Anyone try Apoc/RR/SS?

[Motenten]

@Theytak: Unfortunately you messed up right at the start of the math you posted (the first long post), in your calculations of TP per hit (you calculated TP from round delay and then divided the TP in half; you need to divide round delay in half and then calculate TP for a single hit). It's a common mistake, but it still means the values you use after that are invalid. Also, the manner in which you calculated the difference (whole number of weaponskills within a non-infinite timespan) is subject to fairly large amounts of error due to the short timespan you used.

2) SS increases crit dmg by a %, correct, however no buffs affect cirt dmg and it is a linear stat. A super simple example is this: Hit for 150, crit for 300 (+0% crit dmg, 150 dmg added from crit), crit for 450 (+100% crit dmg, 300 dmg added for cirt). Im sure someone will correct me on the math there but it really is linear.

Err.. Not quite. If a normal hit is 150, and a crit is 300, then when you add +100% crit damage the crit will be for 600 instead.

Anyway, my experience is that judging from the spreadsheet Stringing Pummel should always pull out numbers higher than Victory Smite, yet in-game (at least by eyeballing, never thoroughfully parsed) VS damage seems always more consistant than SP's, altough occasionally SP does more.

In the end I tend to use up VS for the Aftermath, but still this thing always puzzled me. More feedback from other users?

My initial guess is that you're setting up gear sets in the spreadsheet and then swapping between weaponskills to see which does better, rather than optimizing for each weaponskill.

Regarding your other concerns via fTP or accuracy and whatnot, such things are already accounted for in the spreadsheet. DA/TA with limits on max number of hits per weaponskill, proper probability distributions of the number of hits (with the above posts with 70%/77%/etc are not correct about), etc.

I don't play on pup very often, and my anecdotal evidence has lots of issues, but the last time I did a lot of melee/weaponskilling with pup and was messing with both SP and VS, I was doing substantially better with SP. I remember it as being vaguely annoying since SP wasn't doing just 'slightly' better, but a fair bit better. Only caveat is that at that time I believe I was using Rev Fists +2 instead of the Verethragna I later finished, and the extra str on Vere is more valuable to VS than SP.



I also have a long explanation of the effects of slow from Sea Daughter and TA from Apoc, but I'll refrain from the wall of text since the main argument is now a couple weeks old. Short summary: In order for switching from Apoc atma to Sea Daughter to be on par with using Apoc atma, one would have to prove that the 5/tick regain *by itself* can increase your total (melee+ws combined) damage by at least ~50%.

Given a 50/50 melee/weaponskill split (I'm actually hitting this range now with VS), a 50% increase in damage essentially means you need one extra weaponskill every weaponskill cycle. 5/tick regain means +100 TP every minute, therefor you would have to be weaponskilling once per minute for this to be comparable if considered strictly from the master's perspective.

In my experience, extremely high haste situations see weaponskills every 10 seconds. Completely unbuffed situations (but still capped gear haste) see weaponskills about every 30 seconds, including the occasional few seconds of switching between mobs. It would only approach one additional weaponskill per cycle with substantial amounts of time between mobs, though that in itself it not entirely unlikely (eg: have to wait 90 seconds between pops for Abyssea NMs).

If you add in skillchain damage, and assume mirrored damage amounts, that damage increase still has to be proportionately split between master and puppet. Unfortunately, that split varies wildly depending on many factors. Best case for SD's purposes is a split that's heavily in favor of the master, and that the master closes the skillchain.


Overall, modelling the effect is very complex. I'd say it's impossible for SD to beat Apoc for any standard DD, but could actually be feasible for a pup doing frequent skillchains (though difficult to manage given the relative TP gains between puppet and master). The big caveat is, of course, all the issues that Dfoley brings up regarding how much of that damage can actually be used.

[Jar]

Regarding your other concerns via fTP or accuracy and whatnot, such things are already accounted for in the spreadsheet. DA/TA with limits on max number of hits per weaponskill, proper probability distributions of the number of hits (with the above posts with 70%/77%/etc are not correct about)

Explain that part.

[Zhronne]

My initial guess is that you're setting up gear sets in the spreadsheet and then swapping between weaponskills to see which does better, rather than optimizing for each weaponskill.

Nope, altough I have to admit my SP and VS sets are very similar.
Don't have my macrosets in front of me atm but I think they differ only for a few slots.
Cirque+2 body for SP and Usukane for VP, Shadow Belt for SP and Beir Belt+1 for VS, what else? Maybe I use Aesir for SP and Vulcan's Pearl for VS? Not really sure but I think that's about it.
Rest of the gear I have are 2% critdamage/wsdamage Moogle Hat, Cirque+2 hands, Cirque+2 feet, Usu legs, Brutal, Epona, Rajas, Pantin Cape, Ele gorget. Not the best but still pretty good I guess?

Last week I tested a bit while doing Grounds of Valor, over 500 WS used between VS and SP, I wrote down the numbers of every WS I was using, I didn't consider those with missed hits (as far as I could tell from TP returns) and avoided using Aesir Ear Pendant because of the conserve TP.
500 WS is not a huge number statistically, and the targets I chose were really bad for testing, I suppose. Bats and Leeches in Zeruhn Mines... So yeah, I guess my testing is hardly worth anything lol.
From those numbers I wrote the highest and lowest damage. SP's highest was ~2,9k, VS's highest was ~2,8k. SP's lowest was ~800 and VS's lowest was 1036 if I recall, both happened during the leech's attack debuff. I think I still have exact numbers wrote down in a file @ home.
The average number, to my surprise, was very close for both WS. I think SP's average was slightly better? But we're talking about tiny differences.

So from this little test, that hardly counts anything because of how I conducted it, it seems that I was wrong in defining VS more "consistent" than SP, they're close to each other and SP is slightly better (altough the difference seemed smaller than the one suggested by your Spreadsheet, but that's probably because my test had too many unwanted variables).

I also find it a bit hard to use VS for Aftermath and then SP while AM is up. That might be possible if you're in a party with Haste and March, but when you're solo or you don't have enough Haste, you don't get TP fast enough to keep the Aftermath up, so I found it easier to just use VS to maximize the uptime of the Aftermath.
Altough in all honesty I still get the feeling that AM for h2h isn't as relevant as it is for 2 handed weapons.
The proc rate is appearently the same, but it works only on our offhand and its related DA/TA.
This has been discussed several times, they went this way because they thought that the number of hits into a full h2h attack round is higher than the number of hits into a 2handed weapon attack round.
True.
Problem is our hits do half the damage of the hits of a 2 handed weapons. Our whole attack rounds are probably close to each other in terms of damage, but it's like 2 of our hits corresponds to 1 of their.
So our aftermath, when it procs, it's weaker than theirs.
To me this is unbalanced.
They should have reduced the % of aftermath proc (atm it's ~45% if I recall?) but make it so it can proc on all the fucking hits of your attack round.

What's your stance on this Motenten?


Oh and also, very off topic, I have a couple of questions to ask you about your SAM spreadsheet, how can I do that? PM on Allakazham? PM on FFXIAH? Thanks!