Announcement

**Sevv** · 05-02-2007, 08:30 AM

Re: DA Vs. Aggressor Merits (Math/logic Heavy and long)

IMHO as a axe/joy user with a brutal

double attack > aggressor, I think the only time I have a need for aggerssor is skill up pts when im under leveled. I dont think the difference is greater then double attack. The only place I see aggressor greater then double attack is High eva based hnms/gods with Gaxe for sata steel cyclone.

**Ziero** · 05-02-2007, 10:27 AM

Re: DA Vs. Aggressor Merits (Math/logic Heavy and long)

I thought the multihit weapons completely overwrote Double Attack completely. Meaning if you have a Multihit weapon equiped, that hand will not double attack at all. Weren't there tests shown that Multihit weapons never go off during WS as opposed to DA? If that's the case then the Joyeuse shouldn't be considered in this calculation.

Also, iirc there is far more Acc+ gear then DA+ stuff. Since it's easier to stack acc gear and merit DA then to merit your Acc buff and stack DA+ one would assume you would be better off gearing Acc and meriting DA. Both suffer from diminishing returns so an overload of Acc will have the same effect as an 'overload' of DA.

Aggressor will *not* be up every time the timer resets, so your DoT numbers are slightly inflated towards the Aggressor side. In a merit pt, the downtime between pulls, if any, is a loss in DoT on the aggressor end as when switching mobs the time is running down and shortening the duration that Aggressor is actually in use. And in Alliance HNMs, if you in any way shape or form stop swinging while aggressor is active, it will again lower the actual DoT% increase you would see from it being fully merited.

Then there's the fact that kill speed, due to faster attacks, is usually more important in many endgame situations then pure acc. DA will 'increase' your attack 'speed' by giving you extra hits without delay, Aggressor will have no affect on said 'speed'. And though it may *seem* like that extra time with Aggressor may give you more dmg in an hour, you compared it to a solid 100 normal attacks counting the DAs inbetween while ignoring the amount of *time* it takes to make those attacks. The best way to compare them would be to either gauge the Acc% of 100 hits with Aggressor with the Dmg rate of 100 normal attacks plus the additional DA procs. Or even compare the Acc increase over the course of an hour with the amount of DA procs in the same timespan.

Though either way, I can not see any way that increased Aggressor will trump increased Double Attack in terms of dmg output. So while your numbers may not seem accurate to me, we do seem to have reached the same conclusion.

**Saren** · 05-02-2007, 12:07 PM

Re: DA Vs. Aggressor Merits (Math/logic Heavy and long)

Couple of questions.

Wiki is down while the owner moves it and boy do I miss it for posting on stuff like this.

1. Does Agressor allow you to exceed the accuracy cap of 95%?

2. Sorry about linking this post but I went and looked up aggressor and stumbled across a post on Alla (sorry again). The post is here and he talks about a few things but the relevant part was â€˜aggressor adds 25 to accuracy and ranged accuracyâ€™ he used the check method, seemed to know what he was talking about so I am guessing heâ€™s probably right. Annoyingly though he doesnâ€™t mention anything that would answer question one. Can anyone else confirm if he is correct?

The answers to those two really make a big difference to your maths and relative payoffs.

If aggressor doesnâ€™t allow you to exceed the 95% accuracy cap then it becomes a balance of being able to drop some accuracy gear/food for something that allows you to do more damage per hit compared to a DA merited WAR who is hitting a more often but for less damage and evading a bit more . If Agressor does allow you to exceed the accuracy cap then itâ€™s different.

If aggressor adds 25 straight points to your accuracy then itâ€™s not a percentage boost to your accuracy which changes how you would look at things. 25 accuracy points is approaching half of the amount of accuracy it takes to change a High evasion mob into a Low evasion mob which sounds like a lot of accuracy to me but I don't know melee jobs or high level mobs.

As with a lot of job specific merits really, I think it probably depends on what you do most of. Do you need to be able to evade much, do you tend to have more trouble landing hits or doing damage or some of both depending on what you are doing?

I think it also depends on your gear as you said yourself. If you can only have a max of 8 hits per attack round and itâ€™s not possible to proc DA more than once per hand per attack round (both unproven at present) then if you are dual wielding weapons that proc large numbers of multi hits regularly between them then your gear might well be stopping your DA merits from kicking in quite a lot of the time which is wasted merit points.

**Sevv** · 05-02-2007, 12:10 PM

Re: DA Vs. Aggressor Merits (Math/logic Heavy and long)

lol at 95% acc ive been parsed over that more then once on drg >.>

and with wiki down I cant show the difference in da vs aggressor ; ;

**Ziero** · 05-02-2007, 12:24 PM

Re: DA Vs. Aggressor Merits (Math/logic Heavy and long)

Originally posted by SevIfrit View Post

lol at 95% acc ive been parsed over that more then once on drg >.>
and with wiki down I cant show the difference in da vs aggressor ; ;

There is a 'soft' cap on Acc as you can't ever have 100% acc against any mob. So while you can have higher then 95%, that's usually the point where Acc gear stops having any real effect. And I don't think Aggressor is an exact +25 acc (like how Focus has been proven to be an exact +20 acc), I believe it scales up similar to Berserk. With focus, I often hear it become less reliable at higher lvls, but aggressor is always a strong Acc booster. Plus unlike Focus, Aggressor has a significant downside to being used so I'm sure it's bonuses will be far greater then five more acc and 60 more seconds of duration.

**Sevv** · 05-02-2007, 12:27 PM

Re: DA Vs. Aggressor Merits (Math/logic Heavy and long)

Im really thinking aggressor is about a +10/+15 its really like just wearing a life belt, it use swift belt for tp gain, on high eva mobs I use to switch to potent belt (used in ws macro) to help when aggressor was down.

**Jei** · 05-02-2007, 12:54 PM

Re: DA Vs. Aggressor Merits (Math/logic Heavy and long)

To me it's more about what weapon you use. DA is good if you use single hit weapons. Axe/axe, man/jug, something like that.

Berserk+Aggressor is best when you go for multi-hit weapons. Say, every war's dream, Ridil+joy. DA won't have much chance to proc, and ACC from aggressor can really help on top of supa+sword merit,

**Sevv** · 05-02-2007, 12:58 PM

Re: DA Vs. Aggressor Merits (Math/logic Heavy and long)

wasnt thinking of ridill-joy combo, I already wrote a ridill off when i got out of Kings Endgame, </3 afk claims

**Lmnop** · 05-04-2007, 05:39 AM

Re: DA Vs. Aggressor Merits (Math/logic Heavy and long)

aggressor really does seem to be a static +25 acc. Acc doesn't have to scale with level -- it's usually enough to make a mob go from neutral to low evasion.
It will not magically make you go above cap, and Saren, I think this addresses a concern you alluded to but maybe not:

Any good Warrior will have enough acc already that aggressor pushes you into diminishing returns mode. Even if it's enough to push you from high evasion to low evasion, you should have the mob checking low evasion -before- Aggressor (though you could always equip swap on more DD gear while aggressor's up - I tried this back in the 50s and found myself constantly forgetting to hit my macro to go back to acc gear after aggressor went down).

Focus is like 10-15 w/out AF. AF boosts it to 20.

You can obviously hit 3 times in a row and say you had 100% acc, but yes, soft cap over time will show closer to 95%.

As for his double attack rate with joy, it looks like he figured it. In his example of offhand joyeuse + brutal, that would normally be 15% 2A but he has it only giving 7.5 swings, which I believe is right. Because 2A is only proc'ing on half the swings you make (the 50% that joyeuse doesn't activate on).

I probably missed something, but I'm not quite understanding some of the earlier results...

If I have 15% 2A rate, it'll take 34% acc from Aggressor to compensate but if I have 20% 2A, it'll only take 32% acc from aggressor to match? Either there's an ambiguity I misread or that says that 5% additional double attack actually weakens you.

**Hankthetank** · 05-04-2007, 12:07 PM

Re: DA Vs. Aggressor Merits (Math/logic Heavy and long)

Thanks for reading my post. I apreciate the input and some eyes to check for errors.

If I have 15% 2A rate, it'll take 34% acc from Aggressor to compensate but if I have 20% 2A, it'll only take 32% acc from aggressor to match? Either there's an ambiguity I misread or that says that 5% additional double attack actually weakens you.

Its doesnâ€™t weaken you, it just offers a smaler improvement, meaning Aggressor has to do less to catch up. Because of the diminishing returns on stacking DA if you have 15% Double attack and merit another 5% it offers less improvement than if you only have 10% and merit another 5%. Since the resulting improvement is less significant, the amount of improvement aggressor needs to provide is also smaller.

Going from 15% to 20% offers a smaller 4.348% improvement in DoT so we only need a 32.608% improvement in acc between Aggressor up and down to match the increased DoT of the 4.348% increase in double attack.

Going from 10% to 15% offers a larger 4.545% improvement in DoT so you need a bit more, you need a 34.092% increase in acc between Aggressor up and down to match the increase DoT of the 4.545% increase in DoT.

I hope that makes sence. 5% double attack is always an improvement, its just a question of how much of an improvement it actually provides.

Sincerely,
Hankthetank

**Amele** · 05-04-2007, 01:39 PM

Re: DA Vs. Aggressor Merits (Math/logic Heavy and long)

I can't verify if this is the case since I don't own a ridill, but I've been told anecdotally that ridill and soboro sukehiro are functionally identical (excluding the obvious weapon type and delay) and I do own one of those.

in any case, it's an average attack rate of 2 per round (without double attack, /thf for my tests.) and it appears to single attack as frequently as triple attack.

I haven't yet determined the exact single/triple attack rates, I expect it is either:
33/33/33 or 25/50/25 (I'm leaning toward the latter at this time, but it's a pretty small sample size.) I'll see if I can get some people to waste a tag and go poke some excavation duty walls at 50 cap later this week.

could consider running the numbers in both cases probably.

**Jei** · 05-04-2007, 03:10 PM

Re: DA Vs. Aggressor Merits (Math/logic Heavy and long)

I don't own a Ridil either but I have seen tests for Kraken, saying that it evens out between 1-8 hits equally. Joyeuse seems to work the same way, 1-2 hits equally. Wouldn't be too surprising if Ridil uses the same rule, 1/1/1 for 1-3 hits.

Another reason I like to merit aggressor up equally to berserk is that I always use them both together in 1 macro. Having both equally merit and recharging at the same time is really nice and cut down one timing I have to keep track of.

**Saren** · 05-04-2007, 03:27 PM

Re: DA Vs. Aggressor Merits (Math/logic Heavy and long)

O.k. this post is totally lmnops fault. He told me to post maths so here goes. Remember who to blame when you get bored.

Before I start though, Sev and what you said about parsing. I am sure you did. The thing is though that if you knew you had 95% accuracy and you did say 10 batches of 100 hits and recorded how many times you missed, I would expect three or even four of your ten batches to have less than 5 misses in them giving a perceived accuracy of over 95%. 95% accuracy does NOT mean you can be pretty sure to connect 95 hits out of the first hundred hits you do or even 95 000 hits out of the first 100 000 you do (though this is more likely than connecting 95 in 100), this is a simplification to make it easy for people to use. What 95% accuracy means is that if that you picked a mob and hit it for ever then your average percentage of hits would be 95%. They are not the same thing.

Working out Damage per second

Hank, I think your maths is good as far as it goes but you were assuming that aggressor upped your accuracy by a certain percentage.

Before I post more, qualifying statements.

I want to make it clear that I don't expect any of the formulas I will put here or conclusions to be absolutely correct. I have used ffxilopedia as the source for all the information, I have assumed that everytthing said there is correct and I haven't tested anything myself. The chances of nothing I have quoted from wiki here being wrong are miniscule. I have also made some quite sweeping simplifications to stop my brain exploding.
There will be formulas in this post, donâ€™t treat them as gospel: they are at best approximations, and thats assuming the data and forumlas they are using are correct which they may well nto be.
I have ignored critical hits and all warrior job abilities barring aggressor.
At worst, this is me rearranging someone elses mistakes so please please donâ€™t take this too seriously. There is never any substitute for your own experience and testing.

Damage per second, the basic formula:

The stat that every DD wants to max, O.K now a approximation for a â€˜damage per secondâ€™ formula would be:

DPS = NDPS + (WSD/ average time to accumulate 100% TP)

Where
DPS = Damage per second
NDPS = the damage per second from ordinary non weapon skill attacking
WSD = average damage from whatever weapon skill you are using

I have split the damage in this way because for one, there are a lot of different weapon skills you could be doing, each with their own modifiers and gear swaps to max out the damage and many different weapons etc that you could be using. So, if you want to get an idea of the weaponskill side of the damage equation then I recommend looking here for information on TP accumulation and wiki is also a good site for look up the mods for various weapon skills. If you know that you are going to be doing a particular weapon skill most of the time, itâ€™s worth looking at the WSD but I will only be talking about maximising your NDPS.

NDPS

NDPS = average damage per hit * average hits per second * accuracy

Going through all the wiki stuff and simplifying down as much as I can these are what I came up with.

Approx damage per hit
Damage = { [weapons base damage * ((your STR â€“ mob VIT)/4)] * [your attack/mob defence]}

Where [(your STR â€“ mob VIT)/4] will never be less than (weapons base damage/9) and never more than (weapons base damage/9) + 8, the weapons base damage is the damage listed on the weapon and (your attack/mob defence) will never exceed 2. There were other corrections on this ratio but in general more attack the better till you get to your attack being double the mobs defence and the level corrections on this factor are irrelevant because we are talking in general terms so we donâ€™t know what level the mob will be. The only thing I find weird about this is that the level of skill you have with the weapon isnâ€™t listed as a damage factor which I was expecting it to be.

Conclusions:

So, if the above is right, your weapons base damage is by far the most powerful factor in determining damage per hit, followed by your strength (which is also limited by your weapons base damage) and finally your attack (attack being such a small factor surprised me).

Average Hits per second

This is where DA merits are going to boost your damage per second. Assuming that single weilding a weapon with 60 delay swings once per second and that DA can proc with equal chance on any base attack from your weapon then:

Avg hits per second = (avg number of hits per attack round) * 60/ (weapon delay)

If you are dual wielding you would need to put your combined delay in for â€˜weapon delayâ€™

average number of hits per attack round = (base average number of hits from the weapon) * [1 + (DA%/100)]

Conclusions:

Your DA dependant factor ranges from 1.1 up to 1.15 with merits and a +1% DA increases your NDPS by a factor of (0.6 * avg number of hits per attack round)/(weapon delay) which is pretty small again.
Taking into account possible limitations on DA proccing given by a cap on number of hits per attack round or DA not procing more than once per hand makes getting to an equations for average number of hits per second hellish so I left it as it was. I donâ€™t think any limitations would actually be statistically all that relevant, assuming they exist unless you have a large number of attacks per attack round occurring regularly, and by large I mean 2+ single wielding and 3+ dual wielding.

Accuracy:

This one is a pain in the butt and I am working based on some very assumption heavy logic so bear with me. This is the factor that aggressor can help you with.

Lets assume that accuracy is a linear progression, that is:

+X accuracy = +Y% hits landing on the mob.

Lets also assume that the very very tenuous figures put for the by wiki are correct, that is:

At the point of high evasion checkingâ€¦..your accuracy â€“ mobs evasion = -30 and your hit rate is 50%
At the point of low evasion checkingâ€¦..your accuracy â€“ mobs evasion = 10 and your hit rate is 75%

Assuming these are right and that accuracy is linear that means that +40 evasion gained us 25% hit rate. (this would make aggressor, if itâ€™s equal to +25 accuracy, equivalent to a +15.6% hit rate increase).

That this would also mean is that, assuming all the above is true, if you go up to a mob you want to fight, add accuracy gear till you find the point where the mob checks as low evasion, you would need an extra 32 points of accuracy from somewhere to get you up to the 95% accuracy â€˜capâ€™. You can see for yourself though, highly highly tenuous conclusion.

Conclusions:

So, aggressor: itâ€™s really as effective as the gear you have to swap in when you have that extra 25 accuracy that you donâ€™t have to find. You can see for yourself which factors are supposedly more helpful in terms of increasing your damage/hit. You would have to work out your NDPS with aggressor up and again with aggressor down and weighted average the two to see what effected the merits would have.

While we are here though, Wiki lists facrots that improve your accuracy as:

1 level of weapon skill = 1 accuracy (possibly dropping to 0.9 at over 200 skill)
2 DEX = 1 accuracy
+accuracy from gear or job traits or job abilities
+accuracy % from food â€“ these are applied after all the other bonuses and amplify them.

End Conclusions:

O.k. so after all of that, I have this advice.

If you are worrying about what group two merits to put in your WAR, get whichever ones you like and then make sure you have the best weapon you can possibly get because it probably has a much much greater effect on your DPS than your group 1 merits will, which seems to be what other people have said without needing all the gumph.

Sorry that wasnâ€™t more helpful Hank but the number of variables and most of them unqualified factors makes it hard to apply maths to this problem at all.

**Amele** · 05-07-2007, 07:04 AM

Re: DA Vs. Aggressor Merits (Math/logic Heavy and long)

good numbers,

minor issue.

accuracy gains are diminishing over time, starting at the 75-80% mark. (this is to say) +40 accuracy may gain you 25% going from High Evasion to Low Evasion, but will only gain you 5% going from (whatever number) accuracy (85%) to (whatever number+40) accuracy (90%).

this is part of why hank used accuracy gain %'s. - the other big reason is that it's generally accepted that Aggressor scales with level, which implies that aggressor's specific mod (unlike, say, Hasso) *is* a percentage of your accuracy.

(edit: )
also jei, speaking from experience with a kclub, the number of hits was more bell-shaped than linear, (i.e. I hit 3,4,5 times much more frequently than I hit 1 or 8) and since double attack can only proc on 1 hit, it's unfortunately not mathematically sound to just approximate all multihit weapons as 1/1(/1../1etc)
in the case of my soboro, I've always just sort of assumed that rigorous testing will bear out the kclub results I've seen, (that is: x%/100-2x%/x%) probably as: 25/50/25 and I would likely assume the same with ridill, until someone who owns one comes in and can give some definite results

Announcement

DA Vs. Aggressor Merits (Math/logic Heavy and long)

DA Vs. Aggressor Merits (Math/logic Heavy and long)

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment