Shield block numbers?

[Ferien]

Hm...you know what? As much as I hate to say it, lolAlla may be right about this one. Looking at that post again, he gives the formula as being Absorption = Base% + Def/2, with a Buckler's base as being 22%. Ancile's Def rating is 32. Therefore~ Absorption = 22 + 32/2 = 22 + 16 = 38%, which matches up with my number.

Now I need to make sure that the number I found is actually right...

[Neosutra]

Hm...you know what? As much as I hate to say it, lolAlla may be right about this one. Looking at that post again, he gives the formula as being Absorption = Base% + Def/2, with a Buckler's base as being 22%. Ancile's Def rating is 32. Therefore~ Absorption = 22 + 32/2 = 22 + 16 = 38%, which matches up with my number.

Now I need to make sure that the number I found is actually right...

Thats actually not TOO bad since its a buckler type and activates more. May even outparse koenig for total reduction. Hmm. Have to test this.. The plot thickens.

[Ferien]

Tested cap is around 65% regardless. A Koenig Shield with enough skill will proc at that rate or higher anyway. Simple math tells us that Koenig Shield will be better if it procs more than 40% of the time, and many tests have shown that it procs more often than that.

Since we can get them both to cap pretty easily though, you've got 61% vs 38% at the exact same proc rate...and at that point it's really obvious which wins.

[Neosutra]

Tested cap is around 65% regardless. A Koenig Shield with enough skill will proc at that rate or higher regardless. Simple math tells us that Koenig Shield will be better if it procs more than 40% of the time, and many tests have shown that it procs more often than that.

Since we can get them both to cap pretty easily though, you've got 61% vs 38% at the exact same proc rate...and at that point it's really obvious which wins.

Wait, tested cap for buckler type shields? Or tested cap for Ancile? Where did we go from "hard to test block % based on blocks sometimes being as high damage as normal weak hits" to "this is the cap for ancile".? Data? From the data ive seen, buckler types cap much higher then 65%, on the order of 80%+.

80% X 38% = 30% DoT reduction

Still behind Koenig for DoT reduction, but not "holy crap" behind, since you get the added block % increase for keeping shadows up and such. Too bad it has not other good stats.

[Ferien]

Tested cap is around 65% regardless. A Koenig Shield with enough skill will proc at that rate or higher regardless. Simple math tells us that Koenig Shield will be better if it procs more than 40% of the time, and many tests have shown that it procs more often than that.

Since we can get them both to cap pretty easily though, you've got 61% vs 38% at the exact same proc rate...and at that point it's really obvious which wins.

Wait, tested cap for buckler type shields? Or tested cap for Ancile? Where did we go from "hard to test block % based on blocks sometimes being as high damage as normal weak hits" to "this is the cap for ancile".? Data?

Tested cap for Bucklers and Kite shields was both the same. You can see that one on the 1st page of this thread, where I posted my original tests. My tests back in April were done differently than this, with me sitting there without my weapon drawn watching every hit.

Quoted again for reference:

Monastic Cavern, Coffer mobs (EP, L67 max), 302 Shield Skill, 75 PLD

Tatami Shield (size 4): blocked 129/332 attacks, 38.86%, estimated 80% damage reduction. Overall damage reduced = 30%
Koenig Shield (size 3): blocked 265/421 attacks, 62.95%, estimated 60% damage reduction. Overall damage reduced = 37.8%
Bulwark Shield (size 1): blocked 165/256 attacks, 64.45%, estimated 0% damage reduction. Overall damage reduced = 0%

Now, we've pretty much confirmed that my estimate reduction numbers are wrong, but the proc rates are definitely correct. 80% is definitely wrong...Who gave you those numbers?

[Neosutra]

Tested cap is around 65% regardless. A Koenig Shield with enough skill will proc at that rate or higher regardless. Simple math tells us that Koenig Shield will be better if it procs more than 40% of the time, and many tests have shown that it procs more often than that.

Since we can get them both to cap pretty easily though, you've got 61% vs 38% at the exact same proc rate...and at that point it's really obvious which wins.

Wait, tested cap for buckler type shields? Or tested cap for Ancile? Where did we go from "hard to test block % based on blocks sometimes being as high damage as normal weak hits" to "this is the cap for ancile".? Data?

Tested cap for Bucklers and Kite shields was both the same. You can see that one on the 1st page of this thread, where I posted my original tests. My tests back in April were done differently than this, with me sitting there without my weapon drawn watching every hit.

Quoted again for reference:

Monastic Cavern, Coffer mobs (EP, L67 max), 302 Shield Skill, 75 PLD

Tatami Shield (size 4): blocked 129/332 attacks, 38.86%, estimated 80% damage reduction. Overall damage reduced = 30%
Koenig Shield (size 3): blocked 265/421 attacks, 62.95%, estimated 60% damage reduction. Overall damage reduced = 37.8%
Bulwark Shield (size 1): blocked 165/256 attacks, 64.45%, estimated 0% damage reduction. Overall damage reduced = 0%

Now, we've pretty much confirmed that my estimate reduction numbers are wrong, but the proc rates are definitely correct. 80% is definitely wrong...Who gave you those numbers?

I think Bulwarks shield needs to be taken out of that test, as it just isnt a standard shield, has no way other then animation for testing if a block went off (and thats misleading).

Ill do some more tests, but size 1 shields do indeed have a higher block rate % then size 3. Ill try to find some data.

[Ferien]

*shrug* if you find other numbers then great, but from what I remember, the numbers from my test were consistent with others that popped up around the same time. The only way to test Bucklers is by watching the animation, and if you throw out Bulwark, you'd have to throw out Ancile as well, which brings us back to square one no matter what.

[Neosutra]

*shrug* if you find other numbers then great, but from what I remember, the numbers from my test were consistent with others that popped up around the same time. The only way to test Bucklers is by watching the animation, and if you throw out Bulwark, you'd have to throw out Ancile as well, which brings us back to square one no matter what.

No, you dont have to throw out Ancil with the bulwark, due to Ancile higher damage reduction (ancile has 32 def, bulwark has 1 def. Bullwark thus does -no- damage reduction by the shield block reduction formula).

Ill try and write a parser that determines the mode shield block and unblocked for ancile and see if I can just get hit by a mandra for about 6 hours. But it will be about a week or so before I can get around to this.

[Ferien]

*shrug* if you find other numbers then great, but from what I remember, the numbers from my test were consistent with others that popped up around the same time. The only way to test Bucklers is by watching the animation, and if you throw out Bulwark, you'd have to throw out Ancile as well, which brings us back to square one no matter what.

No, you dont have to throw out Ancil with the bulwark, due to Ancile higher damage reduction (ancile has 32 def, bulwark has 1 def. Bullwark thus does -no- damage reduction by the shield block reduction formula).

Ill try and write a parser that determines the mode shield block and unblocked for ancile and see if I can just get hit by a mandra for about 6 hours. But it will be about a week or so before I can get around to this.

>.> I already did that. I can show you the data I pulled from the tests Woozie gave me, and you'll see that there's no discernable "barrier" between blocks and non-blocks. You will not be able to find a proc rate unless you're sitting there watching it the whole time. i already posted the Modes for Ancile about 10 posts up, it was 41 for Blocks and 67 for non-blocks. However, it's impossible to determine the proc rate based on the info I have.

[Neosutra]

*shrug* if you find other numbers then great, but from what I remember, the numbers from my test were consistent with others that popped up around the same time. The only way to test Bucklers is by watching the animation, and if you throw out Bulwark, you'd have to throw out Ancile as well, which brings us back to square one no matter what.

No, you dont have to throw out Ancil with the bulwark, due to Ancile higher damage reduction (ancile has 32 def, bulwark has 1 def. Bullwark thus does -no- damage reduction by the shield block reduction formula).

Ill try and write a parser that determines the mode shield block and unblocked for ancile and see if I can just get hit by a mandra for about 6 hours. But it will be about a week or so before I can get around to this.

>.> I already did that. I can show you the data I pulled from the tests Woozie gave me, and you'll see that there's no discernable "barrier" between blocks and non-blocks. You will not be able to find a proc rate unless you're sitting there watching it the whole time. i already posted the Modes for Ancile about 10 posts up, it was 41 for Blocks and 67 for non-blocks. However, it's impossible to determine the proc rate based on the info I have.

Yep, its impossible to determine the proc rate based on the information -you- have based on someone else's shield.

Hence why im going to run -my- own tests, with my own script, my own shield, and see if I can get more accurate tests, using different mobs that will give me a greater range of mode values to better determine statistical differences.

Whats your problem again?

[Ferien]

Whatever man. You should already know yourself that 38% is well within the variance of a normal hit, and that trying to find blocks vs normal hits with just a list of numbers is going to be pretty much impossible. You'll have two overlapping curves with no clear end to either.

I have no problem, I'm just trying to warn you so that you don't waste your time. If you do run these tests, keep a manual count of how often your shield procs, or you will end up with a bunch of data that you simply can't use for that purpose.

[Neosutra]

You can push down the variance of hits via adjustment of def/stats and by fighting harder mobs. Yes there will be a percentage that will get counted as blocks that were not, but it should be about equal to the amount that are counted as non blocks but were infact blocked as in any normal distribution, if its done correctly. Its not rocket science >.>.