Random Facts Thread

[Byrthnoth]

I'm going to go ahead and draw the analogy between it and Divine Benison. If you think it's ~25% at 90, it's probably 5% per tier level with a tier every 10 levels. What method did you use to test this?

[Martel]

I've posted the results of the Shield Def bonus test data I had lying around over in the job traits thread.

http://www.bluegartr.com/threads/106...=1#post4789103

[Radec]

I'm going to go ahead and draw the analogy between it and Divine Benison. If you think it's ~25% at 90, it's probably 5% per tier level with a tier every 10 levels. What method did you use to test this?

It's rough, but I cast a long spell (AM of all elements) 3-4 times each and ran around to interrupt them, and went with the timestamp difference between casting and interrupted as the new casttime.

If you can cap out cast speed with /rdm, atma etc you could probably get a more exact test as you remove cast speed to uncap, but I don't think there is enough available until near lvl90

[Byrthnoth]

Mana Mist restored 300MP to my level 81 Soothing Healer NPC.

And he used Tachi: Ageha for typical crap NPC damage.

[Martel]

I was browsing the BG wiki the other day and I came across this tidbit in the Shield Skill page.

"In addition to skill, Agility is the determining statistic for shield blocking. Increasing a player's agility can help to increase the block rate. "

I know that rumor has been around forever, and its not an unreasonable assumption, given that all the other defensive skills do benefit from AGI. But it is just that, an assumption. Unless I missed some testing somewhere. The only test I have ever seen was tiny and inconclusive.

As I was scoffing at such an old rumor being on a wiki without even so much as a verification tag I realized I was doing exactly what others were. Assuming it didn't affect it while they assumed it did. Wups. ^^;

So I decided to go test it. And abyssea and atmas make it fairly simple to do so.

Data:

Spoiler: show

LVL 90 AGI test. Koenig Shield EM Monitors	
AGI 64 Skill 369

Blocked
DMG	#
17	2
18	43
19	31
20	48
21	22
22	42
23	28
24	56
25	33
26	30
27	161
28	118
29	90
30	15
31	39
32	23
33	44
34	17
35	1

Block#	843
AVG DMG	26.16

non-block	
DMG	#
61	8
62	3
63	18
64	16
65	23
66	20
67	16
68	18
69	23
70	17
71	18
72	15
73	21
74	17
75	19
76	22
77	30
78	16
79	14
80	15
81	14
82	13
83	17
84	13
85	14
86	92
87	87
88	76
89	87
90	83
91	16
92	15
93	7
94	12
95	15
96	11
97	25
98	12
99	11
100	18
101	11
102	14
103	8
104	7
105	1
106	2
107	1

Nonblock#	1031
AVG DMG	83.61
	
Total#	1874
Block Rate	44.9%

90 Koenig	
AGI 64+90	Skill 369
		
Blocked		
DMG	#	
17	2
18	34
19	31
20	34
21	30
22	41
23	26
24	58
25	41
26	33
27	142
28	158
29	96
30	21
31	32
32	21
33	22
34	13
35	3

Block#	838	
AVG DMG	26.16

non-block	
DMG	#
61	7
62	5
63	16
64	15
65	22
66	31
67	24
68	24
69	15
70	15
71	23
72	21
73	14
74	16
75	15
76	21
77	24
78	26
79	24
80	20
81	14
82	14
83	14
84	10
85	25
86	96
87	106
88	82
89	80
90	95
91	21
92	14
93	19
94	9
95	14
96	13
97	21
98	13
99	10
100	15
101	13
102	12
103	7
104	5
105	7
106	1
107	3

Nonblock#	1111	
AVG DMG	83.5

Total#	1949
Block%	42.9

So.

Koenig shield
AGI 64, Skill 369. Block Rate: 44.9%

AGI 64+90, Skill 369. Block Rate: 42.9%

I'd assume the 2% drop is just variance. But in any case there doesn't appear to be any noticeable gains from adding massive amounts of AGI.

I also did similar tests for VIT and DEX, with the much the same result. I'm suspecting that Shield skill doesn't have a governing attribute like other skill types do. That might be part of why it's so damn hard to cap on mobs of even a decent lvl.

I'll probly do a base stats test, then one with full cruor buffs later to confirm.

[Radec]

The question I have there with having pld16 only, can you push block rate above ~45% otherwise?

[Byrthnoth]

I've always strongly suspected that the AGI rumor (exists for both shield and parrying) was bullshit, and I remembered some testing to the contrary from ages ago.

Now that I think about it, I think I actually removed it from the Parrying Skill page a little while ago. It's gone from the Shield Skill page too now.

[Martel]

The question I have there with having pld16 only, can you push block rate above ~45% otherwise?

Having not done many solid block rate tests(Aside from Ochain), I had to ask the same question myself. Sure, I see the block rate cap listed as 65%, but how much testing is behind that? And was that testing any good?

So, I did a third test, same mobs. But stacking shield skill to confirm that my block rate wasn't capped.

90 Koenig block. Monitor	
AGI 64+90	Skill 412
	
Blocked	
DMG	#
16	17
17	48
18	38
19	54
20	41
21	51
22	40
23	60
24	46
25	37
26	200
27	164
28	123
29	35
30	34
31	27
32	23
33	4
Block#	1042
AVG DMG	24.75

non-block	
DMG	#
58	6
59	1
60	11
61	10
62	9
63	12
64	16
65	14
66	22
67	12
68	16
69	23
70	17
71	23
72	18
73	14
74	19
75	17
76	12
77	12
78	21
79	14
80	19
81	11
82	12
83	20
84	15
85	77
86	73
87	65
88	72
89	67
90	10
91	12
92	11
93	15
94	14
95	11
96	12
97	9
98	7
99	5
100	8
101	4
102	2

Nonblock#	870
AVG DMG	81.2

Total Hits	1912
Block Rate	54.4%

So, after adding +43 skill(total 412 skill) my block rate jumped nearly 10%, going from 44.9 to 54.4%. Certainly doesn't seem like I was at capped block rate for a size 3 shield.

So, I'm pretty sure AGI's not doing anything for block rate.

[Martel]

I've always strongly suspected that the AGI rumor (exists for both shield and parrying) was bullshit, and I remembered some testing to the contrary from ages ago.

Now that I think about it, I think I actually removed it from the Parrying Skill page a little while ago. It's gone from the Shield Skill page too now.

You know.. Parrying/AGI tests would be way easier to do than shield tests. For shield I have to take a fair amount of dmg so I can see the difference between blocked and non-blocked hits.

For Parrying I can just supertank a ton of mobs that are all hitting me 0. Then I'd have my whm mule keep one mob slept out of melee range and I'd engage that one. Then rake in the parse data.

The only downside, is that adding alot of AGI is gonna make the mobs start missing alot, and slow down data collection. I might do that sometime.

Thoughts on what lvl and type of mob might be good? Not sure if it'd be better to test on higher or lower lvl stuffs.

[Lyall]

I'm not so good at the data-gathering but if you're looking for raw number of hits you probably want Mandies or something. Since you're already going to be dealing with a smaller-than-normal size due to evasion anyway, doubling attacks in a round would help. As for which ones to test on, I have no idea. How survivable is the top of Boyahda recently?

[Martel]

I could probly supertank every non-NM mob in that zone at the same time. Assuming I could get them all to stay in front of me, and the sheer lag didn't knock me right off the servers.

But I'm probly going to need to use an abyssea zone, for atmas. As PLD doesn't get a whole lot of AGI gear. And the few pieces I do have also have other stats. I'd prefer to have as controlled test conditions as possible. Test with base AGI, test with Large AGI+ and no other changes.

I might do something like the mandies in abyssea altepa. If I can catch them at a static lvl(base lvl or EM) I think mob lvl affects parry rate... right?

[Byrthnoth]

Mob level does affect Parrying rate, but the problem is more "how many monsters can you keep in front of you at once?" than "how many monsters would be good?"

I'm like ~95% sure Parrying is unaffected by AGI. It looks like a straight dLVL/Skill relationship to me.

[Martel]

Well, since it was such a fast, easy test.. I went and pulled all the fear dearg. I wish I could do block rate tests that fast.. If only blocking gave a different chat message. Or I could dat mod or flag them somehow...

Anyway. Tests.

Test 1
AGI 64 Skill 310
Melee Damage Taken
Player             Melee Dmg   Melee %   Hit/Miss   M.Low/Hi    M.Avg  #Crit  C.Low/Hi   C.Avg     Crit%
Arduwyn                  124  100.00 %       30/2       0/36     4.43      2       0/0    0.00    6.67 %
Martel                     0    0.00 %  3228/1032        0/0     0.00    237       0/0    0.00    7.34 %

Passive Defenses
Player             Evasion  Evasion %   Parry  Parry %   Counter  Counter %   Intimidate  Intimidate %
Arduwyn                  2     6.25 %       0   0.00 %         0     0.00 %            0        0.00 %
Martel                1032    22.48 %     331   9.30 %         0     0.00 %            0        0.00 %

Test 2
AGI 64+90 Skill 310
Melee Damage Taken
Player             Melee Dmg   Melee %   Hit/Miss   M.Low/Hi    M.Avg  #Crit  C.Low/Hi   C.Avg     Crit%
Arduwyn                 1931  100.00 %      84/10       0/43    21.58      3     49/70   61.00    3.57 %
Martel                   346   27.33 %  3186/3594       0/55     0.07    160     0/104    0.87    5.02 %

Passive Defenses
Player             Evasion  Evasion %   Parry  Parry %   Counter  Counter %   Intimidate  Intimidate %
Arduwyn                 10    10.64 %       0   0.00 %         0     0.00 %            0        0.00 %
Martel                3272    45.73 %     374   9.63 %       323     9.20 %            0        0.00 %

So, like you thought, AGI doesn't affect Parry rate at all. Except by making the mob miss alot and making the test take twice as long.

[Foldypaws]

Was that on the test server? 10% parry rate is way higher than I remember ever seeing, I know they supposedly changed it on the test server though.

[Byrthnoth]

It's Kparser, so it isn't test server. Also, superficially nothing has really changed on the test server.

[Martel]

They were EP~DC mobs, so that's probly why.

[Byrthnoth]

If anyone is interested in Avatar pdif/damage calculations, I've been trying to compile the information on them (mostly Kegsay's) so I can write a coherent page for bgwiki. I necroed this thread for that purpose, and consolidated most of the testing from the last few years on my user page. Avatar melee damage seems pretty well figured out, but BP damage is freaking weird.

[Kirschy]

If anyone is interested in Avatar pdif/damage calculations, I've been trying to compile the information on them (mostly Kegsay's) so I can write a coherent page for bgwiki. I necroed this thread for that purpose, and consolidated most of the testing from the last few years on my user page. Avatar melee damage seems pretty well figured out, but BP damage is freaking weird.

If it helps, I believe I have an answer as to why Avatar Blood Pact damage works differently. Back in 2004 they modified BP damage:

The strength of physical "Blood Pact" attacks will not increase as much when fighting enemies considerably weaker than the summoner. The damage ratio when fighting stronger enemies has not been adjusted. http://www.playonline.com/updateus/041209we8yu0.html

The effect was fairly noticeable on easier mobs. It was suggested by myself and some others that the pdif for BPs was capped lower, which fits the description fairly well.

[Yugl]

http://www.bluegartr.com/threads/376...=1#post4795789

CDC Critical hit rate testing.

[Masamune]

http://www.bluegartr.com/threads/376...=1#post4795789

CDC Critical hit rate testing.

Looks still approximate testing method to me... tough kudo for this tedious process.
Protescar results just reduces the possible values for crit% at 100/200/300, like Byrthnoth very well concluded 2 posts below his results.

Where are Blade: Hi critrate test(s) to compare?

Also Blade: Hi and CdC don't have same #hits right ? sounds about right with all crit based WSs having all same initial crit% (at 100TP) but different at higher TP depending on #hits the WS has ?