Shield Asstery

[BerenTebogo]

Ok. Possibly a shoot me question since I've been a PLD main for so, so, so long but I just can't find any of the specific information I'm looking for.

When it comes to shield blocks we all know the following:
Size 1: Low damage + highest activation rate
Size 2: Slightly higher damage + lower activation rate
Size 3: Most common and ideal shield - good balance of damage + can still cap rate
Size 4: Highest damage absorbed and lowest proc rate.

Let's ignore aegis for the purposes of this question and for the purposes of the description let's just assume 0 shield skill gear / merits so it actually makes sense.

My big question is regarding all shields, although specifically for size 3's, how does 1 base defense added to the shield translate into damage mitigated?

For example: Koenig is DEF22 and Kaiser is 23. Assuming a shield block with Koenig will get you hit for 50 damage will Kaiser lower it to 49? 48? 45? 50 - x? Original damage - DEF x static reduction %? (That last one may not be as clear as it sounds to me)

The point of the question is obviously to help PLDs when /WAR or /RDM or in certain situations /NIN to make the best choice for their situation. Terror shield, for instance, is a very good shield but has a DEF rating of 20, which is 9.x% less than Koenig - how will this translate to damage mitigated?

--

Supplementary questions:

-- We all know how fSTR correlates to weapon base damage for the purposes of melee - perhaps there is a VIT or DEF modifier that will apply to shield? I find this unlikely but is it worth exploring?

-- In terms of potential damage mitigation is there a cap for each shield size? I.E. Can a simba buckler (size 1 DEF24) ever reduce as much damage as Koenig (size 3 DEF22) under specific conditions?

------------

I know I'm asking a lot of questions without providing much but I don't recall ever seeing extensive testing on any of these areas and search has failed me. With all the attention we pay daily to min/maxing for melee I feel I've neglected my PLD/RDM, which remains one of my favorite job combos to play and I know many people feel the same.

[Konda]

use -dmg, it's more reliable

[Yarko]

Taken from an old test posted on allakhazam

-Damage %
Shield Type LV DEF Min Max
Hard Shield 3 48 10 54.7% 55.2%
Koenig Shield 3 73 22 60.6% 61.2%
Acheron Shield 3 75 27 63.2% 64.6%
Looking at the numbers, the damage reduction seems to follow the following equation:

Damage_Absorption = BASE + Shield_Defense / 2

In other words, each point of defense on the shield adds +0.5% to the shield absorption %.
The value of BASE depends on the type of the shield. These are are the numbers I found out.



Type BASE
Buckler 22%
Round 40%
Kite 50%
Tower 55??
Relic 55??


Koenig Shield 50+11 %
Aegis 55+20% ?

[Byrthnoth]

I think the verdict of the last testing on this I saw was that Shield size has a base -% Physical Damage Taken, and every DEF on the shield = -.5% Physical Damage Taken during a shield block on top of the -% you already get from the shield size.
So Kaiser would theoretically block .5% more Physical Damage Taken during a shield block than Koenig.
That said, the differences between shield sizes probably make it so that Simba Buckler will never block as much damage as Koenig Shield. The only time when lower-rank shields approach higher-rank shields is when your proc rate starts to drop, and I don't think it's possible for it to drop enough for PLD for a size 1 shield to ever approach a size 3 shield.

[BerenTebogo]

use -dmg, it's more reliable

I idle in full PDT on PLD/RDM while I have SS up, depending on the situation. Still doesn't mean I'm not curious <_<

[BerenTebogo]

Taken from an old test posted on allakhazam

This is actually exactly what I was looking for. Thanks a ton. I hadn't ever seen this test before.

Edit: I would be curious about some input / testing from Aegis owners if possible. Might help me decide if I really will do Aegis finally after all these years after Bravura (hey, I really like soloing!)

[Yarko]

This is actually exactly what I was looking for. Thanks a ton. I hadn't ever seen this test before.

Edit: I would be curious about some input / testing from Aegis owners if possible. Might help me decide if I really will do Aegis finally after all these years after Bravura (hey, I really like soloing!)

No problem.

Being a long time RDM tank and liking numbers, I've parsed and watched aegis in action closely to "study my rivals", and the -75% damage mitigation seems correct. I've also seen a 40-45% block rate on most HNMs. Never really done any tests myself though.
Note: Assuming a 42,5% block rate and a 75% dmg mitigation, aegis translates to about -32% phys dmg reduction

[BerenTebogo]

I'd venture to say the 45% Block rate you parsed is due to Aegis' disease.

It's a fairly common affliction among Aegis upgraders which causes them to throw away all their shield skill gear resulting in a lower block % than even a koenig. Aegis should cap at 65% Block rate as well which is a much higher phys dmg reduction %.

[Yarko]

Actually the 38-40s% come from PLDs using haste gear (on cerb, faf and odin), the 45% from one full timing koenig head and af feet (on cerb).
I'd like to see more tests about shield block rates myself (especially aegis), see how much of an impact shield skill gear really has, as well as reprisal and the level of the HNM.

[Brutala]

I did some koenig shield tests on EM-T steelshells awhile ago on war/dnc. I planned to test on PLD later to see if the same amount of skill would give the same block rate boost in any skill range on the same mob and I was going to repeat the whole thing on higher level mobs to see how it changes with mob-player level difference, but I quit because it was taking too much time just for a small sample. I'll post it in a min, fucking Sabnock popped. brb

Edit: Found it. It's raw and unprocessed and a very small sample, but I think at least 2 things are certain: AGI does nothing whatsoever, and moon phase messes with block rate. I consistently had the same setup parse one block rate on one moon phase and then a completely different one on another. If I knew it would take so long I would have tested only like 2 sets in that time and maybe actually had a decent sample, and I would have started writing down moon phases sooner. Also, these are some references like "today" and "2 days ago", I'll look over my notes and see which parts of the test each referred to.





WAR/DNC on Steelshells. At first I did only level 76s, later also level 75s. Also, before you criticize me for using too many different gear sets and using a job that doesn’t even tank with a shield:
1 – my aim is to come up with a working formula. If the war tests go as planned and a huge sample actually lets me produce a working formula, I expect to already Know how each PLD test will go before I do it. I expect to actually need to spend more time on figuring out what reprisal does than on confurming the intuitions from WAR tests.
2 – I am not using “too many gear sets”. For example, the first 4 serve only to show that a huge difference in AGI does nothing to block rate on these mobs at both high and low total shield skill. I don’t think I will need to increase the sample on these by much. I want huge samples for sets that differ ~only in shield skill as my main goal is to figure out if adding shield skill provides constant returns in block rate regardless of the difference in level between mob and player. And yes, I know just 75 and 76 mobs aren’t enough for that, I simply want a decent sample on them before trying something much higher (that and I’m putting it off because noone will help me and I don’t want to have to use meds and drinks for 10 hours of testing).

Gear set 1 - +25 shield, alky’s, hauby, PDR
89STR 62DEX 92VIT 55AGI 52INT 55MND 72CHR 263shield 230parry 235eva
-20eva hauby +5 ritter +5 ethereal
48/143 0.3357
70/184 0.3804
56/130 0.4308
174/457 0.3807 (session average)
18 parry
27 int. (this meant intimidate from bibiki seashell)
21 eva.

Gear set 2 - +25 shield, AGI, PDR
71STR 63DEX 88VIT 95AGI 57INT 55MND 72CHR 263shield 230parry 235eva
+5eva hachiryu
50/146 0.3425
61/156 0.3910
42/122 0.3442
153/424 0.3608 (session average)
30 parry
27 int.
85 eva.

Gear set 3 - hauby, alky’ss, no shield gear, used haidate and turban instead of PDR in those slots to get more TP to survive
94STR 82DEX 78VIT 55AGI 52INT 55MND 62CHR 238shield 220parry 225eva
+5eva ritter +5 ethereal -20hauby
38/102 0.3725
32/109 0.2936
32/131 0.2443
102/342 0.2982 (session average)
18parry
19 int.
18 eva

Gear set 4 - AGI, PDR, no shield gear
76STR 68DEX 88VIT 95AGI 57INT 55MND 62CHR 238shield 220parry 225eva
+5eva ritter +5 hachiryu
41/147 0.2789
47/154 0.3052
36/122 0.2951
124/423 0.2931 (session average)
25 parry
26 int.
53 eva

Notice how the first 4 sessions have the 55AGI setups slightly outparsing the 95AGI ones. Either AGI does jack shit for block rates, or AGI figures into block rates in such a way that on a koenig shield on level 76 crabs it is already capped somewhere below 55AGI. The Agi difference definitely gave a huge boost to parry and evade rates.
As for the small difference in other base stats… fuck it. Maybe something other than AGI actually affects block rates, but that needs too much damn testing and I don’t want to handle all that bullshit gear. 6 other stats x 1000 hits in one sample, repeated for different total shield skill values. Fuck off. Maybe I’ll get to stats after I roughly establish something based just on total shield skill and mob/player level difference.

Gear set 5 - +35 shield gear, hauby, dusk, feet, haidate
82STR 83DEX 88VIT 61AGI 52INT 55MND 72CHR 273shield 230parry 235eva
-20eva hauby
34/90 0.3778
49/116 0.4224
87/202 0.4307
51/114 0.4474
221/522 0.4234 (session average)
31 parry
19int.
42eva

Gear set 6 +15 shield gear, hauby, alky’s, haidate
89STR 77DEX 88VIT 55AGI 52INT 55MND 72CHR 253shield
48/135 0.3556
39//116 0.3362
59/165 0.3576
146/416 0.3510 (session average)

Gear set 7 - same as 6, just with dusk gloves, fuck. Don’t ask. This was done the day after the gear set 6 test and I put the wrong gloves on without checking what I wore the day before. When I realized it a was a different gear set I just said fuck it and rolled with it. The sessions parse almost the same.
78STR 83DEX 88VIT 61AGI 52INT 55MND 72CHR 253shield
43/114 0.3772
39/99 0.3939
41/129 0.3178
31/90 0.3444
154/432 0.3565 (session average)

Everything up to this point was T steelshell (lv76), from here on I throw in EMs to waste less time trying to claim. My hope is that given huge enough samples there will be a difference between level 75 and 76 crab block rates and that may be a good start to trying to figure out block rate level correction.

Gear set 3 again, added shallops. I’m so tired of staring at crabs I just started using food so I can safely do other stuff while spamming a waltz macro, then counting hits in the log afterwards. If you feel that 4vit 1dex 1int is going to produce different block rates from the noshallops session you’re welcome to treat this as a different test.
As you can see from the 2 tests above I had already stopped counting eva, parry, intimidate. The low AGI and high AGI test should be sufficient to show that there is no impact of AGI on block rate on this mob. As expected, it had a huge impact on parrying and evasion.
94STR 83DEX 82VIT 55AGI 53INT 55MND 62CHR 238 shield
- I know I started at 10% new moon, no idea about moon after that or whether it was rising or falling. This is getting fucking frustrating because the gear sets that parsed 29% 2 days ago are parsing 32% and higher today. >.< I’ll have to start keeping track of bullshit like moon phase and… fuck it. I’ll assume the results from 2 days ago are a fluke since that sample is smaller than this one.

(all T)
43/128 0.3359
38/96 0.3958
37/115 0.3217
33/110 0.3000
36/100 0.3600
34/112 0.3036
48/133 0.3689
269/794 0.3388 (this session)
371/1136 0.3266 (this gear set’s total so far, including the no-shallops run)
(all EM)
27/84 0.3214
34/85 0.4000
30/77 0.3896
33/82 0.4024
26/82 0.3170
150/410 0.3659 (this session)

Gear set 9. No agi, no shield, full PDR, also no alkys or hauby to reduce agi either.
79STR 69DEX 82VIT 66AGI 53INT 55MND 62CHR 238 shield
36% waxing
41/150 (T) 0.2733
48/159 (T) 0.3019
38% waxing
54/158 (EM) 0.3418
37/108 (EM) 0.3426
31/104 (EM) 0.2981
43/112 (EM) 0.3839
40% FQM
57/154 (T) 0.3701
37/127 (T) 0.2913
64/171 (T) 0.3743
50/154 (EM) 0.3247
43% FQM
49/164 (T) 0.2988
54/147 (EM) 0.3673

296/925 0.3200 (Ts, this session)
269/783 0.3436 (EMs, this session)

Gear set 3+shallops again
94STR 83DEX 82VIT 55AGI 53INT 55MND 62CHR 238 shield
FQM 45%
46/124 (T) 0.3710
417/1260 0.3310 (this set’s total so far, again the no-shallops runs included)
I had to stop at this point.

Gear set um… 10. same as 5, no food etc. but full PDR instead of hauby/haidate. Checking if that DEX made a difference or not, but of course, I won’t know until the sample sizes of both set 5 and set 10 are in the thousands.
82STR 63DEX 88VIT 66AGI 52INT 55MND 72CHR 273shield
Waxin’ 62%
84/217 (T) 0.3871
59/129 (EM) 0.4574
45/139 (EM) 0.3237
Waxin’ 64%
70/194 (T) 0.3608
43/133 (T) 0.3233
55/140 (T) 0.3929
48/127 (EM) 0.3780

252/684 (T average in session) 0.3684
152/395 (EM average in session) 0.3848

-next day, same gear
darksday 93% full moon
57/124 (EM) 0.4597
66/149 (T) 0.4426
firesday 95% full moon
62/148 (T) 0.4189
32/92 (EM) 0.3478 W T F
53/147 (T) 0.3605
earthsday 98% full moon
42/108 (EM) 0.3889
67/150 (T) 0.4467
54/118 (EM) 0.4576
65/173 (T) 0.3757
watersday FM100%
64/134 (T) 0.4776
62/155 (T) 0.4000
53/133 (EM) 0.3985
47/112 (EM) 0.4196

285/687 (EM average in this session) 0.4148
439/1056 (T average in this session) 0.4157



-samples aren’t very big, but in the EM sessions it seems that EMs parsed a few % higher a block rate than Ts. More testing is needed to show the exact difference. I’m really bothered how the same gear sets parse completely different on different rl days but seem fairly consistent within the day.

- Neosutra said, in some shadow mantle vs boxer’s mantle argument somewhere I think, that you can expect about a 3-4% boost to block rate from wearing 10 shield skill. I don’t know if that was based on a test or eyeballing, but it’s ~consistent with parts of my small test so far. Wearing no shield skill gave me about a 32% average block rate today, wearing +35 skill made it about 42% 2 days ago. 35 skill, 10%, margin of error, small sample etc. After I get these WAR samples higher for more consistency I’ll move to PLD testing to see if this trend continues all the way to cap (if I can even cap lol; these T steelshells parsed 42% with 273 skill in small sample sizes, if 3 skill keeps adding about 1%, 319 skill should parse ~57% block rate, not capped without reprisal, even if I upgrade pld boots and get sinfender and knight’s earring.)

- if the level difference penalty to block rate is a straight -% per level, what does that mean for level 95 HNMs? Let’s say, for the sake of argument, Tiamat was level 75 and you capped block rate at 65% at a certain amount of shield skill. A level 95 Tiamat would have you blocking 15% if the penalty was 2.5% per level, 25% if the penalty was 2% per level, 35% if the penalty was 1.5% per level, in that same shield skill setup. Question for PLDs: how are your block rates with koenig shield on level 95 shit with reprisal down? I only ever tanked one Tiamat on PLD about a year or so ago and I wasn’t solo tanking. I don’t even remember getting hit.

-what about mobs and the skill they have in their um.. “weapon”? Maybe your shield skill is compared to mob weapon skill in some way and maybe that’s how the Neosutra 3skill=~1% thing works. Maybe the difference I was observeing between 75 and 76 crabs is partially accounted for by the difference in their “weapon skill”. If we proved that 3 skill gave 1% block rate at any level difference between mob and player, and if my sample was big and the 75-76 difference was still 2.5%, how dumb would it be to say that 1% of the penalty is due to the 76 having 3+ weapon skill more than the 75, and the level difference penalty is an additional 1.5%? 1.5% I could live with because it actually allows the high block rates we get on NMs in the 80s with reprisal down, given that their “weapon skill” isn’t

Conculsions I dare make at this time:
-AGI does jack fuck. Some stat other than AGI might work, needs moar testing.

[Neosutra]

If only people could swap into shield gear when their shadows were down, then back into haste gear for atonement afterwards..

Dare to dream.

[Ferien]

Discussed this a few years ago back here. I was doing tests to figure out proc rates and damage reductions for various shield sizes and posted the numbers here: http://www.bluegartr.com/forum/showthread.php?t=46559

I think there was another thread about it as well around the same time, and there was some (actually useful) discussion on both KI and Alla. I should also still have the script I used to parse through my data and the data Woozie was giving me if anyone wants to take a crack at it on their own.

Edit: To clarify what I was posting in that thread, my initial tests I posted to KI and at the beginning of that thread were eyeballed against the EP Orcs in Monastic Caverns. I stood there and just let them beat me. The set of tests I posted later on were based off of data pulled from one of the logging tools (parsers still sucked a lot at defensive stuff back then) and just run through some mean, median, and mode functions to get the percentage reduction. These were done on EM Steelshells IIRC. These were also all done before Reprisal.

[Raen Ryong]

Conculsions I dare make at this time:
-AGI does jack fuck. Some stat other than AGI might work, needs moar testing.

Nice long detailed post :O this conclusion in particular I find interesting given how many (stupid) Paladins I've seen swearing by Drone Earrings and such.

[BerenTebogo]

Nice long detailed post :O this conclusion in particular I find interesting given how many (stupid) Paladins I've seen swearing by Drone Earrings and such.

I don't think the test is conclusive enough to rule out AGI as a modifier for shield blocks. I'd be interested to getting to a conclusive result though since I've always preferred idling in Cerb bow over Lamian +1 on fights where I won't be taking primarily magic damage (AGI+2 on cerb bow)

[Brutala]

Nice long detailed post :O this conclusion in particular I find interesting given how many (stupid) Paladins I've seen swearing by Drone Earrings and such.

Dunno why I wrote that then. It's pretty dumb to say with a that small a sample and without first having data on something much higher in level, like a 2000 hit sample on Aura Statues or something. It could very well be that whatever AGI does to shield on these crabs is already capped at 55AGI so adding 40more did nothing (though it did change parry and eva rates a lot, but they're completely different formulae).

[Raen Ryong]

I don't think the test is conclusive enough to rule out AGI as a modifier for shield blocks. I'd be interested to getting to a conclusive result though since I've always preferred idling in Cerb bow over Lamian +1 on fights where I won't be taking primarily magic damage (AGI+2 on cerb bow)

I think it's interesting in that it casts doubt at least not as if you can really eyeball what AGI does to shield proc rate so I think people have just been taking it for granite thus far.

Normally if I won't be taking magic damage I idle in Fenrir's Stone for the HP/less CE loss Tiphia Sting if I won't be getting hit much.

[BerenTebogo]

Well if it pans out I wouldn't mind going for the enmity ranged item. As it stands now I value a possible shield block rate+ for a non-changing slot over +1/2 enmity.

[Raen Ryong]

Yah, can't blame you there supposing it does increase shield proc rate. There is an enmity+ item now too? I'm so behind since I went inactive.

[Byrthnoth]

I'm not aware of any tests showing a significant effect of Agility on Parrying or Shield block proc rates. I don't even know where the rumor came from really, except that it's really old.

Btw, you can augment Enmity onto a bow if you try hard enough.

[Tricen]

If only people could swap into shield gear when their shadows were down, then back into haste gear for atonement afterwards..

Dare to dream.

NO. BAD NEOSUTRA. THIS IS BG, NO LOGIC ALLOWED.