Item Search
     
BG-Wiki Search
+ Reply to Thread
Results 1 to 20 of 20

Thread: Shield Block Rate Testing     submit to reddit submit to twitter

  1. #1
    Old Merits
    Join Date
    Sep 2010
    Posts
    1,094
    BG Level
    6
    FFXI Server
    Ragnarok

    Shield Block Rate Testing

    Reposting from over on AH.
    So. Some more intensive testing on Steadfast Shield, and some comparisons to other shields.

    But first, a special thanks(sarcasm) to Cirrosiss, for stealing my test mobs, twice. And making this much more difficult and time consuming than it should have been. Jerk. A PLD, screwing with my shield testing... wtf.

    Test conditions.
    MOB:Mourning Crawler, lvl 105(lvl verified via battlemod. <3 Byrth)
    99PLD/DNC
    Shield Skill 442
    PDT 38%

    These conditions were maintained across all tests.

    Steadfast shield
    Data
    Spoiler: show

    Code:
    Damage Taken
    Melee Damage Taken
    Player             Melee Dmg   Melee %   Hit/Miss   M.Low/Hi    M.Avg  #Crit  C.Low/Hi   C.Avg     Crit%
    Martel                822714   94.35 %  14231/739     11/112    56.53    293    27/185  118.58    2.06 %
    
    Martel
      Melee
             11:   34
    +        12: 1951
             13:  559
             14:  391
             15:  483
             16:  229
             17:  391
             18:  241
             19:  489
             20:  206
             21:  333
             22:   79
             67:   20
             68:   93
             69:   77
             70:  659
     ^       71: 1171
             72:  545
             73: 1174
             74:  108
             75:  168
             76:   83
             77:  152
             78:  149
             79:   62
             80:  183
             81:   74
             82:  177
             83:  101
             84:  184
             85:  194
             86:  111
             87:  159
             88:   81
             89:  170
             90:   89
             91:  165
             92:  164
             93:   79
             94:  195
             95:   90
             96:  172
             97:  102
             98:  157
             99:  190
            100:   91
            101:  188
            102:   84
            103:  178
            104:   78
            105:  157
            106:  165
            107:   74
            108:  111
            109:   55
            110:   53
            111:   11
            112:    9
      Melee Crits
             27:    5
             28:    7
             29:    7
             30:    5
             31:    7
             32:    2
             33:    7
             34:    4
             35:    8
             36:    9
    +        37:   10
             38:    7
             39:    8
             40:    5
             41:    1
             42:    2
            127:    1
            128:    1
            131:    2
            133:    2
            134:    4
            135:    5
            137:    1
            138:    6
            139:    2
            140:    9
            141:    4
            142:    4
            143:    3
            144:    1
            145:    6
     ^      146:    5
            147:    6
            148:    7
            149:    4
            150:    5
            151:    3
            152:    3
            153:    1
            154:    4
            155:    3
            156:    3
            157:    3
            158:    1
            159:    5
            160:    4
            161:    5
            162:    1
            163:    5
            164:    6
            166:    5
            168:    7
            169:    2
            170:    3
            171:    5
            172:    1
            173:    2
            174:    1
            175:    4
            176:    2
            177:    9
            178:    6
            179:    5
            180:    4
            181:    1
            182:    6
            184:    9
            185:    2

    Code:
    Summary
    Total hits:14,231
    Blocked:5,480
    Non-blocked:8,751
    Block%	38.5%
    
    AVG Blocked DMG:15.28
    AVG non-Blocked DMG:83.18
    Block DMG%-	-81.6%
    
    Total Block DMG reduction	-31.4%*
    
    *this value is before the PDT on the shield itself. As this is the DMG- value purely from blocking.
    The block dmg- for steadfast lines up perfectly with size3 dmg reduction. 50% for size 3 base. 6% for 99 PLD shield defense bonus. And 25% from the 50 def. 50+6+25=81.

    Block rate is best assessed after looking at other shield's data.

    Next up, some comparative data from other shields. And since there was some question about the accuracy of mob lvls in my old tests, I'll be using new test data. With mob lvls verified by battlemod.

    Adamas
    Data
    Spoiler: show

    Code:
    Damage Taken
    Melee Damage Taken
    Player             Melee Dmg   Melee %   Hit/Miss   M.Low/Hi    M.Avg  #Crit  C.Low/Hi   C.Avg     Crit%
    Martel                191187   96.83 %   3289/171     16/109    57.12     65    36/179  108.28    1.98 %
    
    Martel
      Melee
             16:   16
    +        17:  482
             18:   86
             19:   32
             20:   68
             21:   80
             22:   37
             23:   86
             24:   91
             25:   41
             26:   90
             27:   31
             28:   84
             29:   35
             30:    2
             65:    1
             66:    8
             67:  143
     ^       68:  235
             69:  145
             70:  267
             71:   40
             72:   18
             73:   44
             74:   35
             75:   16
             76:   39
             77:   17
             78:   46
             79:   44
             80:   15
             81:   44
             82:   47
             83:   16
             84:   49
             85:   20
             86:   18
             87:   34
             88:   46
             89:   15
             90:   48
             91:   15
             92:   46
             93:   32
             94:   18
             95:   44
             96:   42
             97:   22
             98:   35
             99:   46
            100:   22
            101:   31
            102:   44
            103:   15
            104:   40
            105:   12
            106:   25
            107:   19
            108:    4
            109:    1
      Melee Crits
             36:    1
             37:    3
             42:    1
             43:    1
             45:    1
             46:    3
             47:    1
             48:    4
    +        49:    6
             50:    2
             51:    3
             53:    2
            128:    1
            130:    1
            132:    1
            134:    1
     ^      138:    2
            139:    1
            141:    3
            143:    2
            144:    1
            147:    2
            150:    1
            151:    1
            153:    1
            154:    2
            155:    1
            156:    1
            159:    2
            168:    1
            169:    3
            171:    1
            173:    1
            174:    1
            175:    1
            176:    2
            177:    1
            178:    1
            179:    1

    Code:
    Summary
    Total hits:3,289
    Blocked:1,289
    Non-blocked:2,000
    Block%	39.19	
    
    AVG Blocked DMG:21.37
    AVG non-Blocked DMG:81.81
    Block DMG%-	-73.8
    
    Total Block DMG reduction	-28.9
    Block dmg- breakdown. 50(base)+6(SDB)+17.5(def/2)=73.5. Compared to the 73.8 result.

    Considering that the 0.69% difference between Steadfast and Adams is very well within variance for adamas sample size, I'd say they have the same base block rate. Note, Adamas is also a completely normal size 3 shield in terms of block rate and block dmg reduction.

    I was actually going to get some data on weathering shield, just to reinforce that point. But some jerk kept taking my test mobs, and I didn't feel up to restarting. So, we'll just have to move on to Aegis.

    Aegis(95)
    Data
    Spoiler: show

    Code:
    Damage Taken
    Melee Damage Taken
    Player             Melee Dmg   Melee %   Hit/Miss   M.Low/Hi    M.Avg  #Crit  C.Low/Hi   C.Avg     Crit%
    Martel                167480   94.71 %   3144/153     12/113    52.23     68    28/191  100.46    2.16 %
    
    Martel
      Melee
    +        12:  480
             13:  124
             14:  131
             15:  113
             16:   67
             17:  124
             18:   66
             19:  107
             20:   56
             21:   98
             22:   40
             68:    5
             69:    7
             70:  110
     ^       71:  230
             72:  121
             73:  227
             74:   17
             75:   38
             76:   23
             77:   34
             78:   20
             79:   19
             80:   31
             81:   15
             82:   30
             83:   20
             84:   45
             85:   44
             86:   15
             87:   34
             88:   20
             89:   33
             90:   15
             91:   26
             92:   26
             93:   24
             94:   34
             95:   19
             96:   23
             97:   20
             98:   39
             99:   30
            100:   14
            101:   38
            102:   17
            103:   28
            104:   19
            105:   22
            106:   36
            107:   21
            108:   33
            109:   15
            110:   18
            111:    8
            112:    6
            113:    1
      Melee Crits
             28:    2
             29:    3
             30:    4
    +        31:    5
             32:    1
             35:    3
             37:    1
    +        38:    5
             39:    1
    +        40:    5
             41:    1
            129:    1
            131:    1
            136:    1
     ^      137:    1
            138:    3
            139:    1
            141:    1
            143:    1
            144:    1
            147:    1
            148:    1
            149:    1
            151:    1
            152:    2
            153:    2
            154:    2
            156:    1
            157:    1
            159:    2
            161:    1
            169:    1
            171:    1
            173:    1
            174:    1
            175:    2
            177:    1
            180:    2
            185:    1
            191:    1

    Code:
    Summary
    Total hits:3144
    Blocked:1,437
    Non-blocked:1,707
    Block%	45.7%
    
    AVG Blocked DMG:15.6
    AVG non-Blocked DMG:84.97
    Block DMG%-		-81.6%
    
    Total Block DMG reduction: -37.2%
    A 7.2% difference in base block rates(Aegis vs Steadfast) is way outside the variance range for these sample sizes. Clearly Putting Aegis in a different block rate range than Steadfast. Amusingly, the block dmg reduction values of the two are identical, if from different sources.

    Aegis: 55%(base)+6%(SDB)+20%(DEF)=81%.
    Steadfast: 50%(base)+6%(SDB)+25%(DEF)=81%.

    A chart for some perspective.
    Code:
    Shield			Skill		Block%		DMG-		Total DMG-
    Adamas			442			39.19%		-73.8%		-28.9%(-33.9% w/ shield DT)
    Steadfast(base)	442			38.5		-81.6%		-31.4
    Steadfast*		452			41.0%		-81.6%		-33.4(-40.4% w/ shield DT)
    Aegis			442			45.7%		-81.6%		-37.2%
    Ochain**		N/A			100%		-66.1%		-66.1%
    
    *For practical comparison, block rate was increased appropriately for the +10 skill on the shield. Likewise, the DT was added to the total DMG- and displayed.
    **Using old test data for Ochain. But it shouldn't make any difference, as Ochain was capping on these mobs at 433 skill. Which is being exceeded by 9 skill for every test.
    Adamas has completely fallen behind, so we'll just move on to Aegis/steadfast

    Aegis Vs. Steadfast

    IF, you're making use of at least 4%~ or Steadfast's DT-, then Steadfast is Physically superior or Aegis. If you're already at capped PDT(PDT set or Turtle mode) then Aegis is ahead via it's higher block rate. But it's not like the DT lets you cap when you couldn't before. So it's basically better PDT when in TP gear or casting. Doesn't raise your max possible DMG reduction.

    But, physical dmg- hasn't been Aegis thing since Ochain came out. Aegis Is about the magic dmg-. Aegis is physically comparably to steadfast, with damn near magic dmg immunity. Thus I'd still rate Aegis as the overall better shield.

    Ochain...

    Is curbstomping Steadfast. There's just no contest. To surpass Ochain, you'd have to push steadfast's block rate over 81%. Which just isn't happening on these mobs. I suppose with the DT, steadfast has an advantage against magic... but, were you having trouble capping MDT? <,<;

    The shield Hierarchy hasn't really changed. Ochain and Aegis still have their uncontested specialties.

    TLDR: Steadfast shield is a good, easy to acquire shield. But it's nothing special/unique. I would say it's currently either the 3rd or 4th best shield atm(I haven't looked into Killedar shield.)

  2. #2
    Old Merits
    Join Date
    Sep 2010
    Posts
    1,094
    BG Level
    6
    FFXI Server
    Ragnarok

    Steadfast test with new +64 Shield skill.

    Some small tests on this. A lot of stuff is messed up on the test server(no spellcast....) but Motenten fixed kparser for it damn fast, Thank you!

    Note: Due to battle mod not working on the test server, I couldn't verify the mob levels beyond "IT." I'm assuming 104~105.

    Steadfast Shield - 496 Skill - ItemLevel 105
    Code:
    Summary
    Total hits:1,570
    Blocked:756
    Non-blocked:814
    Block%	48.15%
    
    AVG Blocked DMG:12.31
    AVG non-Blocked DMG:67.89
    Block DMG%-	-81.8%
    
    
    
    Total Block DMG reduction	-39.3%*
    
    *this value is before the PDT on the shield itself. As this is the block DMG- value purely
    So, gained about 10% block rate with the addition of +54 skill(the shield originally had +10 now +64)

    This puts Steadfast slightly ahead of Aegis, even before the DT on the shield itself. Still doesn't touch Ochain.

    Next test. How high can we get this block rate?

    Steadfast Shield - 539 Skill - ItemLevel 105
    Code:
    Summary
    Total hits:1581
    Blocked:931
    Non-blocked:650
    Block%	58.88%
    
    AVG Blocked DMG:14.05
    AVG non-Blocked DMG:
    Block DMG%-	-81.5%
    
    
    Total Block DMG reduction	--47.9%*
    
    *this value is before the PDT on the shield itself. As this is the block DMG- value purely
    Unfortunately, I'm lacking a lot of the newer(using that word generously) shield skill+ gear. And I forgot a piece of what I do have. So this test is on +43 skill over the last one.

    Even so, gained another 10% block rate from +43 skill. 58.88% block rate. I probably should have done a test on low level mobs, to see how high I could push block rate, and check if it caps. But I didn't think about it at the time. Maybe after this stuff hits the real servers.

    The logical next step is to test a Killedar shield. +215 skill.... I wanna see where that puts block rate. Perhaps it'll hit the cap for size three on these mobs..

    Parse data:
    Spoiler: show

    Code:
    Shield test - Steadfast Shield - Mourning crawler lvl ~105(IT) - 496 Skill - Ilevel 105
    Melee Damage Taken
    Player             Melee Dmg   Melee %   Hit/Miss   M.Low/Hi    M.Avg  #Crit  C.Low/Hi   C.Avg     Crit%
    Martel                 64580   91.34 %    1570/76       9/89    39.52     55    22/151   85.49    3.50 %
    Martel
      Melee
              9:   44
    +        10:  323
             11:   53
             12:   52
             13:   56
             14:   63
             15:   62
             16:   53
             17:   28
             53:    1
             54:    9
     ^       55:   20
             56:   25
             57:  117
             58:  112
             59:   96
             60:   16
             61:   15
             62:   19
             63:   14
             64:   14
             65:    5
             66:   14
             67:   24
             68:   13
             69:   11
             70:   17
             71:    9
             72:   15
             73:   15
             74:   14
             75:   20
             76:   15
             77:   14
             78:   22
             79:   14
             80:   17
             81:   11
             82:   13
             83:   15
             84:   14
             85:   16
             86:    8
             87:    2
             88:    3
             89:    2
      Melee Crits
             22:    1
             23:    2
             24:    1
             25:    2
             27:    1
             28:    2
    +        29:    6
             30:    3
             31:    2
             32:    2
            106:    1
            107:    1
            108:    1
            109:    2
     ^      111:    1
            113:    1
            115:    3
            116:    1
            117:    1
            118:    1
            119:    2
            120:    2
            121:    2
            126:    1
            127:    1
            128:    1
            131:    1
            132:    1
            134:    1
            135:    3
            140:    1
            143:    1
            145:    1
            146:    1
            151:    1
    		
    Shield test - Steadfast Shield - Mourning crawler lvl ~105(IT) - 539 Skill - Ilevel 105
    Melee Damage Taken
    Player             Melee Dmg   Melee %   Hit/Miss   M.Low/Hi    M.Avg  #Crit  C.Low/Hi   C.Avg     Crit%
    Martel                 62714   93.63 %   1581/134     10/100    38.76     39    24/167   75.54    2.47 %
    
    Martel
      Melee
             10:  104
    +        11:  267
             12:   66
             13:   38
             14:   73
             15:   79
             16:   78
             17:   64
     ^       18:   67
             19:   60
             20:   11
             61:   79
             62:   36
             63:   75
             64:   38
             65:   14
             66:   13
             67:   10
             68:    8
             69:   18
             70:   11
             71:   10
             72:    8
             73:   12
             74:   17
             75:    8
             76:   20
             77:   16
             78:    8
             79:   16
             80:   10
             81:    7
             82:    6
             83:   13
             84:   14
             85:   15
             86:   12
             87:    9
             88:   14
             89:    3
             90:   10
             91:   15
             92:   15
             93:   14
             94:    8
             95:   15
             96:    9
             97:   10
             98:   10
             99:    7
            100:    2
      Melee Crits
             24:    2
             25:    1
    +        26:    4
             28:    3
             29:    3
             31:    3
             32:    1
     ^       33:    3
             34:    1
             35:    2
             36:    1
            127:    1
            132:    1
            135:    1
            140:    1
            142:    1
            143:    2
            152:    1
            153:    1
            154:    1
            159:    2
            163:    1
            165:    1
            167:    1


    If anyone wants to bring a Killedar shield to the test server for me... XD If not I may get one eventually, but ppl seem to be hoarding Gabranth horns atm. Or overpricing them.

  3. #3
    Old Merits
    Join Date
    Sep 2010
    Posts
    1,094
    BG Level
    6
    FFXI Server
    Ragnarok

    Killedar Shield(now with 215 more skill!) <,<

    Edit for clarity:Atm(7/22/2013) This post only applies on the test server. Should get these adjustments next update.

    So a killedar shield popped on the AH shortly after I finished my last test posts. lol. Back to work!

    Killedar Shield
    Code:
    Mob				Skill	Block%	DMG-%	TDT-%
    Mourning Crawler(~105?)		647	84.23%	-87.0%	-73.2%
    Mourning Crawler(~105?)		703	94.17%	-86.3%	-81.2%
    Korrigan(72~74)			647	98.25%	N/A	N/A
    Korrigan(72~74)			682	100.0%	N/A	N/A
    Oh, ya gone and done it now SE...

    Well, firstly, it's immediately apparent that there is NOT a hard block cap. Throw enough shield skill on something and.... Honestly, I'm more confused than ever about how the skill/block rate formula works.

    Now for the part everyone's actually thinking about.

    Ochain vs Killedar
    Yeah, Ochain just got screwed.

    Here's a lil comparison of various shields physical defenses. 105 mob, base+merited skill, plus any skill on the shield itself.
    Code:
    Shield				Skill	Block%	DMG-%	TDT-%
    Ochain				432	100.0%	-66.1%	-66.1%
    Aegis				432	43.5%	-81.0%	-35.2%
    Steadfast			496	48.15%	-81.0%	-39.3%
    Killedar			647	84.23%	-87.0%	-73.2%
    So, we have Killedar shield 7.1% TDT ahead of Ochain. Before accounting for Killedar's PDT.. But let's assume you're capping PDT anyway. Not hard to do.

    Consider that Killedar can fully benefit from Reprisal. I have little doubt that Reprisal would cap Killedar from 84.23% block rate. 15.77% block rate increase. fairly sure reprisal can manage that.

    Now we step it up. Killedar shield +1. -.-; It has 13 more skill than the NQ(roughly 2% block rate) 1 more def, and PDT. The def will directly add 0.5% to the shield block dmg- I'd expect the HQ to be at about -75.4% TDT-

    The future for Ochain(and a bit on Aegis)
    SE specifically stated that shields and instruments weren't getting any updates. So things already look bleak.

    But, the reason stated for that, was that they were still the best in their respective specialties. /hope! Perhaps after this adjustment, SE will reconsider that stance on shields. Although that reasoning could exclude Aegis, as it's still flat out best for magic.

    However... Even if we do get Updates to Ochain, simply adding Ilvl and a ton of shield skill isn't gonna do much. Not that I mind auto capping block rate on anything, but the issue here is Ochain's comparatively low block dmg-. 66.1% vs 87.0%. As other shields begin to catch up with Ochain's block rate, it's low dmg- makes it fall behind. It either needs a boost in DEF, or in base block dmg.

    Aegis on the other hand, would benefit immensely from that simple update. It wouldn't catch up to Killedar, but I'd only be about 5.6% TDT behind. Combined with it's magic defenses, it'd make for a pretty incredible all around shield.


    At some point I plan to do some higher lvl mob tests. I've seen some 109 mobs around. I dunno if it'll change anything or not, but more data is always nice.

  4. #4
    Ridill
    Join Date
    Apr 2011
    Posts
    23,736
    BG Level
    10
    FFXI Server
    Bahamut

    Might there still be block caps but Killedar like ochain goes past what other shields have? I thought previous testing with low lvl mobs showed that eventually skill stopped doing anything. It does seem odd that SE is basically going back on it's word though... wonder if they even knew they did that? Though wonder if they consider being able to full time 100% block rate without ja/spells and/or tons of skill making it still the best in it's specialty.

    Always thought it was odd too that aegis/ochain gained no defense with trials and as such gain very little physical defense over the levels to begin with

  5. #5
    Old Merits
    Join Date
    Sep 2010
    Posts
    1,094
    BG Level
    6
    FFXI Server
    Ragnarok

    It's possible that that's the case. But there's really no way of telling. I can't get +215~ skill via anything but that shield. 700 freaking skill when I stacked skill+ gear with Killedar... <,<

    Hmmm, I suppose I could check steadfast on low lvl mobs and see if it exceeds the block rates from previous normal shield tests as well... But for all I know this possible cap breaking could be an effect of having item lvl on the shield. So said test might not even tell me anything about skill/block rate calcs.

    It may be that there's some sort of cap on mob lvls lower than the player, and at some point you cease to benefit from lower mobs levels. Or there could be some kinda tier system in place, gain x amount of skill to continue gaining block rate after hitting the soft cap at x skill. /shrug Even after soooo much freaking testing, I still have no idea how this shit really works.

  6. #6
    Ridill
    Join Date
    Apr 2011
    Posts
    23,736
    BG Level
    10
    FFXI Server
    Bahamut

    Speaking of block testing you ever get around to reprisal/palisade rates?

  7. #7
    Old Merits
    Join Date
    Sep 2010
    Posts
    1,094
    BG Level
    6
    FFXI Server
    Ragnarok

    I've made several attempts, but never got a satisfactory clean sample size. Reprisal testing is a huuuge pain in the ass.

    If you parse, you either have to

    one:Maintain full time reprisal

    For this you need a ton of haste. Haste+marchx2(or embrava but that option sucks now.) +gear will usually do it. And I don't have a BRD alt. Then you have to not get interrupted. Which you wouldn't think would be a problem on a one second cast. But it happens a lot on 105 mobs... get interrupted and then you've got your 3+ mobs spam mele'ing you with reprisal down... skewing the acc of your data. Makes it a bitch to automate too.

    Two: Start and stop parse whenever reprisal goes up/down. This is fucking maddening.maybe if i could bind a key to stop/resume. But doing this manually drives me nuts in very short order.

    Alternative options:

    Parser changes:

    #1: There's a def by time tab in kparser, but it doesn't track reprisal. Even if it did, it only displays avgs, irrc. I need the details so I can separate block from non block individually.(hmm, well, maybe I could do something with the total hits taken during reprisal, then use the # of reprisal spikes for blocks? still, doesn't track reprisal. This also wouldn't work for palisade.)

    #2: Packet based parser. It's possible to tell whether a hit was a block or not from the block animation packet received. A packet based parser could easily parse block rates, then track rate during reprisal or palisade.

    Neither of these is likely atm, as it requires someone to alter or make a program. <,<;

    Battlemod, auto exec, and logger method

    It's possible I could do this by using battle mod to see blocks, the having auto exec send an echo whenever reprisal is gained/lost. The go through the logs from logger and section out all the reprisal up sections and find/count the blocks/total hits.

    But that'd take an insane amount of effort on my part. Id' spend more time processing the data than gathering it.

    Maybe someone else could program something that would search for and count the correct block/hit msgs but only the ones between reprisal activation and los msgs. But that's waaay beyond me.


    Anyway, pardon the rant. But I've spent a great deal of time trying to test reprisal/palisade and got very little result. And even less of it reliable.

    /rantoff

  8. #8
    Ridill
    Join Date
    Apr 2011
    Posts
    23,736
    BG Level
    10
    FFXI Server
    Bahamut

    Hmmm well assuming you could do it with lower lvl mobs you could do it with just gear haste and haste spell and some fast cast atmas /rdm I think. Might be possible in soa areas too if you go all out with haste/fast cast gear and ionis maybe. Not sure. Which would allow you to automate an aquaveil too

  9. #9
    Old Merits
    Join Date
    Sep 2010
    Posts
    1,094
    BG Level
    6
    FFXI Server
    Ragnarok

    Mob level 106~109 shield testing

    40,759 hits of testing later... I have this.

    Code:
    Block rates by mob level. 432 Shield skill
    Mob LVL		 106	 107	 108	 109
    Ochain		98.57%	97.23%	94.96%	93.66%
    Aegis		40.87%	40.38%	38.62%	39.08%
    Killedar*	35.18%	34.58%	33.19%	28.73%
    
    *At the time of this test, Killedar has no skilll+.
    All samples were 3k hits+
    Ochain lost 6.3% block rate from 105~109. And Ochain is supposed to lose 30% block rate on 110~ mobs? Glitched! fix it. Don't band-aid it by nerfing new shields(although, really, I'm fine with Ochain staying best too, lol.) fix the 110~ issue.

    Aegis is being weird. It lost ~3% block rate going from 105 to 106 mobs. Then just kinda stayed there...

    I redid the 108 test again, and got 37.31% block rate. I dunno. Either there's a block rate floor, and/or variance is just screwing with me. Maybe these are actually 40, 39, 38, 37. And they're just landing on extreme ends of the ranges for 3k samples.

    Although, in that case, it initial drop from 105 to 106 is kinda weird. Maybe it's in 1.5% increments? 41.5, 40.0, 38.5, 37.

    Anyway, my head hurts. If anyone else feels like analyzing this, have at! Also, I don't really feel like posting 13 parses worth of data unless someone actually plans to use it.

  10. #10
    Old Merits
    Join Date
    Sep 2010
    Posts
    1,094
    BG Level
    6
    FFXI Server
    Ragnarok

    Skill+ shields, post update.

    I actually did these test the day of the update. Just kept forgetting to write up a post.

    So, killedar now has +100 skill, rather than +218.

    Code:
    105 mobs.
    Shield				Skill	Block%	DMG-%	TDT-%
    Steadfast			517	52.09%	-81.0%	-42.6%
    Killedar			532	60.82%	-87.0%	-52.9%
    So losing 118 skill cost killedar 23.41% block rate

    So, how does this affect the current shield hierarchy?
    Code:
    Shield				Skill	Block%	DMG-%	TDT-%
    Ochain				432	100.0%	-66.1%	-66.1%
    Killedar			532	60.82%	-87.0%	-52.9%
    Steadfast			517	52.09%	-81.0%	-42.6%
    Aegis				432	43.5%	-81.0%	-35.2%
    The changes put Ochain on top again. Although, depending on reprisal's potency, it could put killedar pretty close. 15%~ block rate would put killedar about even with Ochain, on these mobs. Not sure how things would go on higher level stuff.

  11. #11
    Old Merits
    Join Date
    Sep 2010
    Posts
    1,094
    BG Level
    6
    FFXI Server
    Ragnarok

    Beatific shield tests.

    Edit: Errors were later found in my data processing for this test. Corrections and new Beatific values here.

    Effing finally.

    First, massive thanks to Keitaro, for writing me a log parser capable of pulling a list of blocked/non block hits and their dmg out of my log files. Manually searching and collecting the data from logs was freaking painful. And took so long I gave up on using the log data I'd already gathered twice. <,< But now tests for size 1 and 2 shields are relatively simple and painless. Thanks Kei!

    So, data.

    Mob:Serac Rabbit
    Lvl: 111
    Shield Skill: 504

    Total Hits 2073
    Block% 56.92
    DMG%- -50.1
    Total PDT -28.5

    The sample here is a bit small, but I'll probably be doing more tests later now that it doesn't suck horribly.

    I was surprised at the dmg- value. It was quite a bit higher than I'd expected. And I honestly wondered if my test wasn't bad. But after breaking it down it made more sense.

    45 def shield. 22.5 dmg- from def. That leaves 27.5. 6% of that should be from shield def bonus. So 21.5% left. This seems like a fairly Reasonable base dmg- for a size1 shield. It is a somewhat odd number, but the sample is small, so the actual value could easily be 20%.

    The block rate is high. But... not as high as I thought it'd be. On the same mobs, Killedar had a 47.64% block rate. 9.28% less than Beatific. Considering size 1 shields should have higher base block rate than size 3, and Beatific's unique "Increases chance of successful block" Stat, I'd expected a larger gap. Although, killedar does have a 28 skill advantage. Perhaps I need to look into and quantify the base block rate differences between the shield sizes...

    So, where does this put Beatific in comparison to other shields(for physical dmg reduction)?

    Using 111 mobs...
    Code:
    Shield	Block%	DMG-	TPDT
    Ochain	90.98%	-66%	-60.0%
    Killedar47.64%	-86%	-40.9%
    Beatific56.92%	-50%	-28.5%
    Aegis	34.54%	-81%	-27.9%
    Beatific is actually slightly better than Aegis for physical dmg reduction. But then, Aegis hasn't been impressive in that particular field since Ochain was added. Beatific is still miles behind Killedar, and even further from Ochain.

    I have to admit, it's not as bad a shield as I thought it would be. It's not impressive, and there are far better, easily acquired options. But it's not totally worthless defensively.

    Although, this doesn't really change my original conclusions about what this shield is good for.

    Shield mastery TP gain(although, Ochain's better. Even if Beatific has +2 shield mastery, +10 TP on every hit is better than +12 on 57% of incoming hits.)

    Skillup. Again, Ochain's better, but it's a fairly unusual case to have Ochain before you cap shield. <,<

    And blocking WS add effects(or ironclad-esque melee add effects) in the absence of Ochain. In this case, the dmg reduction takes the back seat to just blocking as much as possibly to prevent nasty status.

    I still need to quantify the "Increases chance of successful block" Trait on Beatific. But I'll need other size 1 data to figure that out. And I'll want more Beatific data to make sure my values are accurate.

  12. #12
    Old Merits
    Join Date
    Sep 2010
    Posts
    1,094
    BG Level
    6
    FFXI Server
    Ragnarok

    Block rates vs lvl 126 mobs

    Reposting from over on AH.
    Quote Originally Posted by Martel
    So. Level 126 mobs. Block rates suuuck.

    Target: Serac rabbits(woh gates)
    Level 126.
    The only skill+ gear would be on the shield itself, if present.
    Code:
    Shield		Skill	block%	-dmg	totaldmg-
    Ochain		432	70.23	-66%	-46.6%
    Killedar	532	27.63	-86%	-23.7%
    Beatific	504	36.72	-48%	-17.6%
    Aegis		432	11.91	-81%	- 9.6%
    Weathering	432	 6.18	-74%	- 4.5%
    
    Sample sizes
    Shield		hits
    Ochain		4929
    Killedar	3980
    Beatific	3602
    Aegis		4172
    Weathering	1504
    Getting interrupted like mad while wearing Ochain is really annoying. I guess SE's way of nerfing Ochain, was to make new mobs so freaking high level, you feel like you're playing lvl 75 PLD again. -.-;

  13. #13

    Has anyone done any testing on the block rate of the newly released Blurred Shield or Nibiru Shield? Very curious if they are viable options or if it's still just Aegis/Ochain only, and anything else is inventory-1.

  14. #14
    Old Merits
    Join Date
    Sep 2010
    Posts
    1,094
    BG Level
    6
    FFXI Server
    Ragnarok

    I haven't seen any testing thus far.

    I could run the numbers and tell you exactly where they stand except for a few issues. BG wiki doesn't list the shield size for Nibiru Shield, and we don't know the other augment paths. And blurred... is listed as a size 2, which for that model, and that defense, is freaking weird. I'm inclined to think there was a mistake of some kind and that wasn't accurately dat mined.

  15. #15

    Thanks, Martel.

    Is KParser working? If so, I could do some testing myself, if you want to PM me the methodology you have used in the past.

  16. #16
    Old Merits
    Join Date
    Sep 2010
    Posts
    1,094
    BG Level
    6
    FFXI Server
    Ragnarok

    I haven't checked recently, but there's been an update, so kparser's probably dead again. The most recent post by the guy who's been unofficially maintaining Kparser and posting new memlocs was in April.

    But anyway, all we really need here is for someone to double check the dat mines for the shield sizes and then find out remaining augment paths for Nibiru. It's pretty unlikely that there's some hidden, block rate/dmg- related, aspect to either of these shields.

    Once we have the shield sizes for certain, and we know if Nibiru has any relevant augs, the rest can just be calculated and compared to existing shields.

  17. #17

    Once I have the needed silts, I'll start checking the other Nibiru paths. Will post them up here.

  18. #18

    Nibiru Shield R15 Augments --

    Path A: [1] HP+80 [2] Enmity+7 [3] Phys. dmg. taken -5%
    Path B: [1] HP+80 [2] MP+80 [3] "Fast Cast"+7
    Path C: [1] MND+15 [2] "Cure" potency +5% [3] Potency of "Cure" effect received +5%
    Path D: [1] Accuracy+15 [2] Attack+15 [3] "Store TP"+5

  19. #19
    Old Merits
    Join Date
    Sep 2010
    Posts
    1,094
    BG Level
    6
    FFXI Server
    Ragnarok

    Well, still no confirmation on shield sizes, so I'm just going to do this under the assumption that both Nibiru and Blurred shields are size 3.

    Keep in mind that the numbers here(with the exception of 126 Priwen/Ochain base) are projections using the shield formulas we know, and not actual test results. That being said, they should be pretty accurate anyway.
    Code:
    level 126 Priwen Reprisal Nibiru Reprisal Blurred Reprisal Blurred+1 Reprisal Ochain Reprisal
    Block%    30.21  90.63    30.21  45.31    35.21   52.81    40.21     60.31    70.23  100
    DMG%-     -88    -88      -88    -88     -86      -86      -88       -88      -66    -66
    Total PDT -26.5  -79.7    -26.5  -39.8   -30.2    -45.4    -35.3     -53      -46.3  -66
                    
    level 119 Priwen Reprisal* Nibiru Reprisal Blurred Reprisal Blurred+1 Reprisal Ochain Reprisal
    Block%    40.71  100       40.71  61.06    45.71   68.56    50.71     76.06    70.23  100
    DMG%-     -88    -88       -88    -88      -86     -86      -88       -88      -66    -66
    Total PDT -35.8  -88       -35.8  -53.7    -39.3   -58.9    -44.6     -66.9    -46.3  -66
    
    *As I've mentioned before, Priwen with Reprisal up doesn't actually cap block rate, even though the Reprisal formula SE gave us shows that it should. 
    The block rate is quite high, but not capped.
    Blurred+1 is pretty good. It inches ahead of Ochain with Reprisal up on ~119 mobs. But Priwen still crushes it. And on higher level stuff Ochain stomps it too.

    I think the best thing about Blurred+1 is that it's a really good DD piece while still being pretty good defensively. But if you're focused on defense, then just go with Ochain or Priwen. Also, Blurred+1 price is probably gonna be... scary.

  20. #20
    Old Merits
    Join Date
    Sep 2010
    Posts
    1,094
    BG Level
    6
    FFXI Server
    Ragnarok

    Shield size 1 & 2 base block rates.

    Hell of a necro there, but this seemed like the right thread for this.

    With the recent updates to the BGwiki shield skill page, I wanted to fill in a few of the last blanks in the general shield size info chart. In this case, Size 1 and 2 base block rate.

    Test conditions.
    99 PLD/RDM
    440 shield skill
    No block+ gear.

    Mobs: Blanched Mandragora lvl 100
    Shields used

    Shield (size)
    Bloodbead Ecu (2)
    Simba Buckler (1)
    Gold Buckler(1)
    Aegis (5)
    Adamas (3)

    Aegis and Adamas tests were just for comparison on these mobs. Since size 3 and 5 base rates are already known if should help give an idea of how to adjust the other rates to reach the base.

    text version
    Code:
    Shield	        Block%	S.size	hits*	blocks	sample   +/-%
    Bloodbead Ecu	40.44%	2	3788	2572	6360	1.21%
    Simba Buckler	55.11%	1	2598	3189	5787	1.28%
    Aegis         	51.37%	5	1313	1387	2700	1.88%
    Gold Buckler	55.09%	1	1231	1510	2741	1.86%
    Adamas	        45.06%	3	1168	958	2126	2.11%
    						
    *Note that 'hits' here refers specifically to unblocked hits. 
    It's just how the parse addon categorizes things.
    The odd part here is that everything seems a bit lower than it should be. 99 PLD with merits vs a lvl 100 mob. I should be about 9 skill ahead. So my block rates should all be 1.93% above the base for that shield.

    Nothing actually hits this in this test. Although the Aegis and Adamas test are within range of base+1.9%.

    Despite this weirdness, I just can't see the base rates for size one and two being something as weird as 53% or 38%. So I'm inclined to say the base block rates are probably the following.

    Size 1: 55%
    Size 2: 40%

    There maybe be some issue with my estimated mob skill, or something else making the predictions weird.

    I've tossed the content of the exported .xml file the parse addon made for this here.
    Spoiler: show
    Code:
    <database>
        <Blanched_Mandragora>
            <Arduwyn>
                <total_damage>0</total_damage>
                <defense>
                    <evade>
                        <tally>4</tally>
                    </evade>
                    <nonblock>
                        <tally>2</tally>
                        <damage>31</damage>
                    </nonblock>
                    <hit>
                        <tally>2</tally>
                        <damage>31</damage>
                    </hit>
                </defense>
            </Arduwyn>
            <Martel-Sim>
                <total_damage>0</total_damage>
                <defense>
                    <hit>
                        <tally>2598</tally>
                        <damage>0</damage>
                    </hit>
                    <block>
                        <tally>3189</tally>
                        <damage>0</damage>
                    </block>
                    <nonblock>
                        <tally>2598</tally>
                        <damage>0</damage>
                    </nonblock>
                    <evade>
                        <tally>408</tally>
                    </evade>
                </defense>
            </Martel-Sim>
            <Martel-Gol>
                <total_damage>0</total_damage>
                <defense>
                    <hit>
                        <tally>1231</tally>
                        <damage>0</damage>
                    </hit>
                    <block>
                        <tally>1510</tally>
                        <damage>0</damage>
                    </block>
                    <nonblock>
                        <tally>1231</tally>
                        <damage>0</damage>
                    </nonblock>
                    <evade>
                        <tally>219</tally>
                    </evade>
                </defense>
            </Martel-Gol>
            <Martel-Blo>
                <total_damage>0</total_damage>
                <defense>
                    <hit>
                        <tally>3788</tally>
                        <damage>2</damage>
                    </hit>
                    <block>
                        <tally>2572</tally>
                        <damage>0</damage>
                    </block>
                    <nonblock>
                        <tally>3788</tally>
                        <damage>2</damage>
                    </nonblock>
                    <evade>
                        <tally>419</tally>
                    </evade>
                </defense>
            </Martel-Blo>
            <Martel-Ada>
                <total_damage>0</total_damage>
                <defense>
                    <hit>
                        <tally>1168</tally>
                        <damage>0</damage>
                    </hit>
                    <block>
                        <tally>958</tally>
                        <damage>0</damage>
                    </block>
                    <nonblock>
                        <tally>1168</tally>
                        <damage>0</damage>
                    </nonblock>
                    <evade>
                        <tally>142</tally>
                    </evade>
                </defense>
            </Martel-Ada>
            <Martel-Aeg>
                <total_damage>0</total_damage>
                <defense>
                    <hit>
                        <tally>1313</tally>
                        <damage>0</damage>
                    </hit>
                    <block>
                        <tally>1387</tally>
                        <damage>0</damage>
                    </block>
                    <nonblock>
                        <tally>1313</tally>
                        <damage>0</damage>
                    </nonblock>
                    <evade>
                        <tally>202</tally>
                    </evade>
                </defense>
            </Martel-Aeg>
        </Blanched_Mandragora>
    </database>

Similar Threads

  1. Shield Skill's effect on block rate
    By Martel in forum FFXI: Advanced Math
    Replies: 23
    Last Post: 2011-11-06, 16:24