Summoners, cap pet accuracy or no?

[Ryuko]

I've never seen Garuda miss Contagion. I also believe in another thread about DoTs missing someone posted that "something that hits without an initial damage output cannot miss", stating that Bio & Bio II were examples of this. Contagion doesn't do damage, so I assume it falls under the same properties. It would be nice if a dev commented to confirm / deny this.

I would like some sort of real dev answer on this too. It is true that Miasma seems to miss a lot more (if your acc isn't capped), but Thunder doesn't seem to miss quite as often. I have never seen Bio miss, but I HAVE seen Bio II miss... although I'm not sure if it ACTUALLY missed or the game was simply being wonky and it showed my animation going off but the spell was never actually cast. If I had been parsing, perhaps I would have seen it and known for sure. I haven't done any real testing, but I can only assume that some abilities require more accuracy than others. Otherwise, why in the world would Miasma miss so much but Thunder doesn't?

As far as the pet... if your theory is correct, Contagion wouldn't miss -- but Enkindle would. Again, you can't SEE the numbers pop up like yours would because your pet isn't you, but you could check the parser to know if it hit or not and for exactly how much. Enkindle is a little underwhelming, but it is helpful. I would hate for it to miss.

[Yukagama]

I cap on acc and crit but eather way love to see some offical source regarding to acc on pet ect.

[T0rin]

I cap on acc and crit but eather way love to see some offical source regarding to acc on pet ect.

You don't need an 'official' source. Download a parser (logrep2, ffxiv-app) and go to Turn 5. Have your pet attack Twintania, watch it miss.

SE isn't going to come out and say 'here are the official accuracy caps for all the content', this is left up to the players to test and determine. The players have pegged the SMN accuracy cap at 432-435 and the pet accuracy cap at ~450.

If you take those values as truth, it is not hard to determine how much a 5% loss of accuracy on a pet affects your overall DPS potential.

[Fire-Dragon-DoL]

I think you mean 432-435 and ~450 :P

[T0rin]

I think you mean 432-435 and ~450 :P

Heh, yes. Fixing that now. :P

[xxalucard]

If you take those values as truth, it is not hard to determine how much a 5% loss of accuracy on a pet affects your overall DPS potential.

Key words there: "If you take these numbers as truth."

When it comes to the supposed "DPS loss of 5% pet accuracy" numbers, I don't.

I have no idea what Kevee is basing his math on besides his own theories, especially when there are many factors no parser can even track (such as DoT crits, shadow flare crits, bane'd dots and their crits). There is no battle log evidence to analyze those example crits either, so he (as well as the rest of us) have no idea how many times more our DoTs are critting and how much it's adding up to in a long fight like Turn 5. I would think the longer the fight, the more DPS the extra crit adds, especially for those dots, and since those dots are going to be on MULTIPLE enemies SIMULTANEOUSLY (whereas garuda can only use wind blade on one enemy at a time) those crits will add to your overall DPS that much more.

I really wish I could see for myself how much damage that is. I know it's definitely a lot more than Garuda's single-target wind blade will ever produce with 5% more accuracy, but I do wish I knew how much more.

If just one more of my festers crit on a conflag, that makes up for at least 1 of garuda's 3 attacks missing (since he only has time to land 3 on a conflag even if you instantly target it with obey-- assuming your group has good dps), makes much more of a difference in bursting it, and will allow me to focus garuda back on twin faster. Depending on who's inside the conflag it can sometimes make the difference between firestorm going off or not. Using enkindle means you aren't attacking, and is actually worse DPS in that situation when you and garuda can attack simultaneously (just hitting fester + energy drain + ruin II as garuda attacks is better burst in that situation).

Conflags: Bane dot tic crits, fester crits, energy drain crits (on the fast conflags), ruin/ ruin II crits. Shadow flare crits (for when you're the one inside conflag) All much more important burst.

Snakes: The add snakes die within a handful of wind blades so it's better to keep garuda mostly on the boss snake. While he's on that snake, him missing 5% more is not going to make a difference in killing it. Once it gets 4 stacks of the debuff and your group is going nuts on DPS, it should die with plenty of time to spare. The only time it wouldn't is if you didn't have level 3 LB ready for it or your group messed up and lost the debuff stacks (before it hit 4). In both cases it is unlikely that the few wind-blades Garuda potentially misses because you have 5% less pet accuracy is the cause for failing.

Dreadknights: Your group might burst it faster than Garuda can even get 3 wind blades off on it, making his dps on them simply a bonus, not a necessity, and not an important factor whatsoever in bursting it (compared to, say, a fester crit or bane dot crits off twin / shadowflare crits). If you get chosen as the one who's "stunned" for that dreadnight, then you can't even put garuda on it. Your shadowflare, however, will still be hitting it, and if you bane the moment it appears you can infect him before your stunned.



I see a lot of theories without experience, parsers that can't accurately track the majority of our DPS (being referred to as reliable evidence), and exact number percentages that have no visible formula which factor in all of our different attacks.

I rather stick to reality and what I see working better or not working better... that's just me.

[pandabearcat]

So while I agree with you in theory, you have a bunch of mathematical fallacies in your argument, and I wanted to go over them to make sure you know why people might not agree with you.

You're basically making the argument that "you can see crits, and they're big, and you can't really tell the difference accuracy makes".

Which is true, but doesn't make sense in this argument.

We're not...like talking about losing all crit to increase accuracy.

Here is a delta of like 40 or so stat points, right?

What people are saying is, 14 or 15 stat points is 1% crit gained or lost, and like i dunno, some amount of accuracy (Kevee says 4%) for your pet. This is based on the formulas on the valk site, which I can't remember the link for but is generally accepted as true.

Now the difference is between 1% crit and 4% pet accuracy.

General consensus from modeling and parsing is that pets are between 25-30% of your damage on a single target, that factors in crits, that factors in buffs, that does NOT factor in multidotting.

The more you multidot the less relative damage your pet does (with its own stats, i'm not talking about contagion here).

You also mentioned fester critting...which is irrelevant here because 1% higher chance of fester critting, you might as well say, but oh wait, what if my garuda misses twice during that time, because she has a 4% higher chance of missing. This setup for a scenario doesn't help prove the point one way or another.

You bring up crits again on the dread knight...again this is a fallacy. The chances of you getting one more crit are going to be similar to the chances of garuda missing twice or three times on that enemy...i mean you are really, really overselling the 1% crit and equating it to your entire crit chance, and also as if you could control your chances of critting.

I don't understand your reality of working better or not working better.

Did you try it with 1% crit and with 4% more accuracy? Did you test it hundreds of times? How else can you be statistically certain that any of trials weren't flukes one way or another?

Crit is the gamer's fallacy...in fact if you lost the ability to crit entirely, you still would be able to down these bosses.

That being said one point you made was good: what about multidotting and bane. This is a good question, as without a doubt multidotting does decrease the relative value of pet accuracy. You should look into and test this, but without this idea that somehow 1% crit is going to make or break you.

[Kevee]

snip

Not to mention he/she is still completely disregarding the DET gain in addition the accuracy.



34 Crit (13.8 crit=1%. A little higher than my finding of 13.5, but they had a better sample) = 2.46% crit = 1.23% DPS.

70 Pet DPS(Which is, again, low) x .98 accuracy = 68.6 DPS for 445 Accuracy.
70 Pet DPS x .93 accuracy = 65.1 for 435 Accuracy.
I do apologize, I don't know where I got the 4.2 before, but the difference is 3.5 DPS for the accuracy alone. However, the better your gear and using the more realistic pet DPS, the bigger the gap and more value ACC gets. A more realistic DPS value is ~83-85 DPS. At 83 DPS, it's a difference of 4.15 DPS regarding the accuracy. Remember, crit is static. Accuracy is not. If you dip down to 415 Accuracy and use food to go to 435, you're in trouble. The pet has ~70-75% accuracy then.

For the 1.23% DPS to equal the 3.5 DPS difference from accuracy, you would need to do 284 DPS as the summoner itself(3.5/.0123). For the more realistic value of 4.15 DPS, the summoner itself would have to do 337 DPS to equal it.

While that is more realistic, this is not factoring in the DET gain, and is still too high. This poster is simply on the crit hype-train and thinks the pet is worthless, where it is actually our strongest "DoT."

Not even factoring in the burst DPS of the pet during Rouse/Spur, or the use of Enkindle, where misses have a lot more weight.

And, again, this is not even factoring in the DET gain.

I was trying to say that it's better to invest in even 1% crit over 5% pet accuracy, especially when your own accuracy is capped.

No, it isn't. Especially not 1% crit.

Mostly I noticed that Garuda's wind blade was not helping much in burst situations, and that in special turn 5 scenarios like the snake phase with the 4 stacks of debuffs, the crits you get can be HUGE (i once crit a fester for ~3527 on the big snake in a failed attempt, and my gear is not that great).

I'm not trying to say you won't crit with more accuracy. I'm saying that the potential of even a small amount of extra crits is more worth it than worrying about 5% of garuda's accuracy for the majority (if not all) the situations in turn 5.

There's a reason I have my AF2 chest, but still use Vanya(until I get my AF2 boots or Allagan Boots).

The loss of Accuracy from 442 to 415 is way too much for the pet, and it's consistently been a DPS loss.

Wind blade "not helping much"? Are you joking? It's more than twice as strong as Bio, which is our strongest DoT, and that's not factoring in Rouse/Spur/Enkindle for burst scenarios, where it is more than 4 times as strong, per wind blade, than a Bio tick.

I'm not trying to say you won't crit with more accuracy. I'm saying that the potential of even a small amount of extra crits is more worth it than worrying about 5% of garuda's accuracy for the majority (if not all) the situations in turn 5.

And, again, no.

Besides Fester, Garuda is our strongest attack for all of turn 5. The burst potential, and flat out DPS. If Garuda misses once on a Conflag/Dreadknight, a Bio would have to crit ~4 times to make up for it. You can't make up for that on a short timeframe like Conflag/Dreadknight.

[xxalucard]

I agree that I was making it sound like all crits vs no crits on a 1% difference and that's definitely far, far from the truth. I'm not trying to say that if you get more accuracy you won't crit the same (or even more if you're lucky with RNG).

I was trying to say that it's better to invest in even 1% crit over 5% pet accuracy, especially when your own accuracy is capped.

The point of the examples was more to say that the risk/reward of garuda missing 5% more vs gaining even 1% more crits from everything is more beneficial for turn 5, the turn most people are focused on with the pet accuracy cap.

I actually did run turn 5 at least a hundred times before we beat it. I know that's not high enough for the most accurate numbers, but I believe it's enough times to gauge the effectiveness of certain stat changes-- and I did so in different pieces of gear (around 30+ times in a vanya robe before I got the Summoner's Doublet). Of course the summoner's doublet is better and of course those aren't the specific items in question here, but I believe they are similar enough for me to say that I noticed as you gain in INT your crits get much more value, and when reaching the BiS scenario with the maximum INT possible, you'd get the most bang for your buck out of crit.

Mostly I noticed that Garuda's wind blade was not helping much in burst situations, and that in special turn 5 scenarios like the snake phase with the 4 stacks of debuffs, the crits you get can be HUGE (i once crit a fester for ~3527 on the big snake in a failed attempt, and my gear is not that great).

I'm not trying to say you won't crit with more accuracy. I'm saying that the potential of even a small amount of extra crits is more worth it than worrying about 5% of garuda's accuracy for the majority (if not all) the situations in turn 5.

[FUNgineer]

All this anecdotal evidence :V

if 1% crit is the difference in consistently downing conflags and dreads or wiping then you should reconsider aetherflow management as opposed to worrying about pet missing or not.

that being said, I don't think I'd notice the difference in 1% crit because I haven't done twintania 100 times yet :V