Random Question Thread VI: I'm No Mathologist, But...

[Wolfknight]

Assuming /WAR is your other alternative (which it should be? idk), I don't know how 5% haste would fair against Double Attack and Berserk post-50. You'll be sitting at 12% haste (also assuming you're TPing in Swift Belt/Fumas; positive SAM doesn't have access to other haste gear at that level) @ that level with a Wyvern Earring. Personally I'd stick with SAM/WAR, but I'd look for a second opinion.

I don't think would worth using unless at 50+ you're using soboro since jump, and extra haste work well with it, where as DA doesn't do much. Or maybe if you're getting haste, and double march using GKT at 60~70 level range.
If you're using polearm though /war should win except at maybe 60 ish for acc bonus, but I don't think it would be out zerk still.

[Callisto]

So this has bugged me for awhile...there's this one spot in Sky, on the NW island, at the north end, and there always, always seems to be a random afk JP there. It's sort of by the coffer spawn, but not close enough to be camping the coffer, they're upstairs from the Zip entrance but south of the coffer spawn.

It's not the same person, it's always someone different, usually totally different ranges of jobs, but every single time I go to pop my coffer keys from doing Despot/FoV it never fails that one of them is afk there.

Is there something special there that I don't know about?

[TybudX]

That's where they hide the JP button.

[Argettio]

Sam/Drg with Wyvren Earring pre 75: Worth it or not?

I'm having trouble either convincing parties to let me bring it out or they're complaining I'm taking too many hits (Any at all).

I would say any time you are using a GKT then /DRG would be worth it. Berserk is almost a requirement for polearm use, so /WAR all the way then.

So it all comes down to how you level and where.

If you are doing it 'perfectly' and able to sync to a bird camp (East ron [s], Woodslands, Bhaf etc) then you won't get any use out of /DRG.

[Cairthenn]

Does Sarameya have draw in? Also, how much HP does it have?

[Bloomonkey]

There a tool that let's you repeat macros like /ra <t> so I can skill up archery afk

[Yarko]

Does Sarameya have draw in? Also, how much HP does it have?

about 60K iirc, and no don't remember it drawing people in, but I could be wrong.

[requim]

It's not just that (percent) changes in WS frequency in the dual wield/H2H case depend on hit rate

Hit rate should be above 85% and most one handed jobs use sushi or pizza so it's more likely closer to 95%. While it's not 100%, it's close enough to get a good idea. Yes, it is much more likely to see multiple misses with a lower delay weapon than a higher delay weapon, but if your accuracy is respectable it shouldn't on average by more than a miss per WS round.

Actually, for something even more "obvious," where % changes in WS frequency are concerned it's irrelevant to consider realized double attacks and kick attacks in excess of 100 TP because they don't affect % changes in WS frequency by definition.

I tried to discount the offhand attack, double attack, and kick attack to represent the times when they exceeded 100 TP with Rajas because they nullified the benefit from Rajas dropping a hit. If you need X hits with no STP and X-1 hits with STP there is 0 change in frequency if you end up with X hits before being able to WS. That was my original reasoning for discounting the times where you don't benefit from increased WS frequency, though as I admitted, I think I halved an unneeded time.

The entire point of it was to isolate the benefit Rajas gives to a one handed job for WS frequency, which unlike two handers you're constantly getting at least two attacks per round so it can make it difficult to figure out what the actual average increase is.

cutting it in half is stupid and unnecessary.

As Pchan pointed out you gain twice as much benefit from dropping an entire attack round than just a single wield. I forgot to account for the entire attack round.

[Yarko]

There a tool that let's you repeat macros like /ra <t> so I can skill up archery afk

windower macros...?

[Argettio]

There a tool that let's you repeat macros like /ra <t> so I can skill up archery afk

http://www.autohotkey.com/

Allows you to execute key stroke combinations repeatedly in any program.

[pchan]

Hit rate should be above 85% and most one handed jobs use sushi or pizza so it's more likely closer to 95%. While it's not 100%, it's close enough to get a good idea. Yes, it is much more likely to see multiple misses with a lower delay weapon than a higher delay weapon, but if your accuracy is respectable it shouldn't on average by more than a miss per WS round.

The point of the accuracy in the discussion is not that is 85% instead of 95%, it is that it is 95% and not 100%. The main reason for which about half of the attack rounds apppear on the first swing is accuracy. Obviously it would no be the case with 100% accuracy, where it would always be either the first or second swing (for a given tp return on WS), and rajas would be either always useful or never useful like for 2 handers.


The 50%/50% distribution on the last round would still be the case for a pup with no DA trait for instance or a DNC/NIN with no brutal.

[Mikus]

Does Lindwurm have a Placeholder? Been camping for quite awhile and it just wont' pop. Hasn't been anyone in zone, just me, etc. Same exact thing happened with Friar Rush last week. Killed everything for hours on end and NMs just won't spawn.

[cyc]

RNG in meripos, what subjob should I use? I don't have /SAM yet, so I'd like to know which one is the best in general and which one is the best between /WAR and /NIN (or anyways, some advice about how to choose which one to use)

[Callisto]

Kinda depends what the rest of you party is, but for a normal shooting RNG I'd /WAR, just don't be retarded and WS at 95% when the other DDs are sitting at 15 TP.

[requim]

So the main ambiguity for STP for one handers is finding the probability it will drop a full attack round. I made a hard break down of probabilities for a monk, calculating STP tiers. There definitely may be some mistakes, which I will correct if they're pointed out. Hopefully this will clear up any confusion or arguments based on hit rate, added attacks etc.

Unfortunately my indents got mucked up and need to bounce before making this slightly more legible.

I'll use a monk with destroyers as my main example.

Assuming a static 15% KA and 5% DA (neither applicable for WS)

The TP returns given Rajas (0.2), Usu Feet (0.3), and Usu Body (0.2). Brutal only adds 0.1 if going from 6=>7, 12=>13, or 18=>19, otherwise 0.

So we can basically plug that information into an equation of

MIN of WSACC*[(WS hits*2)+TPACC*[(hits needed)*(TP hits)*(1+KA+DA)] ≥ 94 (max WS return)

Offhand I can't think of how to express hits needed after accounting for KA/DA rates, so I just guessed/checked the min hits and plugged those in (example with KA/DA Rajas needs 15 hits, not 17).

STP = 0 capped acc: 0.95*(4.9*2)+0.95*[(4.9)*2+(4.9*16*1.2)]=98.87, expected min 8 rounds
STP 5-8 capped acc: 0.95*(5.1*2)+0.95*[(5.1*2)+(5.1*15*1.2)]=96.9, expected min 7.5 rounds
STP 12-13 capped acc: 0.95*(5.4*2)+0.95*[(5.4*2)+(5.4*14*1.2)]=96.44, expected min 7 rounds
STP 18-19 capped acc: 0.95*(5.7*2)+0.95*[(5.7*2)+(5.7*13*1.2)]=95.30, expected min 6.5 rounds

That's purely looking at STP assuming capped accuracy for all cases as well as Brutal and 5/5 Kick. You can make any changes to the equation to compare varied STP vs acc, like Rajas/tors vs L/K for TP with identical WS gear.
0.95*(5.1*2)+0.95*(4.9*16*1.2) =99.07
0.95*(5.1*2)+0.9325*(4.9*15*1.2) =95.29

So in that instance you would expect to see the same number of attack rounds for both cases, though variation could let you WS in 7 rounds with Rajas with a 7/8 AF and an additional DA/KA. You could expect to regularly drop a full attack round with usu feet + rajas or usu x 2. The lowest average would be usux2+rajas that would.

Basically Rajas gives a chance to drop a full attack round given an extra miss (~5%)or an extra DA/KA (~20%). So a monk with 6-8 STP would expect to drop a round maybe a quarter of the time. Dodgy final analysis, but I need to head out. If someone wants to calculate the expected chance for Rajas and R+Usux2 to drop a hit would make it clearer since I totally eyeballed it at ~25% chance.

TL: DR
increasing STP in 5-7 increments drops expected round by a half.

Rajas ~1/4 chance to increase WS by an attack round
R+Usu or Usu x2 should regularly drop an attack round
R+Usux2 ~1/4 chance to increase WS by another attack round

[pchan]

S
STP = 0 capped acc: 0.95*(4.9*2)+0.95*[(4.9)*2+(4.9*16*1.2)]=98.87, expected min 8 rounds
STP 5-8 capped acc: 0.95*(5.1*2)+0.95*[(5.1*2)+(5.1*15*1.2)]=96.9, expected min 7.5 rounds
STP 12-13 capped acc: 0.95*(5.4*2)+0.95*[(5.4*2)+(5.4*14*1.2)]=96.44, expected min 7 rounds
STP 18-19 capped acc: 0.95*(5.7*2)+0.95*[(5.7*2)+(5.7*13*1.2)]=95.30, expected min 6.5 rounds

Before even considering the rest of your post, you make 2 mistakes. First asuran fists has 8 hits, the last 6 hits also give tp at the rate of 1.0 if under 10 stp and 1.1 if above 10 stp. The second mistake if that you cannot average the TP return from asuran fists and then deduce the average number of hits required. You have to list all of the a priori 21 different Tp return values, then for each calculate the minimum number of hits required to get 100+ TP and add ~0.5 hit since the last round is equally achieved on first or second hit. Then you divide by the average number of swings per round taking accuracy into account. This gives you an average number of rounds needed given an initial TP return from asuran fist. After that weight all possible value by the probability of that initial tp return to happen and you find a global average number of round.

I will edit this post with an example.

Simple example is with store tp=0. The possible TP returns are

x=4.9*i+1.0*j for i<=2 and j<=6. Put N=ceil( (100-x)/4.9), this is the mimum hits nedded to get to 100 tps. As I said you gotta add 0.5 hits and then divide by the average number of swings per round : R=(N+1/2)/(2.25*0.95). Finally this value is obtained only with probability of hitting i times the first 2 hits and j times the last 3 which is p=binomial(2,i)*binomial(6,j)*0.95^(i+j)*0.05^(8-i-j). Some all of the terms R*p for i,j and you fond your average number of rounds.

In the case of store tp=0, you find 8.71 rounds on average for instance. You can repeat it for every value of store tp. Now for my calculation I will always use the same store TP during WS (12 : rajas + usu feet : 5.4 and 1.0 return on asuran fists) since usu body and brutal are not good.

Spoiler: show

Format : i store tp , average #rounds with i store tp , average #rounds with i+5 store tp , difference
0, 8.358711593, 8.233958553, 0.124753040
1, 8.358711593, 8.233958553, 0.124753040
2, 8.358711593, 7.890015499, 0.468696094
3, 8.236354565, 7.890015499, 0.346339066
4, 8.236354565, 7.768518308, 0.467836257
5, 8.233958553, 7.768518308, 0.465440245
6, 8.233958553, 7.766122296, 0.467836257
7, 7.890015499, 7.766122296, 0.123893203
8, 7.890015499, 7.732552124, 0.157463375
9, 7.768518308, 7.732552124, 0.035966184
10, 7.768518308, 7.313578400, 0.454939908
11, 7.766122296, 7.313578400, 0.452543896
12, 7.766122296, 7.299286310, 0.466835986
13, 7.732552124, 7.299286310, 0.433265814
14, 7.732552124, 7.298246456, 0.434305668
15, 7.313578400, 7.298246456, 0.015331944
16, 7.313578400, 6.954342983, 0.359235417
17, 7.299286310, 6.954342983, 0.344943327
18, 7.299286310, 6.845706419, 0.453579891
19, 7.298246456, 6.845706419, 0.452540037
20, 7.298246456, 6.831354789, 0.466891667

As you see, it's mosty ~0.45-0.46 rounds saved except for stp=0,1,3,7,8,9,15. Then again it could change if you assume uncapped accuracy during melee/ uncapped accuracy during ws etc.

[Tricen]

SO, I was told FFXI Calculator was good for making your game look spiffy. I D/Led it and now it wont go passed the Play button (where it reminds you to not forget about some ppl or some shit, idr) and whenI click play it goes to load and just goes right back to POL. I opened the calc thing and went to settings and hit highest quality and then went to save and apply, then I restarted ffxi. Now it is doin this. Hlap

[Cairthenn]

SO, I was told FFXI Calculator was good for making your game look spiffy. I D/Led it and now it wont go passed the Play button (where it reminds you to not forget about some ppl or some shit, idr) and whenI click play it goes to load and just goes right back to POL. I opened the calc thing and went to settings and hit highest quality and then went to save and apply, then I restarted ffxi. Now it is doin this. Hlap

Well I just did the same thing and it's workin' fine. Your problem! You probably fucked up the registry somehow.

[Tricen]

Well, am I not doin anything right? Idk anything about this shit .

[Cairthenn]

What size is your display/background resolution set to be?