Dragoons: A Rotation Reborn

[Martyn]

Hey Dervy, when you start the new DRG Thread could you please just put in some DPS numbers. Something like approximately expected DPS at Ilvl 110, 115, 120, etc. Maybe the thread finally gets rid of "is 650dps any good? " and "how much dps should I do ".

About that pentamelded stuff I have to disagree with Dervy actually.
Its like the same thing with Zodiac weapons. If you dont have the money to maximaze it there might be way to get a decent compromise.
Take the ring for example. You could get +3 crit and +3 acc if you put in lvl IV materias in slot 4&5 instead of lvl III. Each cost about 500k - 600k compared to 6k - 15k (on my server). I dont even want to guess how much gil you gotta spent for those 3 crit and 3 acc.
In my opinion its better to use lvl III as soon as you reach 11% chance, but that depends on your wallet and your will to spent gil of course. I also dont want to use flint caviar. Its about 50k here which makes it not worth it in my opinion.

I am kind of greedy about stats too, but as soon as I think about spending like 6million gil for 3 more crit which only lasts untill the add on... nah, I wont do that.

Anyways, for those who dont want to spent that much money or simply dont have it, here is one option to use pentamelded stuff, get a decent set and dont waste too much money. It follows the rule of not using lvl IV materia in <12% slots and not using flint caviar. it shows the comparison to the non crafted bis set.

http://ffxiv.ariyala.com/PM4Y

[SunnyHirose]

Hey Dervy, when you start the new DRG Thread could you please just put in some DPS numbers. Something like approximately expected DPS at Ilvl 110, 115, 120, etc. Maybe the thread finally gets rid of "is 650dps any good? " and "how much dps should I do ".

I don't think that will help. One reason is because item level is a deceptive number. Not all slots are made equal, for one: jewelry counts the same as a chest piece, while providing less than half the stats. And there is a significant difference in DPS between i110 with a Liberator versus i110 with a High Allagan Spear. Another reason is it doesn't translate to fight performance all that well. Even an innocent question like "what DPS should I do in T10" is going to be confounded by factors like "do you tank it on top of the puddle?" and "how many times do you get stunned for Wild Charge?".

If you want to know how much DPS you should do on a dummy, try it in-game, because that's 100% accurate to the game and it's not unrealistic to expect a near-perfect performance on a dummy (just keep trying). If you are trying to test a gearset you don't have, well, it takes a simulator to get in the right ballpark (they can't be super accurate until we have a super accurate damage formula, which we currently don't). Unfortunately, the sims that are available have some significant problems, sometimes with accuracy, sometimes with user-friendliness.

I think http://sim.ffxivguild.net/ would be okay if (and only if) several things about it got changed, some trivial some not.

[siverstorm]

Snip

I think http://sim.ffxivguild.net/ would be okay if (and only if) several things about it got changed, some trivial some not.

Yeah the differences between what the sims parse and what I actually parse with my gear is pretty drastic (in some sims it's like 40 points or more under what I regularly do unless the seeds are factoring random mechanics, im not sure), while I love what it is and what it does immensely it's completely unreliable from a drg perspective. I remember it being somewhat close last tier (still off by a large margin, but much more reliable).

...after doing some research before making this post I saw an issue on his github about the changes to some classes from since 2.4 and I think that's the reason why it falls short by so much now.

[SunnyHirose]

Yeah the differences between what the sims parse and what I actually parse with my gear is pretty drastic (in some sims it's like 40 points or more under what I regularly do unless the seeds are factoring random mechanics, im not sure), while I love what it is and what it does immensely it's completely unreliable from a drg perspective. I remember it being somewhat close last tier (still off by a large margin, but much more reliable).

...after doing some research before making this post I saw an issue on his github about the changes to some classes from since 2.4 and I think that's the reason why it falls short by so much now.

There are a few bugs/updates with DRG that I could fix right off the bat, it's just I'm so bad at using git properly. The copy I use personally has a few duct-tape-tier fixes, and certain results seem reasonable (can't tell if they're good, mind) when I use the "run 100000s of sims" option--this takes my personal machine a couple of minutes to run, so it's much too computationally prohibitive to make available on the web version.

When the damage formula is changed, it produces stat weights for WD/STR/DET/CRIT that sound vaguely plausible. The skill speed values must be regarded as pure balonium until the simulation is more properly sensitive to small changes in it.

[LunaHoshino]

Yeah the differences between what the sims parse and what I actually parse with my gear is pretty drastic (in some sims it's like 40 points or more under what I regularly do unless the seeds are factoring random mechanics, im not sure), while I love what it is and what it does immensely it's completely unreliable from a drg perspective. I remember it being somewhat close last tier (still off by a large margin, but much more reliable).

...after doing some research before making this post I saw an issue on his github about the changes to some classes from since 2.4 and I think that's the reason why it falls short by so much now.

Yeah, I was asking because I'd tried that website and it told me I should be doing 445, which seemed way too low considering I was doing 485 in T8 when I was about 7 ilvls below what I am now. Good to know that it is in fact inaccurate where DRGs are concerned!