Maintenance for the week of April 15:
• PC/Mac: No maintenance – April 15
• ESO Store and Account System for maintenance – April 16, 8:00AM EDT (12:00 UTC) - 12:00PM EDT (16:00 UTC)
The issue is resolved, and the North American PC/Mac megaserver is now available. Thank you for your patience!

Build conversion - hybrid build disappearing and builds approaching a 'best' build due to simplifica

Faugaun
Faugaun
✭✭✭✭✭
Multiple threads have mentioned that with the removal of softcaps the hybrid build is going to disappear. As it currently stands I agree that there is a possibility that builds using both stamina based and magicka based abilities will disappear if Zos do not tread lightly.

Essentially this occurs because specialization out performs non-specialization.

A brief example if in total I can get 100 stamina/magicka points then builds with either 100 stamina points or 100 magicka points will be more powerful than a build which only incorporates 50 points in each.

This was a way that softcaps and diminishing returns on stats helped to encourage build diversity. It added a whole new metagame component of "well further points into _____ are not effective. Which secondary stat should I focus on?" This encouraged blended builds and allowed for much more complex interaction of stats/abilities. With soft caps removed there is no incentive to blend different styles of play and instead the incentive is to completely fully invest offensive capabilities into one particular line while ignoring all others. This maximizes your play at that particular role. Further once players have removed 50% of the skills as choices (assuming it is a 50/50 split) then due to a lack of blended encouragement they will simply need to analyze for the best 5+1 skill bar with their chosen role/offense type.

This is bad as it will encourage people to gravitate towards fixed 'ideal' builds in the end no matter how you slice it there will be 6 ideal offensive builds.

1) Magic based Tank
2) Magic based DPS
3) Magic based Healer
4) Stamina Based Tank
5) Stamina Based DPS
6) Stamina based Healer

These will likely be reduced to only 3 optimal builds (a Stamina based healer is likely worse than a magicka based healer, tanks are similar ....and for DPS one or the other will be better).

These builds will largely be able to be determined by fairly simple analysis

Take DPS role for example:

I take 10 characters at a gradient of different stat levels (1/10th max, 2/10th max, 3/10th...9/10th, 10/10). With these 10 characters in place (or one max level character with different gear for weaker levels) a rough plot can be made of the damage points on a graph (Microsoft excel....) Then using built in functions produce a polynomial best fit function for each skill. Finally take a mathematical programming suite (R, SAS, MATLAB, etc...) Write a short program includinging all possible skills assigned to their appropriate functions (what maybe 150 skills total? ) then you insert your current stat point value for magicka and stamina hit play....the program will analyze each function output a damage value for each ability and construct an ordered list of all skills from most powerful to least powerful.

After this to make the best skill bar simply identify the best AoE and the best ST(single target damage skills) put those on then fill in the remaining 3 skills with utility abilities to augment the best skills (guess what the same list created earlier might put some utility skills much higher suggesting they are better to look at).

This is easy to do and costs $0,might take 4 max level characters (1 of each class, and so WW and Vamp)...data collection is simple (equip the gear record the value, rinse repeat 10x for each skill line, 12 class skill lines, WW, Vamp, 6 weapons, then the stragglers from actives from other places: in total 20-25 x10x5 numbers to record....or 1000-1250 numbers to record creating 100-125 plots in excel and 100-125 equations to compare with a program...write the code, enter your values output 30 seconds later) The whole process might take an afternoon (2 if you are slow)...This is the problem with simple systems, they are also simple to break down ...better yet write an add-on that simply grabs these values as you play and exports the data to a database which then generates the best fit equations and produces a data value for all skills and organizes top picks automatically...this just requires installing the add-on and playing then reading the values...again simple systems are easy to brute force test for ideals.

The best way to prevent this is to create a system sufficiently complex that it cannot be brute forced (or prevent access to the data needed to brute force the system, this is probably not an option since players like to know the numbers...). To create a more complex system you need to add variables, soft cap adds a variable (a second or third stat focus). Another variable is synergies, each synergistic ability increases the number of combinations by the number of combinations of possible synergies multiplied by the number of skills (impacted by the synergies).

Let me put this into perspective a bootstrap analysis I did recently generated and then analyzed 1,000,000 data points the whole analysis took about 45 minutes of the computer crunching the numbers to give me an output. It took 3 weeks to write the 150 line program, test it, debug and make it work correctly. Think about that 1,000,000 data points in 45 minutes, now how long do you imagine 1,250 will take?

OK, enough of the mathematical theory ....here's the bottom line

This system is fairly simple already and pushing it to a simpler state (especially in the age of big data) borders on the edge of a brute force for the best ability (heck it could already be done...I am not going to ...but someone playing the game probably has already). The system needs to be made more complex this can be done several ways.

1) increasing the amount of stat interaction behind the scenes to create a damage value (or whatever other value I'm just picking on damage today but the same principle applies to all other roles or catagories). This makes the process much more difficult to analyze ( not impossible though just need enough data).

2) Make automation of data collection more difficult. This is probably pretty hard to do because players like to know values, when values are displayed then that information is avalible to be gathered automatically. Perhaps a less defined system might help to fuzzy the numbers (consider if the player only knows that a piece of armor adds 'substantial' weapon damage...no specific numbers and the overall stat for weapon damage is 'Daedric' without specific numbers). This helps but by itself is not adequate (plus I'm not sure the player community would like the change). In effect what this is doing is creating catagories to be compared data could still create a rough curve based on these generalized stat descriptions then place them on a scale 1-10 and create a rough predictive formula (within the confidence intervals) and have a general idea of which skills are in the top tier (maybe not their exact rankings but pretty close...and more data would put closer and closer).

3) add more variables, like synergies (this can increase the complexity by multiples or possibly exponentials, if done correctly).

4) make immunity possible (kind of like in diablo, where some NPCs would be completely immune to a certain damage type). This renders a highly specified build to be rendered useless....though in the current form it would be physically or magically immune ....perhaps breaking up the stats further (this adds more variables also increasing the complexity) ...maybe holy damage, fire damage, ice damage, lightning damage, piercing damage, blunt damage, slicing damage, projectile damage etc... Where a player may stack for each (or several of these) both offensively and defensively.

5) Create stat packages on gear so a gear would have both a physical and a magical component which would be roughly equal. ( this maintains the integrity of all abilities when min/maxing, it does not help with complexity).

6) more secondary and tertiary effects on skills. As skills become more complex it becomes harder to create simple programs which analyze all skills and increases the time input by a programmer (having to write skill specific code is time consuming, while not a complete fix as the time input needed to solve the problem increases then more people are deterred from attempting). This can probably be worked around by automated analysis of combat logs....unless the combat logs are made fuzzier (instead of x ability causes y damage to target, it shows the target take x 'magical' damage...again players probably won't like this sort of vagueness).

7) Triggered effects (either procs, which could be mathematically calculated unless they are done vaguely, or more ally triggered effects, like those present in the game...even so it could probably be determined how often an individuals friends press x...but it would be more complicated, remeber complexity is our friend).

8) Time of day effects, for vampire and ww skills ... These would actively change the way abilities work based on time of month or time of day. Further weather related changes , say if it is raining then fire magic is lowered but maybe lightning damage is increased ...and in snow storms maybe ice damage is increased. Again complexity makes it harder.

9) More shifting global buffs, food bonuses, and additional triggered and automatic buffs/debuffs related to various events. Make them vague as possible too...remember vauge increases the amount of data needed and makes the answer more fuzzy. This also adds complexity which is our friend.

10) periodic changing of abilities (yes an automated process will quickly overcome this but if a point can be reached where the abilities are changed frequently enough then the program that analyzes the abilities also needs to be frequently updated...this method (while a pain for Zos) is a bigger pain for the person deconstructing the data....if done correctly....this might be disliked by the players though as the gear/spec they build is frequently changed and they cannot specialize...catch 22 here.

With adequate time I am sure I could provide more ideas for improving this aspect...its fairly simple though, increase complexity, increase vagueness and it becomes more difficult (probably never impossible but much like a firewall...the difficulty acts as a deterent to hackers....this makes the overall system more robust and better). This is similar to the tools that the devs probably already use to compare and balance abilities except they are accessed and reverse engineered from the end output while the developer approach is from the inputs and the equations.

I realize your eyes are probably rolling in the back of your head...especially if you have made it this far...that said I hope you understand and I would love some feedback or not your call.
Edited by Faugaun on January 31, 2015 5:17PM
  • joshisanonymous
    joshisanonymous
    ✭✭✭✭
    Disclaimer: Definitely didn't read all that after I got the gist of what you were saying.

    I disagree with the basic premise though. Right now, we don't have "hybrid builds," we have "builds that essentially have everything capped." The removal of softcaps will lead to a lot more diversity, IMO.

    For instance, you're claiming that there will be 3 viable builds because focusing on one specific role and maxing out only a few things that optimally benefit that role, will be the rule of thumb. Of course, you're right, someone who focuses exclusively on DPS after 1.6 will absolutely outperform someone who focuses on DPS but decides they want a couple defensive things also, but they'll be outperforming purely in DPS.

    There's a real choice now, without the softcaps. You can go crazy high DPS machine, but you'll have nothing else. Like, literally nothing else. You can spend every last point in your damage stat and damage passives and damage everything, but you'll have 10k health and 10k secondary resource and no armor and no spell resist. So, if you get your DPS from magicka, you'll die when a fly lands on your shoulder and you'll be able to dodge roll once before being completely vulnerable to stuns.

    The removal of softcaps means there will be some real give or take, and I would imagine you'll see much more diversity because of it. Instead of 3 viable builds, as you've said, there will be a large continuum of possibilities depending on exactly the playstyle each person wants.
    Fedrals / EP / Dunmer Nightblade

  • Faugaun
    Faugaun
    ✭✭✭✭✭
    Disclaimer: Definitely didn't read all that after I got the gist of what you were saying.

    I disagree with the basic premise though. Right now, we don't have "hybrid builds," we have "builds that essentially have everything capped." The removal of softcaps will lead to a lot more diversity, IMO.

    For instance, you're claiming that there will be 3 viable builds because focusing on one specific role and maxing out only a few things that optimally benefit that role, will be the rule of thumb. Of course, you're right, someone who focuses exclusively on DPS after 1.6 will absolutely outperform someone who focuses on DPS but decides they want a couple defensive things also, but they'll be outperforming purely in DPS.

    There's a real choice now, without the softcaps. You can go crazy high DPS machine, but you'll have nothing else. Like, literally nothing else. You can spend every last point in your damage stat and damage passives and damage everything, but you'll have 10k health and 10k secondary resource and no armor and no spell resist. So, if you get your DPS from magicka, you'll die when a fly lands on your shoulder and you'll be able to dodge roll once before being completely vulnerable to stuns.

    The removal of softcaps means there will be some real give or take, and I would imagine you'll see much more diversity because of it. Instead of 3 viable builds, as you've said, there will be a large continuum of possibilities depending on exactly the playstyle each person wants.

    There will still be offensive/defensive give and take there but what I am saying is that there won't be give and take on the stamina vs magicka spectrum it will be either one or the other not both.
  • joshisanonymous
    joshisanonymous
    ✭✭✭✭
    Faugaun wrote: »
    There will still be offensive/defensive give and take there but what I am saying is that there won't be give and take on the stamina vs magicka spectrum it will be either one or the other not both.

    And what I'm saying is it will be both. Dodge roll, one time, because all you cared about was magicka. You're looking at this as if you're playing a static game where the only goal that matters is getting the highest DPS number, the highest heal number, or the highest mitigation number. That's just one aspect of what you may want from your character, depending on your playstyle. You're taking out the largest variables of all, a dynamic game world and humans interacting with it.
    Fedrals / EP / Dunmer Nightblade

  • Bouvin
    Bouvin
    ✭✭✭✭✭
    I haven't enjoyed the changes in 1.6 at all.

    My main even has all skill lines leveled to 50, all skills leveled, and almost all morphs leveled, so I've been able to experiment with every different skill.

    Haven't been impressed with the changes so far.

    LA is much more squishy now. That's fine, but it isn't offset with extra offensive ability. Basically LA is getting owned.

    Medium is meh... the problem is that you are now burning through stamina so fast you're almost instantly out of resources. So again... the main "benefit" of the armor type is non-existent.

    Heavy..seems like the way to go now, especially with shield stacking. You get the most protection without a big hit to DPS or resources. So effectively, it's win win win for heavy. I'm fearing 1.6 is going to be about getting as much Armor + Spell Resist and stacking shields. This puts classes that don't have much for damage shields at a real disadvantage.

    I'm sure there are people that can make LA and Med work. But once 1.6 hits don't be surprised to see everyone moving over to Heavy just like we have with LA now.

    I'm not saying a change wasn't needed, but the current change doesn't solve anything. It just trades LA as FOTM for Heavy. There's no balance and it was very poorly thought out.

    The stat scaling was also poorly thought out and only seems like a reason to motivate us to grind CP and user CP boosters from the store. Not impressed with that either. Progression shouldn't mean "knocking everyone back a peg" so they can "progress" to "where they are now".
  • Faugaun
    Faugaun
    ✭✭✭✭✭
    Faugaun wrote: »
    There will still be offensive/defensive give and take there but what I am saying is that there won't be give and take on the stamina vs magicka spectrum it will be either one or the other not both.

    And what I'm saying is it will be both. Dodge roll, one time, because all you cared about was magicka. You're looking at this as if you're playing a static game where the only goal that matters is getting the highest DPS number, the highest heal number, or the highest mitigation number. That's just one aspect of what you may want from your character, depending on your playstyle. You're taking out the largest variables of all, a dynamic game world and humans interacting with it.

    On a PvP landscape you are probably right because other humans will find and exploit your weakness. I'm not sure it adds to the current system but it does produce an artificial softcap (unlablled where it is BTW). In PvE content this is not the case from what I have seen, maybe others have different observations?
    Bouvin wrote: »
    I haven't enjoyed the changes in 1.6 at all.

    My main even has all skill lines leveled to 50, all skills leveled, and almost all morphs leveled, so I've been able to experiment with every different skill.

    Haven't been impressed with the changes so far.

    LA is much more squishy now. That's fine, but it isn't offset with extra offensive ability. Basically LA is getting owned.

    Medium is meh... the problem is that you are now burning through stamina so fast you're almost instantly out of resources. So again... the main "benefit" of the armor type is non-existent.

    Heavy..seems like the way to go now, especially with shield stacking. You get the most protection without a big hit to DPS or resources. So effectively, it's win win win for heavy. I'm fearing 1.6 is going to be about getting as much Armor + Spell Resist and stacking shields. This puts classes that don't have much for damage shields at a real disadvantage.

    I'm sure there are people that can make LA and Med work. But once 1.6 hits don't be surprised to see everyone moving over to Heavy just like we have with LA now.

    I'm not saying a change wasn't needed, but the current change doesn't solve anything. It just trades LA as FOTM for Heavy. There's no balance and it was very poorly thought out.

    The stat scaling was also poorly thought out and only seems like a reason to motivate us to grind CP and user CP boosters from the store. Not impressed with that either. Progression shouldn't mean "knocking everyone back a peg" so they can "progress" to "where they are now".

    I think the underlying problem (regarding the light/medium/heavy armor) is that light, medium, heavy can all have the same crafted stats applied to them. When in fact they should have different researches applied to them...

    Heavy armor should focus on tanky stats, or supplement to tanky stats or reduced offensive stats in exchange for the armor mitigation.

    Medium armor should focus on close range style builds and augmenting those type stats (could be either magic or physical).

    Light Armor should focus on ranged glass cannon type builds and control utility and healing.

    I think the big problem with armor currently is that the stats are largely available on all types ....so currently its a decision between set bonuses (heavy armor stat bonuses are lacking ATM) and it sounds like after 1.6 it will be based on Set bonuses and mitigation...a step in the right direction unless mitigation is overly controlling effect...so maybe a good second factor (adding complexity, this is good) but maybe to strong of an effect compared to the other effects...with complaints about regen rates perhaps the best course of action would be to increase magicka regen on light armor and stamina regen on medium armor pieces ...still not ideal but a step in the right direction? Maybe?
  • Emma_Overload
    Emma_Overload
    ✭✭✭✭✭
    ✭✭✭✭✭
    I hope ZoS doesn't read all this and decide to bring back soft caps, because they suck!

    As some one already suggested on another thread, all the devs need to do to solve the "hybrid problem" is to base the resource-driven component of the damage formula on Magicka + Stamina COMBINED. They wouldn't even have to rebalance the damage stat at all - just replace Magicka or Stamina with:

    (Magicka + Stamina) / 2

    Problem solved!
    Edited by Emma_Overload on January 31, 2015 9:36PM
    #CAREBEARMASTERRACE
Sign In or Register to comment.