Test: Double Attack during WS

[LyonheartLakshmi]

Re: Test: Double Attack during WS

I just read a theory posted by someone that DA during WSs act in the following manner. The game will check your first swing to see if you get a DA. If it succeeds, it will then also check your next swing to see if you get a DA. If at any time a DA check fails, the game will not make further checks for DA. This caps at 8 total hits for the WS including DAs.

So basically this theory says you can get multiple DAs as long as you keep getting successful checks, but will stop checking the moment you fail your first DA check. This would effectively result in fewer perceived DAs.

Armando, have you heard this theory before? If you rework the probability in this manner, is it consistent with the results of your tests?

[ItazuraNhomango]

Re: Test: Double Attack during WS

I have a macro that straps on my full enmity gear. I see the "equipment changed" lines spamming my chat log, but my character doesn't blink right away. How certain are we that the server receives the change upon seeing that message? Maybe the chat log line is generated client side, and I'm not actually getting the benefit until my character blinks (which is actually "redraw if needed."

Well, can you get the equipment change message while d/c'ing?

Log on to character, set up a gear change macro. Unplug the network cable, and hit the macro ASAP. That should give part of the answer; if you still get the equipment change message, you know it's client side generated. If not, you can reasonably assume it's waiting on the server to give confirmation message, if it's waiting for message at all. (I'm guessing it's not; it's likely "fire and forget", and equipment change message would happen as long as server sends such a command.)

I really should try that SC glitch, though. Can't believe I haven't noticed it all this time.

In any case, none of this is relevant to DA and WS.

[Lmnop]

Re: Test: Double Attack during WS

In any case, none of this is relevant to DA and WS.

Too true, let's try to keep this thread dedicated to that topic alone.

[Armando]

Re: Test: Double Attack during WS

Well, since I had some free time and the update is tomorrow, I decided to get some extra Vorpals for better accuracy. Here's 30 more.

042) 109, 9, 560 --- 98 --- 0
043) 115, 9, 568 --- 100 --- 0
044) 107, 9, 639 --- 100 --- 1
045) 100, 9, 560 --- 98 --- 0
046) 105, 8, 483 --- 96 --- 1
047) 106, 8, 414 --- 96 --- 0
048) 106, 10, 76 --- 898 --- 1
049) 108, 9, 630 --- 98 --- 1
050) 109, 10, 759 --- 96 --- 1
051) 101, 9, 700 --- 98 --- 2
052) 102, 9, 568 --- 100 --- 0
053) 109, 9, 560 --- 98 --- 0
054) 100, 9, 690 --- 96 --- 2
055) 105, 9, 560 --- 98 --- 0
056) 104, 9, 759 --- 96 --- 3
057) 104, 8, 426 --- 100 --- 0
058) 113, 9, 568 --- 100 --- 0
059) 103, 8, 490 --- 98 --- 1
060) 106, 9, 630 --- 98 --- 1
061) 102, 3, 418 --- 98 --- 0
062) 110, 9, 552 --- 96 --- 0
063) 106, 9, 556 --- 98 --- 0
064) 103, 9, 770 --- 98 --- 3
065) 126, 8, 490 --- 98 --- 1
066) 102, 9, 552 --- 96 --- 0
067) 108, 10, 770 --- 98--- 1
068) 116, 3, 560 --- 98 --- 2
069) 106, 9, 700 --- 98 --- 2
070) 107, 9, 630 --- 98 --- 1
071) 108, 8, 420 --- 98 --- 0

Parsed crit% during this second set was 5.44%
Vorpal's crit rate with all 100-115% Vorpals so far: 54/230 = 23.48%

Total 5-hit Vorpals: 7/71 = 9.9% (compared to the theoretical 13.9%)

[Lmnop]

Re: Test: Double Attack during WS

I grabbed some figures for Armando a few weeks ago. All of these numbers were gathered directly after the first 2 handed update, so all my normal swings were a constant 96 damage. I've been awful busy the last few weeks, so sorry to just now get around to posting this info.

code tags never work for me, so let's see how this turns out:


Around 60 Rampages dual wielded. Capped attack and capped STR (I forget attack, 107 str. I know I had enough attack that Armando gave it his blessing).

damage-starting tp-ending tp

888-139-16
962-118-16
786-101-15
824-118-16
1036-143-16
851-102-16
814-106-16
814-106-16
888-125-16
962-122-16
814-111-16
828-109-16
861-109-16
740-119-15
896-123-16
896-101-16
814-147-15
814-110-15
666-108-10 maybe uncapped attack
814-110-15
518-142-09
814-109-16
1036-115-16
962-100-16
592-128-14 overkill
962-107-17
740-121-10
962-117-16
666-111-15
799-101-16 uncapped attack
925-116-17
788-102-10
592-94-14
962-101-16 bio2
740-102-10
962-104-11
888-191-16
1036-113-17
666-112-15
814-100-16
814-109-16
962-104-16
814-111-16
1036-100-11
962-105-17
1147-99-17
888-107-16 overkill
999-101-16
703-131-15
962-125-17 overkill
1036-104-16
1036-107-17
814-98-16
851-107-16
814-102-10
666-109-15
592-119-14
962-116-16
851-106-16
888-107-16
740-102-15
792-124-16

Note that a few had under 100 tp when activating. This was just a laggy day. my client side query was a second behind. And these were ones that I executed directly after the melee swings that put me over 100, and trying to squeeze in before the mob swung at me (which typically meant I got lazy and my shadows dropped). At any rate, just consider these ones as having 105-111 tp.

[Ryoii/Nonomii]

Re: Test: Double Attack during WS

Edit: This was only one really nice 6-hit Penta Thrust. I've come to find that Store TP adds to the TP from the additional hits. Now I know why I never saw a 22TP return with Store TP+46. I'm sorry for my mistake.

I scored a lucky 8-hit Penta Thrust while both Berserk and Warcry were up this Saturday. I was duoing T-VT Darters in The Boyahda Tree with my SAM/WAR and RDM/WHM pair.



I'm quite sure that the 333 DMG hit in the screenshot is a critical since the DMG is displayed on its own line, and you can see a counter for only 133 DMG. 1558 DMG seems about right for an 8-hit Penta Thrust, and this is my personal best while XPing off of T-VT enemies (did not XP during the broken 2h level correction period). I only had Store TP+46 since I was using cheap Meat Mithkabobs, so that would give me 18.9 TP for my first hit from my Engetsuto (480 delay, 13TP base) plus 7 TP for the additional hits. For those of you unfamiliar with Windower, you can see in the lower right that the WS returned 25TP.

My only source of double attack was WAR SJ, so the chance of this happening on a 5-hit Penta Thrust was only 1 in 1000 based on the theory that double attack only checks multiple times if each check is successful. I wasn't getting too many 5-hit Penta Thrusts since I fired off right near 100TP, so I consider this extremely lucky.

This was during roughly a 4 hour XP session at a little over 5k XP an hour, and I finally hit Lv74. Just one more to go ^_^ /cheer. In case you are curious, I believe this Darter was Lv75, and I was still Lv73.

[Armando]

Re: Test: Double Attack during WS

Lmnop, many thanks. I'll analyze the number of crits later. For now, DA rate.

Total WS: 62
Overkills: 3
Observed single DAs: 9 (1 of which was an overkill)
Oberserved double DAs: 0

If we count overkills, the rate of single DAs was 9/62 = 14.52%
If we ignore overkills, the rate of single DAs was 8/59 = 13.56%

Once again it looks like only two DA checks are occurring (18% chance of a single DA * 73.51% chance to land all hits = 13.23% WS's with observable single DAs.) This couldn't come as a surprise, as a 6-hit Rampage can't have more than 8 hits.

That aside, wow, thanks a million Ryoii. That's pretty convincing; for now we can assume 5-hit WS get 3 DA checks. Can you post the level/gear/stats you had at the time?

[Lmnop]

Re: Test: Double Attack during WS

I forgot to mention: natural double attack is 11-12% (unmerited double attack + AF feet which are assumed 1-2%).

[Callisto]

Re: Test: Double Attack during WS

Just a quick couple of questions Armando, I noticed earlier in this thread you had said something about pDIF being capped at 2.0...did you mean cRatio? And if cRatio is still capped at 2.0, would it be safe to assume that to find pDIF you now multiply cRatio by 1.23, since that's the only thing that could get the new higher observed pDIF of 2.46?

[Armando]

Re: Test: Double Attack during WS

During the time between the first and second two-hander patches, S-E broke the PDIF formulas. They were the same as before, except instead of cRatio being capped at 2.0, cRatio was uncapped and PDIF was capped at 2. So when cRatio was 2.0, your PDIF would range from 1.6 to 2.4, but all values higher than 2.0 would get capped at 2. But since cRatio became uncapped, then you could keep increasing your cRatio until eventually the formula for min PDIF would reach 2.0 too, and every swing would have 2.0 PDIF.

Sorry, I don't know exactly what changed with PDIF yet. Haven't seen any threads in BG or anyone post information from Studio Gobli, and my play time has been extremely limited as of late so I haven't been able to do any sort of test.

(I believe you asked me this in another thread, but I was very busy at the time and forgot to reply later. My bad.)

[Callisto]

Re: Test: Double Attack during WS

Ahhh ok, thanks.
I have my theory, I just have no idea how to confirm/disprove it. It would appear that the new max/min pDIFs are 2.46/1.58 from the info I've seen around, as well as my own. Where you used to have cRatio capped at 2.0 then multiplied x 1.2 for max, and then minus .80 for min, I believe they now multiply it x 1.23, then subtract .88...but changing the cRatio cap to 2.05 would also get you 2.46 pDIF, so I have no idea how to figure out which one they changed, and I don't have a parser so collecting data involves me either punching #'s into an Excel sheet or pen & paper, so it's a bit of a PITA.
It would make sense to me that they would make the changes for max and min both on the cRatio > pDIF conversion end because that seems like less work than changing both cRatio and the conversion #'s, but the devs at SE are probably a little brighter than I am.

(I believe you asked me this in another thread, but I was very busy at the time and forgot to reply later. My bad.)

Hehe, no worries, I've actually been making a push to figure this stuff out on my own instead of bothering guys like you for it. I'm trying to keep my brain in shape until I can get back into school, but there's a lot of rust and booze to shake off.

[Ryoii/Nonomii]

Re: Test: Double Attack during WS

Armando, I'll collect my character statistics and gear information late this evening. I might also have information on the stats of the Darter since I did some /check tests a month or two ago when I first started hunting here.

[LyonheartLakshmi]

Re: Test: Double Attack during WS

Once again it looks like only two DA checks are occurring (18% chance of a single DA * 73.51% chance to land all hits = 13.23% WS's with observable single DAs.) This couldn't come as a surprise, as a 6-hit Rampage can't have more than 8 hits.

That aside, wow, thanks a million Ryoii. That's pretty convincing; for now we can assume 5-hit WS get 3 DA checks. Can you post the level/gear/stats you had at the time?

Could it be that the game counts total WS hits (including the offhand), and subtracts that from 8 to give the number of DA checks it does?

[Ryoii/Nonomii]

Re: Test: Double Attack during WS

That would mean that single wielded Vorpal Blade would always get 4 DA checks. I wish this was the case, but we would have seen many more 5-hit Vorpal Blades in Armando's testing:

Chance for 5-hits landing at 95% accuracy: 77.38%
Chance of exactly 1DA with 4DA checks: 34.39%
Chance of a 4-hit Vorpal Blade + 1 DA occuring: 26.61%

Sadly I think you only get DA checks until you miss a DA check. This makes a DA on a 1-Hit WS as likely as a DA on a 7-Hit WS.

I also think that when DWing you may get an additional DA check. This could be tested by using DW with Fast Blade and comparing to Armado's results. I guess you could also DW with a single-hit WS.

I wish I had the time to help test this. I'm rather bummed that Penta Thrust might only get 1 DA check. /sigh It really seems to fit with my experiences.

[Armando]

Re: Test: Double Attack during WS

Sadly I think you only get DA checks until you miss a DA check. This makes a DA on a 1-Hit WS as likely as a DA on a 7-Hit WS.

This is clearly not the case, from my test data (especially the Fast Blades, since those are the most thorough.)

In my Fast Blade tests, if both hits can proc DA independently, there'd be an 18% chance overall of getting a single DA, which when multiplied by the odds of landing all three hits (85.74%) gives us a 15.3% odd of getting a single DA and landing all of them. Experimentally, I had a 15.5% rate of single DAs with all 3 hits landing.

Now, this 18% chance to DA once comes from the fact that there's two possible ways to get a single DA: 1st hit DA while the 2nd doesn't, or 2nd hit DA while the first doesn't. If DAs were checked sequentially until a check is failed (i.e. if you fail to DA the first hit, you can't DA at all; if you DA the first, you have a chance at the second) then we'd basically be back at a 10% single DA rate. In that case I'd have observed less than 10% of my WS get a single DA (due to misses.)

So far, it seems that it works like this:
2-4 hits: 2 DA checks
5 hits: 3 DA checks
6 hits: 2 DA checks
7 hits: 1 DA check