now increases by a nice, consistent amount in return for investing the points/ getting the armour.
Practically, I'd say 1% per point is too much (it's used here because of similar values to existing situations). 0.5% per point produces a more reasonable picture that might be approriate to start off for both armour and resist - they can be diversified if needs be.
"But" some may say "Shouldn't overinvesting be discouraged? This linearity looks like the opposite". Except that the game already punishes overinvestment - in resist, those first 50 points to get your lifetime up to one and a half of what it was without takes only a single skill point each point. To raise it then up to 75% will cost you more. In armour, the same thing happens - the higher up the TLs you go, the less additonal points you see per TL (and this can be refined of course).
How things are at the moment - armour
It shouldn't really be necessary to provide a recap of the current armour system - for those who need it, take a look at the item DB (http://www.techhaven.org/db/). Some key issues to draw out: belts still have a whole set of int-based belts, letting Spies and monks boost their armour in one stat heavily for the torso only. Spies (and monks) can use high tl int-bsaed underwear, which gives better stats than the mid-tier str or int based that PEs can use. Monks have their complete own set of armour, that gives good resists for pretty much every damage type except poison. The fact that there are only two armour slots per target zone, combined with the current armour-resistance curve means that armour that protects against multiple damage types is rather perferable to those that don't. Energy armour is only available from all-over pieces, and the tank-only Duranium.
So, the conclusions from this: carbon armour is a joke at the moment, delivering less posion than the more versatile kevlar - the boost in piercing and force is largley irrelevant considering how much other protection players will get against this. Duranium is tank only. Kevlar at the moment is pretty fantasic armour. One of the key things to note is that PEs gain very little in this whole space from their strength. The kevlar armour that spies wear is more useful to the than anything in the 40-65 category. They lack the int to use the really effective belts. Getting even bad duranium requires some serious gimpage, and the overalls available to PEs are wearker than those for spies or monks. Practically, PEs wind up using either spy armour or newbie monk armour. Meanwhile, monks with their cripplingly low strength get away scot free (other than low force resist/transport_). Spies get to dance around near enough as tough as a PE. Tanks remain happily built of bricks of course, with the best of almost everything available to them. Spies do have to flail a little for any reasonable energy protection, althoug in theory Holovests serve that very well.
How things could be better - armour
Here, there's a need to tear down a lot of elements that are there for old or bad reasons. To start, I'd see an end to all int-based armour - let class's strength be a genuine indication of their survivability. Remove the locational armour system - it's a neat idea, but the reality is that having to cove everything in a few slots really hampers setup viability, and makes one-offs he titan and viper king chest weird. Sort carbon armour out - the poison component needs to be competitive with kevlar - or perhaps better it could have x-ray added to the mix. Bring in a new line of energy armour, (call it Duranit like the old stuff maybe), in the same strength range of inq/carbon. So Spies could spec for the bottom tier as a choice, and PEs would have default access to middle tier and option to spec for top tier.
More dramatically, I'd like to see psi armour gone as well. It's stats remain absolutely fantastic (holy spirit has more energy than duranium, then adds xray and fire too!). Monks are meant to be weaklings, let's see them feel it. As a compensation, perhaps make some low tier (10-30 ish) "vaporisation" armour available - small amounts of energy/fire, maybe x-ray - at the cost of no pierce force whatsoever. This step needs to be considered a lot more cautiously than the others, but it would be nice to see monks who actually consider con and armour setups like other characters do.
The idea over all of this is that characters lacking in strength have to make harder choices about what they are doing with their strength, and at the same time would see benefits from imping/drugging to increase it. That PEs would actually have something to differentiate them from spies, other than being able to use worse weapons.
One thing that is also important to keep, is that the highest TL armour is 90 or so. Tanks shouldn't be able to bump up strength and see damage AND defense go up as a direct consequence.
Some notes on Power Armour
One of the general principles in Neocron is that setups trade damage for defense (in theory, they should trade off mainskill, weaponskill and defence, but for the moment the first two almost always go togehter). Power armour right now spits in the face of that idea. Higher levels of PA, for every class, give more armour, more mainskill and more weaponskill. This is wrong. The simple fix that works for this is simply to fix all PAs for a class at the same armour values (still allowing for camo to be different). It would be interesting to see PA do different things, but that's changing the game, rather than tweaking it.
Conclusion
Set out above is a plan to make the resistance more comprehensible to players, and more supportive of the class balance and differentiation. The two stages, adjusting the resistance calculation and adjusting the armour sets are related, but the implementation of one needn't require the other - I'd place much higher priority on adjusting the armour set than on tweaking the way things are calculated.
For those interested in the numbers part above, you can find the spreadsheet I did calculations on, with some factors that can be adjusted, here. (http://cmaster.linxsoft.co.uk/images/forumstuff/nc%20armour%20toying.xlsx) - You also see what a linear and an exponential resistance system would be like.
Anyway, thanks for those who bothered reading the whole way through, and open to comments we go!