What's Better?

[eyedea]

Been second guessing my WAR RR and KJ sets lately. Anybody able to do the math and/or make suggestions would be appreciated.

Relevant info is Elvaan with merits being G.Axe 8/8, DA/Berserk 5/5. Speaking mainly for birds and Einherjar.

RR:

Perdu/Pole/XX/Bomblet
Heca/Snow/Brutal/Triumph
Zahak/Heca/Rajas/Triumph
Cuch/Warwolf/Haidate/Heca

Sometimes I feel I lack accuracy with this set. Was wondering if using Haub+1 and Adroit Ring would be better off using. Mind you, both sets keep me at DEX117 for crit cap.

Also for KJ im using the following:

Perdu/Pole/XX/Bomblet
Heca/Breeze/Brutal/Triumph
Zahak/Heca/Rajas/Triumph
Cuch/Warwolf/AF2/Heca

Should I throw on Haub+1 instead? Any other pieces I should be going after?

Playing on PS2 so I'm not able to parse or anything. I just kind of eyeball it.

[Hegarty]

Also for KJ im using the following:

Perdu/Pole/XX/Bomblet
Heca/Breeze/Brutal/Triumph
Zahak/Heca/Rajas/Triumph
Cuch/Warwolf/AF2/Heca

Should I throw on Haub+1 instead? Any other pieces I should be going after?

Playing on PS2 so I'm not able to parse or anything. I just kind of eyeball it.

RR looks ok but for this set you should use Haubergeon +1 (relevant stats being Accuracy+16, Attack+16, STR+6) or Hecatomb Harness (Accuracy+10, STR+12, Attack+9) depending if you need the 6 Accuracy or not, gaining this accuracy would allow you to change your back to Foragers (Amemet+1 for gypsys or those with RNG 75) for the attack which far outweighs 1 STR.
You should also try and get a Warrior's Stone at which point Haub+1/Heca, WAR stone and Cuch mantle with Alky braces MIGHT be better than Heca hands.

[Callisto]

I would also recommend the Haub+1, KJ performs best at the highest possible levels of attack compared to RR(if you aren't having Berserk up and getting at least one attack buff like Chaos I would use RR), the crit rate isn't doing anything and the extra acc will let you make up the STR elsewhere such as Alky's.

[FailureMidgard]

I already answered this part. You are reproaching me for not to allowing the fumas monk to use usukane on WS, but why would you ? The OP asked "what's better" for TP, without mentionning what you would assume in other slots. If you want to compare the damage of MONK1 and MONK2, monk1 having fumas and no usukane and MONK2 having usukane you have to do the computation the way I did it first. At this point, and you admitted it your last post, it's obvious that you only wanted talk shit, by trying to discredit my maths by pointing out a completely irrelevant detail that doesn't influence the result. I mean if the question was "how much better is usukane > fumas" this would be relevant, but here we're talking what's better. You can WS with usu feet, no feet or w/e you feel like putting in that slot, this will have ZERO impact, so there is absolutely no point starting shit in the thread for the ONLY purpose of doing a negative post.





This is exactly what I said, you don't understand some part of the math and therefore you claim it wrong. I mean, talk to your mog, go punch some bunnies, and you will see that monk's attack go by pair for the most part. Why do I add 1/2 of a hit you say ? It's fairly simple. Let's say that I'm am in a situation where I need 15 hits (using my computation method) to get 100 TP, which means a minimum 15 hits. MNK is not SAM. The 15th punch comes with a total round, where you can

- miss one of the hit
- get kicks or Double attack procs

What I claim is that due do the randomness of miss/DA/KA occuring in the previous attack rounds, you can never be certain that this 15th hit is part of the the 1st swing/2nd swing/DA of 1st swing/DA of 2nd swing/KA. A sequence of hit/miss/hit/hit etc allowing you to reach 100+ TP can be obtained equally (approximately) by having the 15th hit be either the first swing or the 2nd swing. As a consequence, and this is an approximation, mainly due to not accounting for the DA/KA part on the last round,

- 50% of the time you are getting [ 15 th hit (*) , 16 th hit ] on your last melee round where (*) is the hit at which you reach 100 TP for the first time, but you require 16 hits to finish that round.
- 50% of the time you are in the [ 14 th hit , 15 th hit (*)] configuration, where 15 hits are enough.

So on "average", you need 50% * 16 + 50% * 15= 8 + 7.5 = 15.5 hits.

There is the 1/2 hit that I added. If I were using Faith bag. I would have added 1 total hit, I'll let you guess why.. You know, this is what is is usually referred to as TP loss. With a normal weapon it's about 0.5 hits and with faith it's 1.0 hits (once again, it's an approximation). I'll let you compute, even approx., what happens for Kclub/ridill and other joy toys.

Ok, so basically you have no clue what the word average means, gotcha. The "correct" way to do the math is still to use the averages without adding that half a hit. Your numbers are still way off, and if you think your 50% is right, then you shouldn't be using the 2.25 swings per round, because that's in direct contradiction with what you just posted.
The numbers I posted are correct. It should be 7.65 and 7.09, meaning a 7.9% increase. Your adding of half a hit is wrong, and goes against the entire idea of using the average number of rounds. As you said in your last post, if you model how many rounds it will take with misses/etc, you'll get the right number, that number will be one I am giving, not one where you add .5 hits

Finally, there is nothing wrong with pointing out faulty math on these forums. The original poster did ask which is better, but if you want to go with your assumption that he doesn't have Usu feet right now, it is entirely relevant for him to know how much better before he decides to join a salvage group. It is also insulting to people who do math correctly when you come on here and butcher shit like that. This same thing happens virtually every time you post, and sometimes it actually does matter.

Finally, for the record, if you want to estimate and guess, here's what you should do:
If you are trying to prove A > B, and you have to guess at some stuff on B, we'll call this B', then give B' every benefit of the doubt, and overestimate it, this way, if you prove A>B', and B'>B, then A>B without a doubt. If you need to make assumptions at A, underestimate it all, so that if A>A', but A'>B still, then again, without a doubt, A>B. This is simple logic, and when you have to estimate things, this is how you should be doing it. But that being said, you still have no reason to add half a hit to averages when all we are concerned with is averages...Either leave them as decimal averages, or use some approximation method to round it to nearest rounds, because again, the averages I posted are the exact averages it takes to hit 100 TP.

[FailureMidgard]

Oh ya, as far as knocking the method but not the conclusion, pretend that someone went to a merit party and said "So, I went to an hour merit party, for the first 30 minutes I used Usukane feet, and for the second 30 minutes I used Fumas...I didn't parse or anything, but I have a keen eye for these things, and Usukane feet are definitely the better choice!".

Just because they may have the correct end result does not mean that anyone should give a shit what they said, and I would fully expect you and others to inform them (and others before they take the advice) of the problem.

[ZhiGaruda]

Finally, for the record, if you want to estimate and guess, here's what you should do:
If you are trying to prove A > B, and you have to guess at some stuff on B, we'll call this B', then give B' every benefit of the doubt, and overestimate it, this way, if you prove A>B', and B'>B, then A>B without a doubt. If you need to make assumptions at A, underestimate it all, so that if A>A', but A'>B still, then again, without a doubt, A>B. This is simple logic, and when you have to estimate things, this is how you should be doing it.

That's a mighty fine example, actually.

[Ezhno]

Sentinel Shield vs any other endgame shield in a situation where atonement is viable(ie the only time to use pld tanks)

I've been thinking about this too for a little bit and here's my thoughts:

The biggest issue is loss of % reduction of damage per block for any hits that may get through shadows, but when I looked at that even that doesn't seem like that much. If you go based on the Shield Asstery thread which cited the formula Block PDT = 50 (Kite shield base) + (Shield's Defense / 2), then you're only looking at a 6% difference during blocks (61% v. 55%).

That difference is obviously going to get less and less the better you can keep shadows up, and the 3 enmity won't be an issue with Atonement spam and a Cure Cheat.

I've been using Sentinel for a while for Limbus bosses (the only things I go to anymore that would even merit a PLD tank), so I'd say go for it.

[Yugl]

Been second guessing my WAR RR and KJ sets lately. Anybody able to do the math and/or make suggestions would be appreciated.

Relevant info is Elvaan with merits being G.Axe 8/8, DA/Berserk 5/5. Speaking mainly for birds and Einherjar.

RR:

Perdu/Pole/XX/Bomblet
Heca/Snow/Brutal/Triumph
Zahak/Heca/Rajas/Triumph
Cuch/Warwolf/Haidate/Heca

Sometimes I feel I lack accuracy with this set. Was wondering if using Haub+1 and Adroit Ring would be better off using. Mind you, both sets keep me at DEX117 for crit cap.

DEX: 5+10+4+5+4+5+15+3 = 51 + 65 (Base DEX for sub SAM) = 116DEX
ACC: 282 (Skill) + 87 (DEX) + 5 + 6 + 10 + 5 = 395

Short for birds by 12ACC. In Einherjar, where you can use Pizza, you're fine for ACC. Your DEX, however, is lacking in either case (Especially Einherjar when we discuss higher tier mobs).

With this, you have two good options:
Make two separate sets; One for Einherjar and one for Colibri
Colibri: Hauby+1; Adroit Ring; Minuet Earring
Einherjar: Minuet or Pixie Earring

Make one single set for both
Both: Use the Colibri set listed above

Also for KJ im using the following:

Perdu/Pole/XX/Bomblet
Heca/Breeze/Brutal/Triumph
Zahak/Heca/Rajas/Triumph
Cuch/Warwolf/AF2/Heca

Should I throw on Haub+1 instead? Any other pieces I should be going after?

Playing on PS2 so I'm not able to parse or anything. I just kind of eyeball it.

Anywhere you can use food, you're using Pizza, so the goal is to reach roughly 390-392 (I Prefer 392 to include Troll MNKs and BLU Mamools).

ACC: 282 (Skill) + 75 (DEX) + 5 + 6 + 10 + 5 = 383 ACC

Looking for 9 ACC in your case. Here's a couple choices:

1. Hecatomb Body; Warrior's Stone; Ares's Legs (Best choice in my opinion)
2. Hauby+1

[LordTrey2]

On DRK before 60, GA + shield break (with a few sturmwinds thrown in) or other weapons and WS?

[pchan]

Ok, so basically you have no clue what the word average means, gotcha. The "correct" way to do the math is still to use the averages without adding that half a hit. Your numbers are still way off, and if you think your 50% is right, then you shouldn't be using the 2.25 swings per round, because that's in direct contradiction with what you just posted.
The numbers I posted are correct. It should be 7.65 and 7.09, meaning a 7.9% increase. Your adding of half a hit is wrong, and goes against the entire idea of using the average number of rounds. As you said in your last post, if you model how many rounds it will take with misses/etc, you'll get the right number, that number will be one I am giving, not one where you add .5 hits

No, you still didn't understand what TP excess is; 7.65 and 7.09 are the average number of rounds you would get only when you are getting 100 TP on the last hit to reach 100 TP, half of the rounds will come with an extra attack round (at least) which you absolutely don't consider in your calculation. I'll give you a pretty simple exemple :

Let's say you get 100.0 per melee hit and 0.0 TP return on every weaponskill. 95% acc, 0% DA and 0% KA. But you have exactly 2 swings per round.

What you claim is the following

(100-0.0)/100.0= 1.0 hits (this is your 16.36) required
1.0/(2.0*0.95)=0.526 rounds required (this is your 7.65)

...Do I really need to explain why you need a bit more than 1 round on average ?

[Eole]

Hume Samurai Targetting Greater Colibri

= Tactical Points =

main "Tomoe"
sub "Pole Grip"
ammo "Black Tathlum"

head "Ace's Helm"
neck "Ancient Torque"
ear1 "Fowling Earring
ear2 "Brutal Earring"

body "Usukane Haramaki"
hands "Askar Manopolas"
ring1 "Rajas Ring"
ring2 "Toreador's Ring"

back "Cuchulain's Mantle"
waist "Virtuoso Belt"
legs "Byakko's Haidate"
feet "Usukane Sune-Ate"

= Penta Thrust =

main "Tomoe"
sub "Pole Grip"
ammo "Black Tathlum"

head "Champion's Galea" (Weapon Skill Accuracy +15 Accuracy +10 Attack +5 STR +4)
neck "Shadow Gorget"
ear1 "Fowling Earring"
ear2 "Brutal Earring"

body "Usukane Haramaki"
hands "Bandomusha Kote"
ring1 "Rajas Ring"
ring2 "Flame Ring"

back "Cuchulain's Mantle"
waist "Virtuoso Belt"
legs "Byakko's Haidate"
feet "Usukane Sune-Ate"

What's better than this (in the obtainable equipment category) ? Why ?

[Hyriu]

Swift/Velocious/Speed belt.

Dusk Gloves.

[Eole]

Swift/Velocious/Speed belt.

Dusk Gloves.

Haste +5/+7 vs DEX +5 Accuracy +9 (+3.75 = 12.75)

Let's say we do 1024 Attacks.

(Hasso) 91% Hit Rate +5/+7 Haste vs 95% Hit Rate
1095 Attacks dealt (985 with 90%HR) with +7% Haste vs 1024 (972 with 95%HR)
1075 Attacks dealt (978 with 91%HR) with +5% Haste vs 1024 (972 with 95%HR)

(Seigan) 86% Hit Rate +5/+7 Haste vs 92% Hit Rate
1095 Attacks dealt (930 with 85%HR) with +7% Haste vs 1024 (942 with 92%HR)
1075 Attacks dealt (924 with 86%HR) with +5% Haste vs 1024 (942 with 92%HR)

Are my calcultions right ?

[Fottiti]

No.

Haste gives increasing returns and you're calculating assuming no other gear haste (which you've already shown you have in your tp setup) or spell haste which is completely unrealistic. You didn't even include hasso's 10% haste in your calculations.

If you want the "why" you can work it out for yourself quite easily. There is no need for hypothetical hit rates you are quoting when you are targetting greater colibri as their stats are listed on wiki.

These calculations aren't complicated and for colibri its been shown here on BG many times already.

[Kuroyou]

for Samurai:

Acc +7, STR +4, DEX +6, EVA -32, AGI -5
vs
Double Attack +4%

I have 8/8 Great Katana Merits.

Can you give any info/help on which, and why? Try to look at the spoiler after deciding on the stats, to avoid blind gear-bias

Spoiler: show

Strap: Rose Strap vs Pole Strap
Neck: Chiv Chain vs. Peacock Amulet
Body: Haubergeon +1 vs. Askar Korazin

-- CURRENT --
Hagun / Pole Strap / None / Balm Satchet
Wal. Turban (WS: M.Cap) / Peacock (WS: Gorgets) / Brutal / Fowling
Askar Body / Askar Hands (WS: AF+1 Hands) / Ecphoria Ring / Rajas
Foragers / Swift Belt (WS: Warwolf) / Byakko's (WS: Shura) / Fuma Legs (WS: Rutter)

The real question, I'm trying to decide if the benefit of the stats in the first set, is really worth the 30,000 AN. Or losing the 4% DA.

Also, because of macro limits, I'm not able to swap out body items for WS. So factor in that the body that I wear, will have to do for WS as well, wether I like it or not.

I ask that the answers stick to the gear/stats I'm comparing without suggesting any items outside of that, please.

Thanks for any help

[Cream Soda]

4% DA is going to be less than a 4% increase if you have any other DA, or a 4% increase if you have no other DA. 7 acc is going to be a 3.5+% increase before even looking at any of the other stats, provided you're in a situation where you can use all 7 points of acc. Then you have the other stuff to throw in. including the dex, which gives more acc. (And you only need 8 acc to 4% hitrate = 4+% increase if you can benefit from all 8 points of acc). Easily your first choice.

[Kuroyou]

4% DA is going to be less than a 4% increase if you have any other DA, or a 4% increase if you have no other DA. 7 acc is going to be a 3.5+% increase before even looking at any of the other stats, provided you're in a situation where you can use all 7 points of acc. Then you have the other stuff to throw in. including the dex, which gives more acc. (And you only need 8 acc to 4% hitrate = 4+% increase if you can benefit from all 8 points of acc). Easily your first choice.

Thanks for the breakdown, I really appreciate it. Also I think I only need 11 STP for the 6-hit right (forgot to mention in OP I'm merited 5/5 STP, but that maybe was assumed already)? Rose +4, Brutal +1, Ecph. Ring +1, Rajas +5. That makes 11. So I may be able to keep the PCA for an extra +5 Accuracy even still. Was thinking the the Chiv +1 would give me a safe +12 in case, but may not need it.

So my next Q, is, do I need all that ACC for typical say, merits (doubt)? And what about EG NMs? I usually eat those Coeurl Subs as they aren't too dare.

[Wolfknight]

Rosa + Haub + PCC, you don't need chiv chain since you're using Ecphoria Ring because rajas[5] + brutal[1] + rose[4] + Ecphoria[1] = 11 you need for in TP as long as you ws in the same amount, even then you're better off swapping AF+1 legs, or hachiman sune-ate[or body] in instead of that if you're changing rings, and also I am assuming PS2 reason for low gear swap in which case I'd recommend using the macro book/pallet swap command to "chain" macro gear.[ie first one 5 swaps > changes pallet up one to next swap, and then other leads to same thing for TP gear, so you just hit enter few times for more gear swapping.]
Edit: unless you're using pizza you'll probably want the acc in most places, and even then it's hard to say.

[Cream Soda]

Most likely not if you're eating pizza. Though, an observation. You have pcc vs chiv. You have enough stp w/ rose strap, rajas, ecphoria, brutal for 6hit, so no need to give up 5 acc.

[seiji]

Rosa + Haub + PCC, you don't need chiv chain since you're using Ecphoria Ring because rajas[5] + brutal[1] + rose[4] + Ecphoria[1] = 11 you need for in TP as long as you ws in the same amount, even then you're better off swapping AF+1 legs, or hachiman sune-ate[or body] in instead of that if you're changing rings, and also I am assuming PS2 reason for low gear swap in which case I'd recommend using the macro book/pallet swap command to "chain" macro gear.[ie first one 5 swaps > changes pallet up one to next swap, and then other leads to same thing for TP gear, so you just hit enter few times for more gear swapping.]
Edit: unless you're using pizza you'll probably want the acc in most places, and even then it's hard to say.

no, 9stp is enough to WS in (rose+rajas(+brutal)):
11.5*1.46]*5]+11.5*1.44=16.7*5]+16.5=83.5*16.5=100!