[dev1027] [dev1032] Status Allocation System + Damage Calculation Broken/Unbalanced

I didn't get any response in my last post so I am putting up the dev tags and writing out my conclusions clearly on this forum. Skip to #9 if you're impatient.

The following 3 posts offer empirical data supporting all the statements I am making in this post:

Test Central: Log Parse Results = MAX Attribute+Gear Effects
http://ffxiv.zam.com/forum.html?foru...7176564&page=1

Log Parse Results! Effect of (INT PTY RANK ELE)on THM Spells
http://ffxiv.zam.com/forum.html?foru...07363078848581

Log Parse Results! Effect of (INT PTY RANK) on Phantom Dart
http://ffxiv.zam.com/forum.html?foru...32715260171115
------------------------------------

1) Empirical Test Conclusion: Increasing Piety has ABSOLUTELY NO EFFECT OR CORRELATION TO INCREASING ACCURACY OR LOWERING RESIST RATES for Phantom Dart, Damnation, Scourge/II, or Shadowsear.

In none of my tests did Piety have any effect whatsoever, even with dramatic changes from 40 to 100 PTY. PTY is an utterly useless stat to invest in.

2) Empirical Test Conclusion: Increasing INT values does steadily increase Phantom Dart average damage. However it may not affect Weapon Skill (Damnation) or Nuke (Scourge/Scourge II) damage. Either the cap was met at a lower value in my test, or INT does not affect Weapon Skill or Nuke damage. Phantom Dart does not yet appear to cap at all.

Increasing INT with Phantom Dart increased damage steadily 0.25% (R43) to 0.5% (R48) PER 1 INT. See test for more details.

3) Empirical Test Conclusion: With Phantom Dart and Scourge, average non-critical and critical damage increases in direct proportion when the change in (PlayerRank/MobRank) increases. Further, there seems to be a point where the rank difference dramatically changes damage. This was observed on the Rank 59 mob.

All tests showed consistently that the R59 mob defensive abilities increased 20% from 58 to 59 while from 55 to 58 each step is only 3%. This proves the calculation system was unbalanced due to player and mob rank. This may be because the R59 mobs' innate stats were raised inconsistently.

4) Empirical Test Conclusion: Per #3 and #2 above, the effect on damage output of increasing your Rank by 1 is as influential as increasing your INT by as much as 16 in this test. Compound this with the players ability to load the INT stat and you have a very unbalanced system (see 5).

5) System Conclusion: Per #3 above, the status point allocation system forces the need to make player rank far too influential in damage calculation, just to mitigate the unbalance caused by loading up stats. The player's ability to so drastically change status parameters makes the effect of changing parameters nearly useless compared to changing rank. Being able to so widely set the range of INT is just plain sloppy and useless.

6) System Conclusion: Per #5 above, the status point allocation system makes any gear status bonus utterly useless, as the range you can increase status points makes a typical gear's +1 to +7 bonus pale in comparison. Per above, a +7 INT bonus inconsistently increases damage, and in the best case the increase is not even half the effect of +1 player rank. It was also untested if gear status points even contributed at all. It's true that increasing VIT and MND with gear bonus has no effect on HP or MP.

7) Appeal: Please produce a combat system where increase of player stats will directly and consistently impact damage output, and lower misses and resist rates.

8) Appeal: Consider smooth curve functions when players and mobs change their stats as rank changes. Dramatic inconsistent differences with just a small variable change will only confuse players. As evidenced, the rank 59 to 59 mob results in a 20% damage decrease compared to the typical 3% change as rank changes by 1.

9) Final Appeal: Implement the best solution, eliminate player and mob rank values from the calculations completely, and eliminate the status point allocation system completely. Rely solely on the primary stats (STR INT DEX PTY MND VIT) that are increased and FIXED accordingly depending on rank and class. This will enhance class (and mob) uniqueness.

This will dramatically increase the effect of raising stats with gear. Raising your STR by +10 via gear bonus should result in a damage output as if they had ranked up 2 levels. This will exponentially increase the value of elite end-game items.

It will create a strong baseline for balancing the calculation system by directly modifying ONLY the primary stats on players or mobs. The current system is too complicated as was admitted by Naoki Yoshida in a recent fansite interview. It worked just fine in FFXI. If it wasn't broken, don't fix it.

Players. if you agree with this assessment, please "like" the post and/or post here to ensure the dev team understands the importance of the issue. If you disagree, please note the # above and explain why.

I really hope the system we get in the Big Battle Patch follows your suggestions or otherwise addresses your valid concerns. The devs have stated that the current formulas are too complex to balance with a practical amount of work, so the patch should bring very different and simpler formulas for damage. It sounds like relations between stats, regardless of rank, would make the most sense, even if I no longer get XP. Oh wait...

Liked, appreciate the time and effort you've put in, this gives some interesting insight and backs up my own trial and error testing. Fully agree, although with these issues in development, I imagine they have already created their own answer to the problem.

Tweaked my points a bit more. I hope they actually have created their own answer to the problem, and will share it with us.

Totally agree with this post.

Ideally stats need to be given to the class as a fixed incremental per level, balancing this kind of system would be much easier not only for healing/damage output but for balancing monsters and NM's.

Removal of Physical levels also as they would no longer be relevant, this would at least go in-line with the 2nd poll results.

Didn't the Delta rank ratio play a large part in FFXI's damage formula also?

Ah, this is fantastic work. Thanks a lot. One thing is that I've seen with my own informal tests that CON spells do indeed scale with INT and with Elemental Allotment. I brought up the disparity in my own thread:

http://forum.square-enix.com/ffxiv/t...nd-As-Intended

But it's absolutely great that you've provided some solid data backing up my claims with respect to THM nukes. Your figures on PIE are pretty damning, I'm wondering it extends even into Bio/Dia/Poison, as SE really pushed the influence of PIE on DoT spells.

Quote:

---

Originally Posted by Zephy

Ah, this is fantastic work. Thanks a lot. One thing is that I've seen with my own informal tests that CON spells do indeed scale with INT and with Elemental Allotment. I brought up the disparity in my own thread:

http://forum.square-enix.com/ffxiv/t...nd-As-Intended

But it's absolutely great that you've provided some solid data backing up my claims with respect to THM nukes. Your figures on PIE are pretty damning, I'm wondering it extends even into Bio/Dia/Poison, as SE really pushed the influence of PIE on DoT spells.

---

No empirical data to prove it, but I've felt like DoTs definitely benefit from higher PIE
Also... now, most experiments are purely for entertainment purposes, since the formula's are changing.
For example.... Before the November patch VIT had a much bigger impact on damage taken....
And logging in after the patch, we all saw a serious decrease in our HP
So with each patch, their is a chance any previous experiments will be made irrelevant...

@OP I do commend your efforts and your grasp of the scientific method, especially in regards to having as few variables as possible....
Very nice work....

I can supply some additional data for you (PM me if you want to see the data gathering spreadsheet).

For Conjurer Cures, each 2 points of Mind adds 1 point to cure amount, and each 3 points of Magical Potency adds 1 point to cure amount. No cap that I could discover (and if there is a cap, it is above the MP pool based on Mind).

Just these two value can completely predict cure amounts, and I have been unable to find any other factor.

Awesome I'd love to see your numbers. I can integrate it into my thread and give you credit for it. Really appreciate anyone who can contribute test data.

If delta rank played a part in XI, it was modest to the point where a decent number of people say your level has no effect on performance at all.

In XIV, level completely dominates every outcome in all the disciplines. Almost no other parameter significantly affects any relevant outcome parameter like HQ rate, successful synth completion rate, sp per hour. They only have very minimal effects on sub-outcomes like accuracy %, damage, completion progress, quality gain, etc. Those outcomes don't actually matter if they aren't that relevant. 45 damage really isn't better than 38 damage if it doesn't unlock higher SP/hour.

Did you mean to say a R50 with minimum INT will outperform a R35 with max int? Heck, a R40 with INT 50 will probably outperform a R37 with maximum INT. A Naked R40 smith beats a fully geared R37, that's for sure.

Once you hit level 50, almost no character progression you devote to will make that much of a difference. Even the difference between Jade Crook and Jade Crook+3 on a R50 conjurer really won't unlock any significant outcomes for you. Your nukes will functionally perform about the same.

The stat and attribute system in XIV is probably the most placebo-heavy RPG system I've ever seen. It's as if they thought we wouldn't notice the stats did anything, and continue merrily getting them as if they did. It certainly makes balancing the game easy when the stats don't really do anything valid.

Quote:

---

Originally Posted by Peregrine

If delta rank played a part in XI, it was modest to the point where a decent number of people say your level has no effect on performance at all.

In XIV, level completely dominates every outcome in all the disciplines. Almost no other parameter significantly affects any relevant outcome parameter like HQ rate, successful synth completion rate, sp per hour. They only have very minimal effects on sub-outcomes like accuracy %, damage, completion progress, quality gain, etc. Those outcomes don't actually matter if they aren't that relevant. 45 damage really isn't better than 38 damage if it doesn't unlock higher SP/hour.

Did you mean to say a R50 with minimum INT will outperform a R35 with max int? Heck, a R40 with INT 50 will probably outperform a R37 with maximum INT. A Naked R40 smith beats a fully geared R37, that's for sure.

Once you hit level 50, almost no character progression you devote to will make that much of a difference. Even the difference between Jade Crook and Jade Crook+3 on a R50 conjurer really won't unlock any significant outcomes for you. Your nukes will functionally perform about the same.

The stat and attribute system in XIV is probably the most placebo-heavy RPG system I've ever seen. It's as if they thought we wouldn't notice the stats did anything, and continue merrily getting them as if they did. It certainly makes balancing the game easy when the stats don't really do anything valid.

---

Now this I can agree with... Rank needs to not matter so much

We're definitely on the same page. As far as my statement illustrating the impact of max vs min INT, I'm by no means making an empirical conclusion here as I did not complete this kind of test case. My main point should have been that INT is proven to make a difference in certain limited contexts (Phantom Dart but not so much Nukes), and being able to so widely set your INT value is just plain sloppy and useless compared to the effect of rank.

Rank obviously dominates as the most influential factor, and I suspect this was done as a quick fix to a broken system. And of course as a consequence the system is not balance-able, and there is no way to improve your performance once you are max rank.

Well said Peregrine. Delta rank is over the top in XIV. I can go from doing 800 pts on a spell to 200 from raising one star >.<

I don't mind them hitting harder or having more HP, but instantly doing 25% is a little nutty. This also has something to do with the current issue with leves ... its either EZmode or a dirtnap, very few have a "fun" setting lol!

Nice job Deus. Stats need to matter more.

After this patch, we can see how the mob ranks go up and get a feel for how consistent or inconsistent the change is.

The move to an 8 person party is good. 15 is just too many to develop a tight strategy. Suppose you were the only highly optimized player in a 15 person party, it wouldnt even make a difference in the fight.

All this new gear is great, but without a solid stat system underneath, it's all for show.

What is the point of making a gear class specific if it does not directly enhance the efficiency of that class?

Quote:

---

Originally Posted by Deus

After this patch, we can see how the mob ranks go up and get a feel for how consistent or inconsistent the change is.

The move to an 8 person party is good. 15 is just too many to develop a tight strategy. Suppose you were the only highly optimized player in a 15 person party, it wouldnt even make a difference in the fight.

All this new gear is great, but without a solid stat system underneath, it's all for show.

What is the point of making a gear class specific if it does not directly enhance the efficiency of that class?

---

Agreed, but I think it's really just more ground work while keeping people occupied & testing the system live simultaneously. Also this can give people some feel of uniqueness being that the new items carry their class/rank requirement.

I've completed another test case as 30 MRD wearing several new NM gears. Test 1 is wearing all the gear plus some accessories. Test 2 is completely naked except for the NM weapon. Expect the results next week. Analysis will take some time as I have to build out the parse script more.

I agree with you wholeheartedly. Your research explains my personal observations.

They need to take away stat allocation, implement a skill grants stat system like Blue Mage from FFXI, and make it so 1 point of each stat is a drastic bonus over not having that extra point.

Having no visible change for 5 to 10 points in a stat makes the game seem like you are not getting anywhere with your stats, It makes gear seem even less necessary.

I AGREE 100% with Final Appeal. GET RID OF RANK BEING PART OF THE FORMULA FOR JUST EVERYTHING.

I know that the solution they chose by making rank a strong factor in just any formula is one way to make sure "soloing" works, but PLEASE PLEASE either get rid of it completely, or tweak it in a way where Rank isn't the only damn thing really affecting anything in the game.

Just as an example. A Gladiator R48 in Full Plate Gear will get a shitton more damage when getting hit by NMs(Buffalo etc.) then a GLA50 with Leather Gear(more then 100def less) just because and only because he is 2 Ranks higher. This is not working and this will make gear whatsoever completely useless as long as Rank is the only factor that really does anything in the game right now.

/signed.

Delta rank is far to large of a factor. Hoping it will be a thing of the past.

I like most of the appeals, with a few oppositions...

-I think most gear should be more cosmetic than impacting on the PC's performance (save "invisible" gears which exist only to produce bonuses)
-Stat allocation is a nice feature, skills/spells should encourage every job to use all 6 stats, and the elemental affinities should make it viable to use elemental spells while using a martial job
-Physical level should apply as a form of "tutorial" to give the player their pool of malleable points
-Discipline rank should only exist to increase attack options (the gamut of attack types available) which only get a modifier for using the same weapon they are native to, not the rank of that discipline (I think the OP and I are on the same page here, but still a bit unsure)

Is the case the same with Dex? Cuz I could use more points in Vit, especially if Dex doesn't really do anything. lol

If they remove physical rank, I hope we can still have point allotment. I think it's unfair if it's automatically distributed. With an auto-stat distribution that means I would have to rely only on gear to increase my damage.

I am only going to be a disciple of magic, with the exception to lancer for Muscle Memory I and II. My mind is at 105. My Int is 145. I want to be a DD conjurer, not a healer. I want to have Intelligence at the highest I can go, without good gear so I can do decent damage. So that way when I someday do get good gear for damage dealing, I could hopefully do way more damage than most people. People who are build to be other jobs, or healing SHOULD never be able to touch the damage I can do when I build to be a DD conjurer. With gear they could, and that's bullshit to me, with auto-stat distribution.

What I don't understand is why this problem with most stats being ineffective is not being fixed immediately. It's the one of the first things you mess with in the beginning of any game, when building a character. All these patches we had so far, this should have been among them. But all I can do is wait, and pray.

I have completed tests on the effects STR and DEX. DEX proves to be useful. Also tested was the effect of NM Gear Stat Bonuses.

See MAIN TEST THREAD for links and details.

Deus, you're awesome.

<3

I reviewed Deus' data, and I need to toss in that it doesn't quite support the general conclusion that PIE is a useless stat. It does support that changing from 40 to 100 Piety against one particular creatures (Doblyns) has no effect. It also supports that rank is a very significant factor in the calculations.

My intuition is that creatures are built with stats and affinities just as players are. Some creatures have very high evade rates (Opos), some have very high defense (Coblyns, Crabs). Certain stat assignment will probably be more effective against some creatures then others.

Let me play devil's advocate here for a moment with removing the class rank from the picture, given the current system : what happens when a physical level 50 player starts off a new class at rank 1?

If stats become the dominant, and class rank is insignificant, those stats would allow the class rank 1 to be essentially as effective as rank 50, barring access to skills and gear. Such character in party play would completely dominate over lower physical level characters, diminishing their enjoyment (unless the goal is to powerlevel those character, letting them sit on the sidelines).

My intuition is that the class rank was brought in as the dominating factor in order to address this, but the side effects are significant. I think that any change to the way that stats work will have to go hand-in-hand with a change to the way that physical level works, and the way that points are assigned for classes of different ranks.

Quote:

---

Originally Posted by Amineri

I reviewed Deus' data, and I need to toss in that it doesn't quite support the general conclusion that PIE is a useless stat. It does support that changing from 40 to 100 Piety against one particular creatures (Doblyns) has no effect. It also supports that rank is a very significant factor in the calculations.

---

I don't think he was concluding that PIE is useless- just that it's dependent on the spell cast, or rather the spells he liked to use had no use for PIE. I could have read wrong, but the data seemed to prove only that PIE doesn't effect some spells/skills and that rank holds significant sway in damage calculation.

Quote:

---

Originally Posted by Jinrya-Geki

If they remove physical rank, I hope we can still have point allotment. I think it's unfair if it's automatically distributed. With an auto-stat distribution that means I would have to rely only on gear to increase my damage.

I am only going to be a disciple of magic, with the exception to lancer for Muscle Memory I and II. My mind is at 105. My Int is 145. I want to be a DD conjurer, not a healer. I want to have Intelligence at the highest I can go, without good gear so I can do decent damage. So that way when I someday do get good gear for damage dealing, I could hopefully do way more damage than most people. People who are build to be other jobs, or healing SHOULD never be able to touch the damage I can do when I build to be a DD conjurer. With gear they could, and that's bullshit to me, with auto-stat distribution.

What I don't understand is why this problem with most stats being ineffective is not being fixed immediately. It's the one of the first things you mess with in the beginning of any game, when building a character. All these patches we had so far, this should have been among them. But all I can do is wait, and pray.

---

I also am and will be an elemental caster type, but don't see the same issue as you, IF stats are automatically alloted to a class then it makes gear all the more important to seperate one of us from the next.

Lets not forget that gear and progression is the main reason most of us play these games.

I would like to see a compromise though in the way that stats are alloted automatically but elemental (stat points) affinity is user defined.

That being said I am pretty sure they plan to make stat points user defined across all classes and then save them to that class.

Battle Reform Blueprint by Lead Battle Planner
http://forum.square-enix.com/ffxiv/t...995#post138995

I'm pleased with their responses, see my thread here:

http://forum.square-enix.com/ffxiv/t...ulation-Broken

I wish they would just ffxi level up stats

Deus, a bunch of us are thinking about soloing and the effect of INT on damage. A lot of people are concluding that INT has no effect based on your test and are zeroing out their INT. However, your test was with INT nearly maxed out! I'd love to see the results of your test with 0 in INT, 40 in INT, and so on.

Quote:

---

Originally Posted by Neptune

Deus, a bunch of us are thinking about soloing and the effect of INT on damage. A lot of people are concluding that INT has no effect based on your test and are zeroing out their INT. However, your test was with INT nearly maxed out! I'd love to see the results of your test with 0 in INT, 40 in INT, and so on.

---

It makes a difference if you're going from 16 to 40~80.

After 100, not so much of a difference. Diminishing returns.