Forum Settings
       
This Forum is Read Only

Log Parse Results! Effect of (INT PTY RANK ELE)on THM SpellsFollow

#1 Mar 21 2011 at 1:38 PM Rating: Excellent
*
56 posts
Note: This test is part of a compilation of tests. The goals of these tests and the scopes are defined in the following thread. This thread is a subthread of the master thread:
TEST CENTRAL <== Massive updates, thread rewritten!


========= I. BACKGROUND

I've updated my Perl script to output more data values, more organized, and consistent. I've also switch the display of the data to use Google Spreadsheets. I think you'll find reading the data much more enjoyable. The field+format enhancements will carry to any type of skill, not just spells. The following 3 tests were executed before the February SP patch.


========= II. PURPOSE

TEST 1 and 2 show the effect of INT and PTY on spell damage and resist. TEST 3 was to test the effect of elemental point difference, but due to the Febrauary patch, I was forced to rank up during TEST 4. The impact of ranking up hugely affects calculations. So TEST 4 is not provided. TEST 3 is provided to illustrate more data to compare with TEST1&2 which are all player Rank 48. The variation in mob rank illustrates the effect of (DELTA)PlayerRank/MobRank.


========= III. EXPERIMENT CONSTANTS, VARIABLES. EXPERIMENT RESULTS
The spreadsheet and definitions within should be self-explanatory.
Note the spreadsheet has multiple pages.
https://spreadsheets.google.com/ccc?key=tekgszFtn5t2v2wGffQclRw&gmrcpt


========= IV. SCOURGE CONCLUSIONS

  • 1. Effect of (DELTA)PlayerRank/MobRank: 48 VS 55,56,57,58,59
    • Conclusion:
    • Scourge's HITAVG & CRITAVG increases due to a function in direct proportion when the (DELTA)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.
    • Data Citation:
    • See spreadsheet. Normally the DMG change a mere 1-4%, but between mob ranks 58 and 59, the DMG changed a whopping 16-20%! This may be due to the mob stats not being set consistently, not necessarily the damage formula.
    • Discussion:
    • The scaling is consistent and matches my Phantom Dart test exactly. Damages scales exponentially, such that the closer the player rank is to the mob rank, the less damage changes when mob rank changes. Apparently from this data, the (DELTA) Mob/Player Rank dramatically affects the damage inconsistently, not on a smooth curve. These conclusions are the same as in my Phantom Dart experiment.


  • 2. Effect of INT: 174 VS 166
    • Conclusion:
    • Scourge's HITAVG & CRITAVG has NO CORRELATION with change in INT!
    • Data Citation:
    • See spreadsheet page "Conclusions: INT". The damage barely changes between tests, and it changes with no pattern.
    • Discussion:
    • Raising your INT has no effect, under these testing conditions. Perhaps the INT cap was met, or INT is not even in the formula. You can definitely get away with a LOWER INT when only nuking.


  • 3. Effect of PTY: 40 VS 100
    • Conclusion:
    • Scourge's resist rates, or PART%+MISS% has NO CORRELATION between changes in PTY. This is final and definitive, and the main goal of this experiment.
    • Data Citation:
    • See spreadsheet page "Conclusion: PTY". Even with 335 and 434 trials with a GAIN OF 60 PTY, there is no consistent pattern, sometimes increasing, sometimes decreasing, but always not far from the average.
    • Discussion:
    • In the spreadsheet, I added the MISS% and PART% because they should both be directly related to PTY, but also presented them separately.



    ========= V. SCOURGE II CONCLUSIONS

  • 1. Effect of (DELTA)PlayerRank/MobRank: 48 VS 55,56,57,58,59
    • Conclusion:
    • Scourge II's HITAVG & CRITAVG increases due to a function in direct proportion when the (DELTA)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.
    • Data Citation:
    • See spreadsheet. Normally the DMG change a mere 0-3%, but between mob ranks 58 and 59, the DMG changed a whopping 19-23%! This may be due to the mob stats not being set consistently, not necessarily the damage formula.
    • Discussion:
    • The scaling is consistent and matches my Phantom Dart test exactly. Damages scales exponentially, such that the closer the player rank is to the mob rank, the less damage changes when mob rank changes. Apparently from this data, the (DELTA) Mob/Player Rank dramatically affects the damage inconsistently, not on a smooth curve. These conclusions are the same as in my Phantom Dart experiment.


  • 2. Effect of INT: 174 VS 166
    • Conclusion:
    • Scourge II's HITAVG & CRITAVG has NO CORRELATION with change in INT!
    • Data Citation:
    • See spreadsheet page "Conclusions: INT". The damage barely changes between tests, and it changes with no pattern.
    • Discussion:
    • Raising your INT has no effect, under these testing conditions. Perhaps the INT cap was met, or INT is not even in the formula. You can definitely get away with a LOWER INT when only nuking.


  • 3. Effect of PTY: 40 VS 100
    • Conclusion:
    • Scourge II's resist rates, or PART%+MISS% has NO CORRELATION between changes in PTY. This is final and definitive, and the main goal of this experiment.
    • Data Citation:
    • See spreadsheet page "Conclusion: PTY". Even with 157 and 217 trials with a GAIN OF 60 PTY, there is no consistent pattern, sometimes increasing, sometimes decreasing, but always not far from the average.
    • Discussion:
    • In the spreadsheet, I added the MISS% and PART% because they should both be directly related to PTY, but also presented them separately.



    ========= VI. SHADOWSEAR CONCLUSIONS

  • 1. Effect of (DELTA)PlayerRank/MobRank: 48 VS 55,56,57,58,59
    • Conclusion:
    • Too few trials to conclude. May differ based on mob's current HP value.
    • Data Citation:
    • See spreadsheet.
    • Discussion:


  • 2. Effect of INT: 174 VS 166
    • Conclusion:
    • Too few trials to conclude. May differ based on mob's current HP value.
    • Data Citation:
    • See spreadsheet.
    • Discussion:


  • 3. Effect of PTY: 40 VS 100
    • Conclusion:
    • Too few trials to conclude.
    • Data Citation:
    • See spreadsheet.
    • Discussion:



    ========= VI. OTHER CONCLUSIONS

    Proposed INT Soft Cap, Does It Exist?

    I've learned a lot with this experiment, because while Damnation damage was not increased by INT, surprisingly, none of the these spells were as well. This means either the cap was met, or INT is not in the equation at all! This means so far, only Phantom Dart is proven to benefit from INT. Future tests with a larger change in INT, at a level FAR below where a presumed cap might be will shed more light on this.

    In the spreadsheet I also calculated the impact of ELE points, even though the difference was small, and my test had been interrupted by an unexpected rank up, it was worth just going through the motions to show that the ELE points did not contribute to damage nor resist rates. I did not include ELE points in the conclusions above since it was so inconclusive, but am pointing it out here for you to look at the data.

    As stated, the main goal of this experiment was to establish the benefit of PTY. Here is irrefutable proof that PTY is useless, so far for Scourge, Scourge II, Damnation, AND Phantom Dart...

    Which is better? Phantom Dart or Scourge Nuking for SP/hr?
    If you are going to be Scourge nuking instead of Phantom Dart, you can get away with a lower INT. You would do well to try both methods and note your SP/hr. I found Phantom Dart was MUCH higher SP/hr (yes I did track it while I did my tests). With Phantom Dart, my highest recorded period of SP return was 21167 SP/hr over 32min. With Scourge, my highest recorded period of SP return was 8454 SP/hr over 54min. Those times are during a high-efficiency grind with no interruption from other players and a steady spawn circling routine. Note these values are PRE-FEB patch, the SP/hr was roughly doubled at that time, and coblyns later nerfed, so use only for comparison purposes between the two.

    Conclusion, when soloing/duoing SP grind mobs, stick to Phantom Dart and max your INT, and minimize your PTY.


    Normal skills (Phantom Dart) Weapon Skills (Damnation) and Spells (Scourge) relate to Disciple of War (STR DEX)
    I would go so far as to suggest with the level of detail I have revealed with damage calculation that the skills I have tested as DoM apply to DoW. I will next be testing similar skill types as 50 DoW, after the development team fixes this unbalanced, unpredictable calculation system.

    A Final Recommendation for stat balancing for SP grind to accommodate DoW AND DoM jobs
    STR: Max it as high as possible to affect basic attack damage. Will cap out or not be relevant for weapon skills.
    VIT: Maximize after you are happy with your other parameters. Affects Max HP and may also contribute to reducing raw damage.
    DEX: Minimize, its useless.
    INT: Max it as high as possible to affect basic wand attack damage. Will cap out or not be relevant for weapon skills.
    MND: Don't increase it past 40. You shouldn't ever have issues getting MP with Siphon MP.
    PTY: Minimize, it's utterly useless. May still affect enfeeble accuracy and duration, but this is untested and would only matter for NMs. Not even worth testing in the current calculation system.

    Appeal to Naoki Yoshida and the Dev Team
    Please produce a combat system where increase of player stats will directly impact damage output, and lower misses and resist rates. This is the only way gear with stat bonuses will have any impact. Even after the November patch where spells became usable, spells still give inferior SP/hr over using basic attacks, and are not at all affected by increasing your stats.

    Further, results like I observed in both tests, where the damage was dramatically reduced going from the R58 to R59 mob show inconsistent formulas are being applied. Consider smooth curve functions, in this case, when mob rank changes. Dramatic inconsistent differences with just a small variable change will only confuse players.

    Get rid of physical level, it will diminish any value from having gear with stat bonuses, in addition to providing the players too many variables for things they don't need to change, which just confuses them. Too many custom variables also keeps you from having a balanced system. Fix stats with respect to each class and rank, just like it works with Mobs and in FFXI. FFXI worked. If it wasn't broke, don't fix it.


    Edited, Mar 26th 2011 5:21pm by DeusBahamut
    ____________________________
    Deus Diabolus, Gstar Raw, Besaid Server
    Leader of Lifestream LS
    http://lifestreamls.enjin.com
    #2 Mar 21 2011 at 2:52 PM Rating: Decent
    Scholar
    **
    844 posts
    Thanks!
    ____________________________
    See your face upon the clear water. How dirty! Come! Wash your face!
    #3 Mar 21 2011 at 3:59 PM Rating: Good
    *
    56 posts
    The main TEST CENTRAL thread has been massively updated as well.

    Edited, Mar 21st 2011 6:07pm by DeusBahamut
    ____________________________
    Deus Diabolus, Gstar Raw, Besaid Server
    Leader of Lifestream LS
    http://lifestreamls.enjin.com
    #4 Mar 21 2011 at 6:12 PM Rating: Good
    Avatar
    ***
    2,000 posts
    First of all, thanks for taking the time to actually test this out.

    My one suggestion would be to increase the difference between the allocated stat values before your next series of tests. 174 --> 166 is not much of a difference compared with the recent player discussions here and on the offical forums.

    At the moment there are basically three schools of thought when it comes to how mages should allocate Int and Pie:

    1. "All mage stats are important, so you should use an even distribution such as 120 Int, 120 Pie. I've noticed an increase in spell/debuff damage/accuracy/duration whenever boosting Int or Pie."
    2. "There are severely diminishing returns above 80 or 90 points in Int or Pie. Allocate Int and Pie up to 80-90 each, then dump the extra points into Vit/Mnd."
    3. "Int is virtually worthless compared with the other stats. Leave Int at base levels, take Pie to 80 or 90, cap Vit, and put the rest in Mnd."

    I would personally be very curious to see comparisons of spell damage and accuracy at each of the "point thresholds:"

    17
    40
    80
    127
    174

    This type of testing would at least let mages make informed decisions about how they allocate their stats. At the moment we only have casual observation and unverified anecdotes...
    ____________________________
    #5 Mar 22 2011 at 12:36 PM Rating: Good
    *
    56 posts
    Thanks for the feedback. This kind of discussion is what I'm looking for. I still need to go over the results in both of my tests in a more thorough organized fashion to draw more conclusions for this test, but here's my initial response.

    My prior Phantom Dart/Damnation test proved that there is still a return from increasing INT from 166 to 184 for those skills. At that time I was attempting to disprove that a soft cap existed, a point at which increasing INT gave ZERO return. I proved that EITHER the cap is somewhere between 166 and 184 OR there is no cap (under these specific testing conditions at least).

    So this time around, in nearly the exact same testing conditions, I tested spells. You can see there is nearly no return from going 166 to 174 INT. This means the soft cap for INT was hit *for spells* (under these specific testing conditions). So I proved there is a soft cap, and it differs from the cap of Phantom Dart and Damnation. Now that we know there is a soft cap, we will have that in mind when doing tests at levels (40-80-100-122-174).

    The main objective of this test was once and for all, to squash PTY. I've had it with investing in a stat that does nothing, and requiring a HUGE number of trials to get any sense of "hit % up" return. And I've done just that. As you can see PTY HAS NO EFFECT/CORRELATION ON REDUCING OCCURRENCE OF RESISTS OR PARTIAL RESISTS when going 40 to 100 under these testing conditions. The same statement holds true in my Phantom Dart tests.

    Conclusion, if you are going to be Scourge nuking instead of Phantom Dart, you can get away with a lower INT. You would do well to try both methods and note your SP/hr. I found Phantom Dart was MUCH higher SP/hr (yes I did track it while I did my tests). With Phantom Dart, my highest recorded period of SP return was 21167 SP/hr over 32min. With Scourge, my highest recorded period of SP return was 8454 SP/hr over 54min. Note these values are PRE-FEB patch, the SP/hr was roughly doubled at that time. Note your SP/hr may differ under other conditions.

    Conclusion, when soloing/duoing SP grind mobs, stick to Phantom Dart and max your INT, and minimize your PTY. There's your answer.

    This is the only school of thought as far as I'm concerned. I've seen no tests supporting any other.

    Further I would go so far as to suggest Phantom Dart and Damnation behave like all other NormalAttacks and WeaponSkills with respect to STR and DEX.

    While it would be an interesting prospect to test PTY effect on status effect (it's on my TEST CENTRAL thread to-do list), it would have zero impact on SP grind, and is now counterproductive to your SP/hr to raise PTY for that purpose.

    The cost to get 40 PTY is relatively low, so I recommend putting any stat you expect any return for at 40. I also recommend 40 MND if only for the MP and healing capabilities.

    The rate of return of increasing a stat and the damage that results, or the "scaling" of the damage may depend on the mob type and its defensive capabilities. It will require tests on weaker mobs and highest rank NMs to conclude the full range of scaling. In other words, a high INT may be "overkill" and have no "returns" on your typical SP grind mob, but may prove to do double damage to an NM. One interesting thing I noted is I never got 1/4 resists on these SP grind mobs, but get 1/4 resists all the time against Uraeus. In my opinion it's most important to optimize at the SP grinding mob type, because optimization is most beneficial when you are ranking up or farming. Also note Phantom Dart is devastating to mobs lower rank than you, while spell damage caps much lower and scales much less when used on lower rank mobs.

    Regarding additional testing at this time, we can safely assume that our results will not be consistent or scale well at all. The damage calculation system is broken, as spoken by Naoki Yoshida himself that it's "not balanced" . So I won't be doing any more battle damage stat optimization testing until after it's fixed.

    Not to mention the entire Physical Level Stat Point allocation system is going to be scrapped. And this is a very good thing. This system makes any equipment bonuses worthless in comparison. Which makes the "best" gear worthless...

    Edited, Mar 22nd 2011 3:17pm by DeusBahamut
    ____________________________
    Deus Diabolus, Gstar Raw, Besaid Server
    Leader of Lifestream LS
    http://lifestreamls.enjin.com
    #6 Mar 25 2011 at 4:17 PM Rating: Good
    *
    56 posts
    Fully parsed all data to match the methodology in my Phantom Dart post. Integrated my post above with these conclusions.

    Added my final recommendation for all players how to allocate their stat points.

    Added appeal to Naoki Yoshida on exactly how he needs to fix the current combat calculation system.
    ____________________________
    Deus Diabolus, Gstar Raw, Besaid Server
    Leader of Lifestream LS
    http://lifestreamls.enjin.com
    This forum is read only
    This Forum is Read Only!
    Recent Visitors: 22 All times are in CST
    Anonymous Guests (22)