Looking to beef up the power of my BLU

You may have responded to the posts, but you haven't taken to heart what has been presented. I'll summarize your responses for you.

'I don't need accuracy or attack gear, because I would have to modify my macros, and maybe even press an additional macro in the process. I'm satisfied with my mediocre melee damage and 88% accuracy while using sushi, because I mistakenly believe that the effect of my melee damage is negligible when compared to my spell damage. I realize that I'm the only one who believes this, but I don't care.'
 
Nyosan~ said:
You may have responded to the posts, but you haven't taken to heart what has been presented. I'll summarize your responses for you.

'I don't need accuracy or attack gear, because I would have to modify my macros, and maybe even press an additional macro in the process. I'm satisfied with my mediocre melee damage and 88% accuracy while using sushi, because I mistakenly believe that the effect of my melee damage is negligible when compared to my spell damage. I realize that I'm the only one who believes this, but I don't care.'
Click to expand...

Zing.

yanzi said:
he doesn't have sea, so not like he can spam disseverment.
Click to expand...

In general, ACC rings will do more than the STR rings for him was what I was getting at.
 
Vilurum said:
Er wait, that 88% was with *sushi*?

That's, uh ... yeah ... if you're not hitting the 95% soft cap with sushi, there's something wrong.

Sword merits + acc gear + meat = win. Added perk is that meat typically costs less than sushi.
Click to expand...

88% acc isn't all that bad. I have a large acc build and usually hit around 90-93% on mamool ja's outside Nyzul Isle, but that's as WAR with a BRD while lacking many many merits. From my personal experience, the best exp parties I've had are where mobs are best fought using Sushi rather than meat. If you're using meat, it usually means the mobs are lower level and yield less experience.

Aside from that, I agree with Vilurum. In order to get the most out of your damage, you really need to focus both on DoT(melee) and spike (WS and spell damage). I don't know a whole lot about BLU, in fact, I really know nothing about it besides the most obvious and the overlapping factors with other jobs, but I don't think it makes the idea any less relevent. While focusing your damage on spells alone will improve your own damage in the long run, it's not neccessarily the best approach.

I noticed gil is an issue for you, so you should think of it like this: vast improvements in gear will require larger costs of money so improving both would normally be more cost efficient (I think Vil already hinted at that). Excluding pieces which will help both melee and spells, spending large amounts on major gear choices will often net less improvent in damage then spreading minor improvements throughout your equip set. A good example of this is buying a HQ hauby. A HQ hauby runs about 6 million, but if you spread that same 6 mil on other gear choices, you can get 2 snipers, Merman's (HQ coral), and HQ STR rings which will net more improvement in damage than the hauby+1 would for far less gil.

Apply this to BLU. Now, using all your gil to only part of your damage, in this case magic, is no different. If you're buying expensive gear which only effects the magic parts of BLU, you are missing out on the melee parts which can also benefit just as much from minor improvements especially when you might not have anything currently in your gear slots which improve melee stats. To me, the best way to go about making improvements is to buy the most cost effective improvements. Spend the gil on an item which will net you the most damage increase per gil. Good examples of this are hauby for melee and NQ staves for mages. Both give a noticeable improvement on damage, curing, resting, etc., but cost relatively little to obtain.

However, I will stress this a lot, everyone here is just making recommendations. You are free to take them of your own choice or not if you chose to do that as well, but most people here gain nothing from giving you this advice and nothing to lose either. In other words, it's in your best interest and no one else's to make the best decision for yourself so I suggest you go through what you have read and find what works best both as a whole and for you. At least if you have an educated decision based on facts and you decide on your choice based on a solid reason, then it's better than going blindly without any reason.
 
Lavi said:
88% acc isn't all that bad. I have a large acc build and usually hit around 90-93% on mamool ja's outside Nyzul Isle, but that's as WAR with a BRD while lacking many many merits. From my personal experience, the best exp parties I've had are where mobs are best fought using Sushi rather than meat. If you're using meat, it usually means the mobs are lower level and yield less experience.
Click to expand...

As a side note, it all depends on your setup and merit status. It's definitely possible for certain jobs to be using meat at Mamool ja.
 
yanzi said:
he doesn't have sea, so not like he can spam disseverment.
Click to expand...

This applies to Frenetic Rip and Hysteric Barrage, too. Also, obtaining Disseverment won't automatically yield a huge boost in performance. He'll still need to balance MP expenditures with MP recovery, it's not like he can just cast Disseverment inbetween the other spells and call it good.
 
dedalius said:
As a side note, it all depends on your setup and merit status. It's definitely possible for certain jobs to be using meat at Mamool ja.
Click to expand...

Not if you plan on keeping chain while killing THF and NIN mobs. Using meat increases the likelihood you will lose chain and even with a fairly good accuracy build while bergressor is up, I still notice a good drop in accuracy on THF mobs and when it's down, I pray to keep that chain alive.
 
Lavi said:
Not if you plan on keeping chain while killing THF and NIN mobs. Using meat increases the likelihood you will lose chain and even with a fairly good accuracy build while bergressor is up, I still notice a good drop in accuracy on THF mobs and when it's down, I pray to keep that chain alive.
Click to expand...

Like I said, depends on your setup, the people I go with just throw a Feint on the lurkers and we're fine. NIN mobs are never a problem for us.
 
Vilurum said:
Er wait, that 88% was with *sushi*?

That's, uh ... yeah ... if you're not hitting the 95% soft cap with sushi, there's something wrong.

Sword merits + acc gear + meat = win. Added perk is that meat typically costs less than sushi.
Click to expand...

Lol, saw that too. The Drk landed higher percentage of swings, though I thought it was impossible to breach 95%?
 
Lavi said:
Not if you plan on keeping chain while killing THF and NIN mobs. Using meat increases the likelihood you will lose chain and even with a fairly good accuracy build while bergressor is up, I still notice a good drop in accuracy on THF mobs and when it's down, I pray to keep that chain alive.
Click to expand...

Grab a thf and use feint^^ Even when I'm in mediocre parties a thf mamool lasts 20-30 seconds max when that hits.
 
Ewok said:
Lol, saw that too. The Drk landed higher percentage of swings, though I thought it was impossible to breach 95%?
Click to expand...

I've parsed 96 I think, and I believe Fly hit 97% once.
 
The soft-cap in game is 95% accuracy, but due to the random nature of hitting/missing, and non-infinite sample sizes you get numbers higher than that occasionally.

Edit: For those that don't understand what a softcap is, it is a programmed limit. Theoretically, you could stack ACC to increase your accuracy higher than 95%, however there is a programmed limitation on it to cause you to miss atleast 5% of the time. Just like in Diablo 2, you were unable to put more than 20 points into a skill, so it soft capped at 20 (naturally), however the actually data type used to store the number was able to go up to 255.
 
Emphasis on "soft cap". You can, I'd imagine, push past it, but you get major diminishing returns. Kinda like cure spells. (And yeah, there's the whole issue of randomness and statistics and sample sizes too.)

Plus your average will be drawn up by guaranteed hits like SA or Assassin TA. And multi-hit WS are practically guaranteed hits, because the probability of missing the entire multi-hit WS when your accuracy's 95%ish anyway, is abysmally low. So that'll drag up average too (and in the case of a merit party DRK, the use of multi-hit WS is more likely than the use of SA).
 
Vilurum said:
Emphasis on "soft cap". You can, I'd imagine, push past it, but you get major diminishing returns. Kinda like cure spells. (And yeah, there's the whole issue of randomness and statistics and sample sizes too.)

Plus your average will be drawn up by guaranteed hits like SA or Assassin TA. And multi-hit WS are practically guaranteed hits, because the probability of missing the entire multi-hit WS when your accuracy's 95%ish anyway, is abysmally low. So that'll drag up average too (and in the case of a merit party DRK, the use of multi-hit WS is more likely than the use of SA).
Click to expand...

Not exactly right. Your second paragraph is accurate, however the first paragraph isn't. Soft-cap is not a point of diminishing returns, that's the point of diminishing returns. A soft cap is an actual programmed limitation, i.e. someone in SE decided that you should miss at least 5% of the time. Similarly, you evade atleast 5% of the time. However, like I said, due to the random nature of the game itself you get fluctuations which sometime go outside the soft-cap extrema. If you were to parse an infinite number of hits you would never go above 95% ACC.
 
Last edited:
In that case, perhaps it would be better to describe it as a bypassable (see my second paragraph) hard cap.

A "soft cap" is exactly that - soft. Which is to say, you can keep on improving past it, it's just that the formulas are different so it's a lot harder to get improvement. Kinda like getting a Cure 1 to heal more than 30 HP by adding healing skill and MND - you can do it, certainly, but at nowhere near the skill/MND to +1 HP healed rate, that you got before you hit 30 HP. That's a soft cap. Pulling numbers out of thin air here, say +1 MND is +1 HP healed before you're healing 30, and +30 MND is +1 HP healed after you're healing 30.

What you're describing - a flat "if you *can* miss, you *will* miss 5% of the time" effect - is more akin to a hard cap. I was under the impression that melee accuracy worked in a soft cap fashion similar to cures (possible to push higher than 95%, but requires piling on way more accuracy than needed to get *to* 95%), but if it's a hard cap fashion as you describe, that makes sense too. But it wouldn't be called a soft cap in that case. :P

As for diminishing returns, that's more a matter of approaching a mathematical limit. Even if there's no soft cap in place, depending on formula it may simply not be possible to get to a point where you actually hit 100%. (It really depends on whether the formula is ultimately based on multiplication/division, or addition/subtraction, really.) As such, you can approach something but never actually reach it - adding a stat may make you approach it by a certain rate relative to what you were at prior to adding the stat, but it could be a very tiny effect in an absolute sense.

For example, going from whiffing 60% to 50% is - relatively - the same as going from 6% to 5%. You're whiffing 5/6 as much as you were before. However, in the latter case in an absolute sense, you're only whiffing 1% less, while in the former case you're whiffing 10% less. That's a diminishing return, and you run into them with practically every game formula that has to do with percentages.
 
Vilurum said:
In that case, perhaps it would be better to describe it as a bypassable (see my second paragraph) hard cap.

A "soft cap" is exactly that - soft. Which is to say, you can keep on improving past it, it's just that the formulas are different so it's a lot harder to get improvement. Kinda like getting a Cure 1 to heal more than 30 HP by adding healing skill and MND - you can do it, certainly, but at nowhere near the skill/MND to +1 HP healed rate, that you got before you hit 30 HP. That's a soft cap. Pulling numbers out of thin air here, say +1 MND is +1 HP healed before you're healing 30, and +30 MND is +1 HP healed after you're healing 30.

What you're describing - a flat "if you *can* miss, you *will* miss 5% of the time" effect - is more akin to a hard cap. I was under the impression that melee accuracy worked in a soft cap fashion similar to cures (possible to push higher than 95%, but requires piling on way more accuracy than needed to get *to* 95%), but if it's a hard cap fashion as you describe, that makes sense too. But it wouldn't be called a soft cap in that case. :P

As for diminishing returns, that's more a matter of approaching a mathematical limit. Even if there's no soft cap in place, depending on formula it may simply not be possible to get to a point where you actually hit 100%. (It really depends on whether the formula is ultimately based on multiplication/division, or addition/subtraction, really.) As such, you can approach something but never actually reach it - adding a stat may make you approach it by a certain rate relative to what you were at prior to adding the stat, but it could be a very tiny effect in an absolute sense.

For example, going from whiffing 60% to 50% is - relatively - the same as going from 6% to 5%. You're whiffing 5/6 as much as you were before. However, in the latter case in an absolute sense, you're only whiffing 1% less, while in the former case you're whiffing 10% less. That's a diminishing return, and you run into them with practically every game formula that has to do with percentages.
Click to expand...

No, a soft cap is a limitation which the programmers put on intentionally by limiting the amount of whatever something can do. In my Diablo 2 example, the programmers limited it at level 20. The hard cap is 255, the actual limit of the data structure used to store the skill level information.

The actual accuracy cap is 100%, due to the fact that you can never hit more than you swing (lol), but the soft cap (95%) is coded in by the programmers.

As an addendum, I was wrong with the minimum 5% evade, its actually 20%.

http://wiki.ffxiclopedia.org/Accuracy

All about it.
 
Guess we agree on the meaning but differ on the vocabulary. You see soft cap as meaning arbitrary and hard cap as meaning unavoidable (for example, a 20% to 95% accuracy soft cap vs. a 0% to 100% accuracy hard cap).

I see soft cap as meaning effectively capped but not an absolute stopping point - you can proceed further but with great difficulty. And hard cap as an absolute stopping point that you can't pass no matter what - whether that's due to plain old mathematics rendering percentages beyond 0% to 100% meaningless, due to computer hardware design limiting numbers to 0 to 255 (although that's an implementation choice as well, there's no rule that says they *have* to use only one byte, or even use an entire byte - see larger numbers taking up 2, 4, or more bytes, and also see bitwise coding), or whether that's just due to an arbitrary programming decision.

Either way, I think we agree on everything *but* the terminology, wouldn't you say?
 
Vilurum said:
Guess we agree on the meaning but differ on the vocabulary. You see soft cap as meaning arbitrary and hard cap as meaning unavoidable (for example, a 20% to 95% accuracy soft cap vs. a 0% to 100% accuracy hard cap).

I see soft cap as meaning effectively capped but not an absolute stopping point - you can proceed further but with great difficulty. And hard cap as an absolute stopping point that you can't pass no matter what - whether that's due to plain old mathematics rendering percentages beyond 0% to 100% meaningless, due to computer hardware design limiting numbers to 0 to 255 (although that's an implementation choice as well, there's no rule that says they *have* to use only one byte, or even use an entire byte - see larger numbers taking up 2, 4, or more bytes, and also see bitwise coding), or whether that's just due to an arbitrary programming decision.

Either way, I think we agree on everything *but* the terminology, wouldn't you say?
Click to expand...

Essentially.
 
You must log in or register to reply here.
Back
Top