Test: Double Attack during WS

[Lmnop]

Re: Test: Double Attack during WS

I don't see how that's exactly forgotten.

See, Armando's taking into account that double attacks will often go unnoticed because they'll instead fill in for attacks that missed. And then double attacks themselves will often miss, so we'll have a lot of imperfect data. Solution: you get a bigger pool of data.

Really, all we need is a screenshot of a Fast Blade that returns 9 tp, but we may as well use the time to collect more double attack data.

[Armando]

Re: Test: Double Attack during WS

Whew. Glad to see so many people discussing this ^^

Celeal, thanks for going through the trouble of showing all the possible outcomes, but I can't find a way to apply that to the data, since the only way to differentiate between the outcomes is when the first hit misses

Regarding jacking up your DA rate through other means for the test: Like Saren said, I do want to see a double DA happen with just the trait alone, because if I have other factors, then there's always that one person that'll say "I don't buy it. It could've been X, Y, or Z affecting you." However, that is a very good suggestion, and honestly I had forgotten about Warrior's Roll (haven't been playing enough this semester, I guess it shows XD.) As things stand now, I wouldn't be able to drag a COR and a WAR out to do my tests since I can only play on the weekends I go back home, and even then I'm often busy on weekends. However, if anyone wants to contribute a few tests with DA + Brutal + Warrior's Roll + Rampage, I certainly wouldn't turn it down. Like Lmnop said, all data's good data if you record the right things. Shouldn't take too long to see a double DA under those circumstances against TW mobs, but if anyone's going to try this, remember: no dual wield! That's part of the reason I'm doing this, after all.

Macht, I've already taken into the account the fact that some DA's will miss. That's why I was asking about the probabilities, to be able to match up the experimental number of no-miss DA'd WS to what the probabilities say I should get. You do bring up a good point, though, I didn't get any 2 TP Fast Blades, but I think the odds of that are low. A double DA should happen 1% of the time, 0.10 * 0.10. On the other hand, there's only a 5% chance of missing any given hit in the WS, and out of those 5% there's the other 10% chance of getting a DA to deal with. Not to mention that it's a 5% chance to miss any one hit, but we'd need to see the first hit specifically miss, so only half of those Miss-Hit-Hit scenarios would show up as 2 TP WS. I'm not even going to bother doing the numbers for the sake of not screwing up, but the probabilities are less than 0.5%.

[Celeal]

Re: Test: Double Attack during WS

References:
TP per hit = 6.2
1 TP = First hit missed
6 TP = Second hit missed
7 TP = 2 hits
8 TP = 3 hits
9 TP = 4 hits

Data:
Exactly 6/100 Fast Blades missed the first hit (~95% accuracy cap)
At least 2/100 Fast Blades missed the second hit.
At least 15/100 Fast Blades had one Double Attack.

Unless something is wrong with the possible out of Fast Blade I listed in previous post, I think something is missing in the reference of TP return (Macht in above post already covers it):

Other TP return of Fast Blade should be:
2 TP = Double Attack must be involved for Fast Blade to return 2 TP.
3 TP = Only one possible way for this to happen.
(See below for detail)

1 = One hit landed on the mob during multi-hit WS.
0 = One miss on the mob during multi-hit WS.
X = Double Attack does NOT trigger during that hit.

Note: there are total of 30 possible outcome of Fast Blade, assuming Double Attack is available (without Triple Attack, DW, or using weapon like Joy-toy, etc).

2 TP = 2nd hit landed and one double attack landed the mob:
1st D.A 2nd D.A
--- --- --- ---
0.. X.. 1.. 1
0.. 0.. 1.. 1
0.. 1.. 1.. X
0.. 1.. 1.. 0
Double Attack must be involved for Fast Blade to return 2 TP.

3 TP = 2nd hit land and two double attack landed
1st D.A 2nd D.A
--- --- --- ---
0.. 1.. 1.. 1..
Only one possible way for this to happen.

[Macht]

Re: Test: Double Attack during WS

Unless something is wrong with the possible out of Fast Blade I listed in previous post, I think something is missing in the reference of TP return (Macht in above post already covers it):

Other TP return of Fast Blade should be:
2 TP = Double Attack must be involved for Fast Blade to return 2 TP.
3 TP = Only one possible way for this to happen.
(See below for detail)
1 = One hit landed on the mob during multi-hit WS.
0 = One miss on the mob during multi-hit WS.
X = Double Attack does NOT trigger during that hit.
Note: there are total of 30 possible outcome of Fast Blade, assuming Double Attack is available (without Triple Attack, DW, or using weapon like Joy-toy, etc).
2 TP = 2nd hit landed and one double attack landed the mob:
1st D.A 2nd D.A
--- --- --- ---
0.. X.. 1.. 1
0.. 0.. 1.. 1
0.. 1.. 1.. X
0.. 1.. 1.. 0
Double Attack must be involved for Fast Blade to return 2 TP.

3 TP = 2nd hit land and two double attack landed
1st D.A 2nd D.A
--- --- --- ---
0.. 1.. 1.. 1..
Only one possible way for this to happen.

Yes, this was exactly what I was pointing out. DA will not fill in for the first hit so if first hit misses it's impossible to still get 6+TP when a DA kicks in and lands. Which means you should of seen at least a 2 at some point with those 6 were the first attack missed.

Though I was also pointing out that getting 9 TP was not the only way to prove that DA triggers for each hit in the WS, getting 3 TP also proves that. Hell from how it's looking it seems you might want to prove that getting 2 TP is possible, if that is possible then it posses more potential for your initial theory.

That's were I would start anyway, because out of all the times I've used Fast Blade with DA I can't recall a single instance I've ever seen 2 TP happen.

[Armando]

Re: Test: Double Attack during WS

Oh, ok. I get what you're saying now. Yeah, you're right about that - a 3 TP return Fast Blade really would prove the theory. However, the odds of a 3 TP Fast Blade are downright abysmal. The odds of 2 TP are also lower than the odds of a 4-hit Fast Blade, so I'm bound to see a 9 TP one way before I see a 2 TP or a 3 TP one. Either way, thanks for pointing that out, I'll definetely keep an eye out for those if they do happen.

[Celeal]

Re: Test: Double Attack during WS

If you acc+ against the mobs in the test is 95%, it would be hard to miss a hit during Fast Blade

[Macht]

Re: Test: Double Attack during WS

Well also out of a potential of 4 attacks, if the theory is right, it kind of also means that you don't completly need 95% accuracy to prove it. All you need is TP values.

Then also I would think to guarentee that first hit never misses to improve odds would be do like WAR/THF and use backstab with Fastblade. That will guarantee that first hit will never miss which means you only have to do odds of 3 other potential attacks landing. Which if I read right has higher and better odds of getting the result you are trying for.

[Armando]

Re: Test: Double Attack during WS

Regarding SA: That's a good suggestion. Actually, I'm not sure how much of a difference it'll make, since SA will only guarantee the first hit, and the first hit is the one hit in the entire WS that we can afford to miss since we can tell without doubt when it does hit by the TP returns. Problem is, I'd like to finish the Fast Blade tests without SA so that we can add all the data together, and I'll need to be PLD main for Vorpal Blade. Lmnop should be able to do it, though.

[Lmnop]

Re: Test: Double Attack during WS

if the lazy bastard would log on...

[Armando]

Re: Test: Double Attack during WS

Y H'LLO THAR DOUBLE ATTACK.

Attempt #120 (120 including the previous 100, that is.) Oh, and Lmnop, it's cool.

[Saren]

Re: Test: Double Attack during WS

Very nice! Congratulations on prooving your theory ^^

[Celeal]

Re: Test: Double Attack during WS

very nice, cheer ^o^!

[Armando]

Re: Test: Double Attack during WS

Thanks ^^ I'll finish the remaining 70 Fast Blades and combine the numbers later. Have stuff to do soon. Either way I'm glad to have found that. Oh, yeah, fun fact - just before I headed out for my testing I was in Giddeus, and I got DA on Circle Blade, but the DA only applied to the target, and not the Yag that took collateral damage.

I'd also like to point out that there's proof in the Allakhazam Thief forums that DA and TA can both proc within the same WS even when single wielding, so in theory a THF/NIN with Brutal and TA merits has a 10% chance of TA and a 5% chance of DA on each of Dancing Edge's 6 hits. You can find the thread [url=http://ffxi.allakhazam.com/db/jobs.html?fjob=10;mid=1177270401175316328;num=4;page=1]here[/url.]

EDIT: What the...a space keeps showing up on the link and it's breaking it. Oh well.

[LyonheartLakshmi]

Re: Test: Double Attack during WS

Nice. Congrats. I guess the question now is, does this hold true only for certain WSs (one of which is Fast Blade), or does it apply to others as well?

[Armando]

Re: Test: Double Attack during WS

I would expect this to apply to every WS. I don't see why S-E would make DA stack with every (non-ranged) physical WS, and then make it so only some of them can DA more than once. They already have an 8 hit cap on WS, which would prevent the more extreme multi-hit WS like Dancing Edge, Rampage, and Asuran Fists from getting too many DA's or something.

Anyways, the second round of 100 Fast Blades is here!
101) 103, 7, 162
102) 143, 8, 206
103) 107, 7, 159
104) 104, 7, 148
105) 102, 7, 152
106) 114, 8, 236
107) 105, 7, 137
108) 102, 8, 235
109) 130, 7, 156
110) 100, 7, 125
111) 106, 7, 137
112) 120, 7, 157
113) 108, 8, 235
114) 107, 8, 251
115) 105, 7, 161
116) 106, 8, 211
117) 113, 8, 247
118) 134, 7, 155
119) 120, 8, 242
120) 107, 9, 322
121) 110, 7, 142
122) 108, 7, 148
123) 100, 7, 150
124) 102, 7, 167
125) 104, 7, 149
126) 104, 7, 123
127) 100, 7, 147
128) 104, 7, 157
129) 100, 6, 78
130) 101, 7, 119
131) 101, 7, 162
132) 102, 7, 165
133) 104, 7, 146
134) 100, 8, 239
135) 102, 7, 162
136) 100, 7, 150
137) 100, 7, 177
138) 105, 7, 147
139) 102, 7, 173
140) 104, 7, 154
141) 110, 7, 159
142) 101, 7, 162
143) 100, 7, 132
144) 127, 7, 146
145) 100, 7, 137
146) 100, 7, 160
147) 101, 7, 153
148) 106, 7, 141
149) 103, 7, 133
150) 102, 8, 218
151) 106, 7, 169
152) 104, 7, 136
153) 106, 8, 234
154) 103, 7, 147
155) 100, 7, 163
156) 102, 7, 126
157) 103, 8, 230
158) 105, 7, 138
159) 107, 8, 215
160) 105, 7, 161
161) 106, 7, 177
162) 114, 7, 149
163) 111, 7, 166
164) 100, 7, 154
165) 102, 7, 152
166) 103, 7, 145
167) 110, 7, 143
168) 104, 7, 165
169) 104, 7, 164
170) 120, 7, 184
171) 110, 6, 81
172) 101, 7, 177
173) 105, 7, 157
174) 111, 7, 166
175) 109, 7, 149
176) 102, 8, 254
177) 108, 7, 143
178) 104, 7, 151
179) 101, 7, 138
180) 104, 7, 152
181) 108, 7, 152
182) 118, 7, 177
183) 100, 7, 128
184) 100, 7, 132
185) 104, 8, 222
186) 103, 7, 150
187) 104, 7, 172
188) 108, 7, 135
189) 102, 7, 148
190) 100, 8, 223
191) 105, 1, 89
192) 108, 7, 160
193) 165, 7, 169
194) 100, 7, 163
195) 105, 7, 157
196) 104, 7, 148
197) 101, 1, 89
198) 102, 6, 72
199) 103, 7, 169
200) 106, 7, 159

Now to add up the data...
Single DAs: 31/200 (at least)
Double DAs: 1/200 (at least)

Like I stated earlier, Fast Blade has an 18% chance of a single DA. Theoretically that would mean 36 of the 200 Fast Blades should've DA'd. Only 85.74% of those will land all their hits, so we should've seen ~30.87 of them land. That's pretty close to the 31 I got. That proves that each hit gets its own individual chance at DA, and the 9 TP, 322 damage Fast Blade is proof that the game allows multiple DAs within the same WS. I only got a single 4 hit Fast Blade, but that's to be expected, since the odds it occurring are only .8145% with accuracy factored in.

On an unrelated note, only two of my two hit Fast Blades exceeded the maximum expected damage of a two hit 100 TP Fast Blade (182). However, the one that exceeded that value the most still only exceeded it by 3 damage, and that can be attributed to having 135 TP instead of 100. All other Fast Blades fell within their expected ranges. Since those ranges were calculated using 1.6 and 2.4 PDIF as the minimum and maximum, this further supports the oft-disputed fact that only WS with crit modifiers can crit naturally in the first place.

Anyways, I'll probably start testing on Vorpal Blade next weekend.