Gear Discussion/Help SAM

Woot, personal attacks, sorry buddy, been playing this since day one NA release and I obviously understand the math that goes into it better than you.

Honestly, he has to be a bought account. He kept using a set as if he used it when it didn't even give 100tp. Sorry but that's a major flag that he doesn't either have it or never used it. People argue for no reason too much for him to need a reason. Imo all he did was toss a few numbers in something and wanted to back it up, which since his inital claims were fail he obviously has no experience to back it up.

I was posting as he entered that (took me a while since I am multi-tasking while posting here). And while it backs those claims which even true still are flawed as being better, it really doesn't change anything. Even getting capped acc on zanshin hits, 3% haste and double tp on 38% of misses will not cover those missed attacks.

He doesn't want parses most likely b/c he knows the results. He can throw numbers he never actually came up with from experience all day, idc, his credibility is nil as far as I'm concerned.

I'm sorry but unless you're getting extreme amounts of outside haste it's not. You can think you understand it all you want. You have made far too many critical errors for me to believe you. Try using the job on the char you're playing and show us some parses showing what you claim, then I'll listen. You're going to need more haste than 3% to make up that 12%.

You say you have but what you've shown so far is reliance on numbers you never used and false claims that you kept having to change when shown false. You basically show traits of a person that has not played the job.

Lol, I've also addressed this, you really should try reading my posts instead of just spouting more stuff. This is not the set I actually use, my personal set is somewhere in the middle of page 3. I created this as a set that should be not too hard to obtain but still is similar in criteria to my actual set and that's why I didn't know it didn't work as is. And I discussed with Vincent the solution to it, though atm I'm not sure which of our math is off for tp calculation.

You say that all those factors won't cover it, but where's the proof? Your doing this just because you "think" that it's not enough to cover it. But my math shows that it does, so please, if the math is wrong, find where it's wrong, otherwise you're not adding anything useful.

And like I said, I don't trust parses, there's too much human involvement. Math is what it is, if there's something wrong with the calculations, you can fix em, but you can't do much about all the human factors that go into a parse.

I think this could work well, just so that on the initial miss, you throw on all your ACC gear so not to hit both hits. I mean you probably have like a .001 second time frame to do it which is plenty.

The math was posted earlier that you need an amount of haste that surpasses what the gear cap+ haste spell give you. Or did you totally skip that post?

With server lag gl macroing before the server actually calculates it.

What are all these critical errors? There's one which was the store tp thing and that was fixed without affecting anything in the calculations. All you see is 12% and 3% without looking at them in context. Haste by it's nature, the more you have, the better it gets, it's not additive. The 12% on the other hand, if you look at it in the overall picture, it's 12% of the times you missed, so it's 12% of 13% (assuming 87% acc) which is what, 1.56%, suddenly that haste ain't looking so bad eh?

No, I responded to it with all the things that were wrong with it, he got back to me, and I responded with some things he still had wrong, currently waiting on his response to that.

Edit: The following was edited to correct an omission in the melee data.
--

I have had a request to review the base math posted by Tarowyn earlier in this thread. Below is my analysis of the comparison between the two configurations (askar + dusk vs. hauby + hachiman), with some adjustments for factors left out.

I'm sure many people will tl;dr the math part. That's none of my concern. However if you wish to say that this is wrong, please indicate exactly where there is an error, and show your work.

Given that Ikishoten's TP is applied to the base TP of the weapon before Store TP, 4 Ikishoten merits gives 2 hit's worth of TP on a Zanshin hit (+12 TP on top of the base of 11.5 TP);

Given a 40% proc rate on Zanshin, and the presumed capped Zanshin accuracy if base accuracy is 78% or higher due to the acc boost;

Given that Haubergeon puts you at exactly capped accuracy;

Hauby hit distribution:
95% hit
3.1% miss
1.9% zanshin hit

Askar hit distribution:
88% hit
7.44% miss
4.56% zanshin hit

Total avg hits' worth of TP per round:

Hauby: .95 + 2 * .019 = .988
Askar: .88 + 2 * .0456 = .9712

w/DA (assuming Brutal & Pole Grip only)
Hauby: .95 * 1.05 + 2 * .019 = 1.0355
Askar: .88 * 1.07 + 2 * .0456 = 1.0328

w/DA (as /war with Brutal & Pole Grip)
Hauby: .95 * 1.17 + 2 * .019 = 1.1495
Askar: .88 * 1.19 + 2 * .0456 = 1.1384

Value of Zanzhin hit reduced to a single-hit 20% of the time when it is the last hit to 100 TP (going from 83 to 116 TP). While it technically also improves weaponskill damage from the higher TP, I'm not going to try to calculate that.

w/War +Brutal & Pole Grip
Hauby: .95 * 1.17 + 2 * .019 * .8 + .019 * .2 = 1.1457
Askar: .88 * 1.19 + 2 * .0456 * .8 + .0456 * .2 = 1.1293

This gives the average number of hits' worth of TP per attack round, and can be directly converted to weaponskill rate as a ratio of hits needed to weaponskill. Since both are 6-hit builds, a/b works fine.

Difference of 1.45% on TP, favoring Hauby.

Now apply the differences in haste from dusk vs hachiman.

Hauby: turban (5%) + b.haidate (5%) + fumas (3%) + swift belt (4%) = 17%
Askar: as above + dusk gloves (3%) = 20%

TP rates without any buffs (solo or generally unsupported):
Hauby: 1.1457 / .83 = 1.380
Askar: 1.1293 / .80 = 1.412

Advantage: Askar by 2.3%

Buffs can provide an additional 26% haste in a moderate haste setting (using the spell + 11% March, and ignoring Hasso).

Total haste:
Hauby: 43%
Askar: 46%

TP rates then shift to:

Hauby: 1.1457 / .57 = 2.010
Askar: 1.1293 / .54 = 2.091

Advantage: Askar by 4.0%

That pretty much gives you the range of difference in damage from the weaponskill half of your damage equation. Then calculate out the difference in melee damage.

Hauby: +14 acc (10 acc + 5 dex, assumed rounded up), +4 str (hachiman)
Askar: +4 att (5 dusk + 2 askar - 3 hachiman), +2 DA, +3% haste (dusk)

+4 att is worth approximately 1.0% melee damage, long-term.

Given 100 attack rounds, and a comparitive baseline of 100 = 100 swings @ 100% acc with a hagun:

Hauby: 100 * 1.17 (DA) * (.95 + .019) (acc+zanshin) * (81/80) (+1 fStr, assuming base of +5 fStr with hagun) = 114.790
Askar: 100 * 1.19 (DA) * (.88 + .0456) (acc+zanshin) * 1.01 (att) = 111.248

Then apply haste:

No extra haste:

Hauby: 114.790 / .83 = 138.30
Askar: 111.248 / .80 = 139.06

Advantage: Askar by 0.55%

Moderate haste:

Hauby: 114.790 / .57 = 201.39
Askar: 111.248 / .54 = 206.01

Advantage: Askar by 2.3%

Assuming a 50/50 split of melee and weaponskill damage*, net relative gains gives:

No extra haste:

Hauby: +0%
Askar: +1.4%

Relative: Askar by 1.4%

Moderate haste:

Hauby: +0%
Askar: +3.15%

Relative: Askar by 3.15%

*Aside: a 60/40 weaponskill/melee split is also possible/probable. Given the above numbers, however, that further favors Askar since it's further ahead on the weaponskill side.

However, the above puts the value of the Hauby's acc at the lowest relative worth without actually going over cap. Zanshin is supposed to be strongly worthwhile at 78%+ acc, so I'll redo the calculations assuming Askar has a base accuracy of 80% and Hauby has a base acc of 87%.

Hauby hit distribution:
87% hit
8.06% miss
4.94% zanshin hit

Askar hit distribution:
80% hit
12% miss
8% zanshin hit

Total avg hits' worth of TP per round w/war sub, brutal and pole grip:
Hauby: .87 * 1.17 + 2 * .0494 * .8 + .0494 * .2 = 1.1068
Askar: .80 * 1.19 + 2 * .0800 * .8 + .0800 * .2 = 1.0960

Weaponskill side:

No extra haste:

Hauby: 1.1068 / .83 = 1.333
Askar: 1.0960 / .80 = 1.370

Advantage: Askar by 2.8%

Moderate haste:

Hauby: 1.1068 / .57 = 1.942
Askar: 1.0960 / .54 = 2.030

Advantage: Askar by 4.5%

Melee side:

Given 100 attack rounds, and a comparitive baseline of 100 rounds = 100 swings @ 100% acc with a hagun:

Hauby: 100 * 1.17 (DA) * (.87+.0494) (acc+zanshin) * (81/80) (+1 fStr, assuming base of +5 with hagun) = 108.914
Askar: 100 * 1.19 (DA) * (.80+.08) (acc+zanshin) * 1.01 (att) = 105.767

Then apply haste:

No extra haste:

Hauby: 108.914 / .83 = 131.22
Askar: 105.767 / .80 = 132.21

Advantage: Askar by 0.75%

Moderate haste:

Hauby: 108.914 / .57 = 190.90
Askar: 105.767 / .54 = 195.86

Advantage: Askar by 2.6%

Assuming a 50/50 split of melee and weaponskill damage, Net relative gains gives:

No extra haste:

Hauby: +0%
Askar: +1.8%

Moderate haste:

Hauby: +0%
Askar: +3.55%

So in high accuracy situations, Askar wins by 1.4% to 3.2%, and in moderate accuracy situations Askar wins by 1.8% to 3.6%. Higher haste would push it further ahead, as would using a 60/40 split instead of a 50/50 split.

Hauby +1 should be about 2.5% better than the NQ, which makes it roughly competitive with the Askar results.

The above is based on a number of assumptions (eg: that you -always- weaponskill at 100 TP) and 'given's (eg: Zanshin proc rate, that you've merited Ikishoten, etc). Should any of them be incorrect, it may skew the results in various unidentified ways.

Also, someone please explain why several people keep insisting that 46 Store TP (5 Askar + 5 Rajas + 1 Brutal + 25 trait + 10 merits) gives you 16.6 TP per hit. 11.5 base TP per hit * 1.46 = 16.79 == 16.7 TP per hit.

I'm not seeing your gear set. I don't see how you're getting that high of acc on something like a colibri with askar though.

The only time I saw any math for zanshin you had full merits in it. I'll re-read some tomorrow, but anything over 3 merits in Iki needs to be thrown out.

I've never said you need full iki merits, just 4. Some other people have mentioned it but its not required for a 450 GK.

It's funny, I'm of the opposite mind, anything below 4 Iki merits is just wasted to me, lol.

Blade bash and shiki is too helpful not to have.

Blade Bash I can kind of understand (and I think during the course of this thread I decided I'd put my last merit in there), but I figure I can live without Shiki. I guess it depends what you use your SAM for, but considering the avg DD doesn't even take a minute to get TP, not having it just won't bother me that much.

I didn't see the point of shiki until I got it. There are many times before a fight that giving someone some tp to start off with helps.

Besides the numerous times you get requested to shiki someone, it's kind of nice to be able to give someone a tp boost so they can ws and you can sc off them. Espc helpful if you're helping someone break a ws latent, for example save to almost 300 tp while someone is getting up to 100 tp (not too hard with a 140-160 meditate), sekka > sc with them > shiki them > sc. It can really speed up that kind of thing.

I wouldn't put more than 1 in it though(I've known sams that put 5 >.>), anything past that I find to be a waste.

I can't really see being able to sacrifice either tbh. I'd rather 4 bash than iki for faster stuns personally.

I tried to make the case for Shiki merits many times, but outside of the occassional situation of wanting to boost a PLD's TP for Chivalry or beginning a SC quicker. Blade Bash on the other hand has saved my *** numerous times. lol I know a few argued a SAM shouldn't ever need to do the stunning, but many times ***just doesn't work out perfectly and playing BLU taught me the true beauty of stunning capability. lol

Oh and gawdamn, Moten. :Q__ Thank you for taking the time do all that!

I haven't checked his numbers but off hand the only thing missing that I see is the 5 dex affecting critical hit rate. Probably not a big difference but it is there.

Shiki's kinda cool to play with and stuff as you mentioned, but the timer on it just detracts from it's actual usefulness I feel like.

And the dex thing is probably very hard to calculate since it's a curve unlike acc and dmg being pretty much linear. SAM doesn't really focus much on crits anyways so yeah, don't see it being a major factor.

Lol, Mote, one tiny thing I noticed, are you not including Zanshin hits for melee dmg at all? Pretty small difference but just noticed it rereading through your post.

Edit:

Did a sample calc, for Askar at 88% this look right? (.81 inverse of DA ratio)

100*1.19*.88*1.01 + 100*.81*.95*.0456*1.01 = 109.311

Also noticed that for TP calculations, I think you're not taking into account the lowered percentage for zanshin due to DA where zanshin and DA can't overlap? Man, totally didn't want to include DA because it makes things even more complicated!

I indeed erred. Can't believed I missed such a basic element. I'll correct the above post momentarily.

Wait...did I just read something to the effect that SAM's do not focus on critical hits?

Lol, why aren't you asleep Mote?



Yeah, there are very serious crit heavy builds out there that people use for war and nin, but since sam's less dot focused and more ws focused, better off keeping 6 hit and piling on the haste.

The above data has been corrected to account for melee damage due to zanshin. This notably changes the balance of damage done, and actually puts melee damage in Askar's favor instead of Hauby's, giving the Askar set a strong overall lead in the range of 1%-4% depending on haste.

The difference is actually a bit surprising to me, as I would not have expected that much of a lead with the Askar setup even under really good conditions.

However this does still depend on two aspects of Zanshin that haven't been fully verified (there's enough variance in the results of the page Tarowyn linked to that I would at best call the proc rate 'tentative'), so there's still a degree of uncertainty in the final conclusion.

And for DA/Zanshin overlap... I thought I had it properly separated, but may have made a mistake there as well. I'll need to look into it further.

Assuming Zanshin is determined first, as it's pretty much impossible to say which takes priority.

1000 swings

950 connect
17% of those (161.5) also DA, of which 95% (153.425) connect.

50 miss.

40% (20) of those Zanshin. They cannot DA.

60% (30) do not Zanshin. There is a 17% chance of DA on those (5.1), of which 95% connect (4.845).

Total non-Zanshin swings:

Hit: 1,108.27
Miss: 58.33
Total: 1,166.6 (effective 16.66% DA rate)

Hypothesis:
DA rate == DA% * (1 - Miss% * Zan%)

If so, 88% hit rate and 19% DA rate should produce 18.088% DA rate.

1000 swings

880 connect
19% of those (167.2) also DA, of which 88% (147.136) connect.

120 miss.

40% (48) of those Zanshin. They cannot DA.

60% (72) of those do not Zanshin. There is a 19% chance of DA on those (13.68), of which 88% connect (12.0384).

Total non-Zanshin swings:

Hits: 1,039.1744
Miss: 141.7056
Total: 1,180.88

Confirmed: 18.088% DA rate.

I'll work out how to apply that to the calculations tomorrow.

This is what I want to know. :V

From my experience on sam/war, I wanna say that DA gets checked first, and if that happens zanshins just can't happen. As a programmer I'd also look at it as, DA check is a pre round check while zanshin checks after a swing actually occurs, but yeah, it'd be hard to figure out which way it goes based on pure testing alone and you pretty much need iki to be able to tell the difference. Though you're pretty much out of luck if both hits miss, lol.

DA checked first is also easier to calc right? Guess whichever works since we're not sure.

So many personal attacks in this thread :( let's attack each other's arguments, not the individual themselves!

I combined everything as damage% just to save me the effort honestly <_< and you're right that the TP/WS damage ratio will be different from mob to mob but if you look at the fights at the very extreme end that you simply Meditate and then run in, Meditate favours neither anyway because neither are restricted on WS gear.

Askar will win slightly if you are only fulfilling NQ Haubergeon's base accuracy exactly it seems which is honestly different to how I thought previously, so this thread has been a valuable discussion. Also important to note is that this discussion concerns primarily Great Katana where WS damage is all-important and meat only really serves to increase melee DoT damage if using Gekko.

So basically, as conclusions it seems that:

NQ Haubergeon:
High accuracy and/or high haste: Askar
One or other lacking: Hauby
HQ Haubergeon:
Pretty high accuracy and/or pretty high haste: Askar
One or other lacking: Hauby

/facepalm x1000

This is exactly why this discussion should have ended pages ago. Askar only wins under very specific conditions with a ton of assumptions about Zanshin's proc rate and acc boost. You're also(from what I can see) assuming an 88% hit rate while using askar which is unrealistic for the average SAM. However you just posted that "conclusion" without stating any of that and a lot of SAMs who shouldn't be using Askar will think it's better than Haub(when in most cases it is not).

I actually disagree with the conclusion but for different reasons as I would say Askar actually came out better looking than this conclusion stated, lol. I would like to say again that this is for a VERY SPECIFIC case. You must have at least 4 Ikishoten merits and are using Askar to gain 3% haste over using Hauby while maintaing 6 hit. You must also have a reasonably good haste set to begin with or getting lots of outside haste, this case was tested with 17% haste with Hauby and 20% with Askar. In the hope that any SAMs reading this won't get the wrong idea, hehe.

As for the 88% hit rate, that's based off capped acc (95%) using Hauby and then subtracting the amount of acc that hauby gives. Mote also ran the numbers on 80%/87% hit rate above while still retaining pretty much the same results. Does 80% sound fairer?