WAR MT + PLD OT vs. PLD MT + WAR OT in terms of DPS

[The_Badger]

Your parses don't seem to reflect the damage a PLD will gain with the SE debuff, that's 11% damage missing (something like 11.5% if you want to get anal, iirc), which is also stated in the first post linked. That would set the parses at nearly equal, but I also think that SE > BB combos is higher damage over all because you lose SE debuff on a full combo going SE > BB > BB. I'll also add that your data points seem to be fairly small, RNG could have easily skewed the data and we don't know if you cherry picked the best and worse.

SE wasn't used for either WAR or PLD in the second set, which is more accurate as both are using i90 weapons as opposed to the first set where PLD had i95 vs WAR i90, so the point is moot, as they'd both gain from adding SE in the mix. The first set of tests, where SE was used, the PLD has a weapon advantage and Circle of Scorn was hitting two dummies due to poor positioning, giving even more of an advantage yet WAR still won every test. The second set of tests are more accurate as both have identical gear and both using i90 weapons and CoS only hit one dummy. I encourage you to read the posts and methodology instead of just glancing at the numbers.

I've parsed it many times and posted quite a decent sample size. I have more results than are posted, but it's just not necessary to post more as the results are very consistent. If you doubt my findings run the same tests yourself and post results to compare. The results are not RNG and I have no bias toward either class as I play and enjoy both equally. So no, these aren't cherry picked and the tests are easy to replicate. Feel free to dispute my findings but back it up with actual numbers or tests instead of just guessing or hypothesizing. This is the problem with this topic, almost no one posts any actual results and just guess this or that and claim this and that.

Please see THIS post for updated info.

[Skull_Angel]

I've parsed it many times and posted quite a decent sample size. I have more results than are posted, but it's just not necessary to post more as the results are very consistent. If you doubt my findings run the same tests yourself and post results to compare. The results are not RNG and I have no bias toward either class as I play and enjoy both equally. So no, these aren't cherry picked and the tests are easy to replicate. Feel free to dispute my findings but back it up with actual numbers or tests instead of just guessing or hypothesizing. This is the problem with this topic, almost no one posts any actual results and just guess this or that and claim this and that.

The user base has no clue how you garnered your results from the data that you show; at this point you're not really arguing against any person, your arguing against math that's been shown to be close enough to RL results to come to the current conclusion (neither job has a distinct advantage in DPS when both are standing on equal grounds in their damage stance).

Your opening statement ["To keep this short and sweet; PLD will NOT out DPS a WAR in equal gear, ever."] is just silly in that it implies that WAR has a sizable advantage over PLD; this contests what math shows. The sample size you present is no where near decent and all we have to go off of is your word that your other parses follow suit.

You may have the wrong idea about how theorycrafting works if you are asking people to contest incomplete data that is supposed to go against what's been shown to be mostly accurate. You have to compile the data that shows otherwise with full confidence, incomplete data and word of mouth won't cut it for most people as well as telling others to spend their time to replicate your tests.

[Kitru]

This is likely due to 3 inherent facts -

And you're forgetting a few others:

1.) Brutal Swing is 100 potency every 8 GCDs, whereas SW is 300 every 12 and CoS is 250 every 10.
2.) SwO Oath is an extra ~55 potency per GCD.
3.) Fight or Flight (30% +dam for 30 secs every 90 secs; 10% +dam averaged) is significantly stronger than Berserk (~40% +dam for 20 seconds every 90 secs w/ a ~71% loss of DPS for 5 seconds every 90 secs).

The 20 extra potency per combo (6.67 pot/GCD) is easily outstripped by SwO (~55 pot/GCD) or *either* of the off-GCDs (25 pot/GCD each).

If you want to go on about all of the advantages that WAR has, you should factor in all of the advantages that PLD has that everyone seems to forget about, especially since, except for Maim, the WAR advantages pale in comparison to the PLD advantages.

I just did a test myself.

Using identical gear (only difference was PLD w/ Curtana Animus and TW Shield v. TW Axe; PLD had 20 less crit and 21 less DET; 10 STR natural disparity preserved), over a 90 sec interval (1 Berserk/FoF) in SwO, using off-GCDs (CoS was only hitting one target because, seriously, it's not that hard to attack from the side) and RoH combo spam, PLD managed 23056 damage (256.18 DPS). Over the same interval, outside of Defiance and using off-GCDs and SP>BB (maintain Maim without getting SE debuff), WAR managed 22563 (250.7 DPS).

Every single time I've tested it, the numbers turn out exactly the same: PLD in SwO just *barely* edges out WAR outside of Defiance, and I've been testing it since 2.1. The fact that I have theory backing up my numbers and yours are only backed up by supposition that the theory is completely and totally wrong (not to mention that your numbers would be mediocre output even back when the gear you're talking about was actually good) leads me to place a helluva lot more credence on my own.

[The_Badger]

The user base has no clue how you garnered your results from the data that you show;

Sure you do, read the posts.

The sample size you present is no where near decent and all we have to go off of is your word that your other parses follow suit.

I encourage everyone to challenge my findings, but do so by posting actual results and numbers from identical tests; 3 minute parses in identical gear on a Training Dummy. Or post your own methodology and I'll test it. But please stop just saying you don't believe this or that, or that the math doesn't add up. Post your own tests. Otherwise you're just contributing to the confusion and not helping the cause.

You may have the wrong idea about how theorycrafting works if you are asking people to contest incomplete data

Yet you challenge data with no data?

Look, I don't have a dog in this fight so I don't care whether WAR or PLD are the better damage dealer of the two tanks. I'm simply posting information pertaining to the situation.

Using identical gear (only difference was PLD w/ Curtana Animus and TW Shield v. TW Axe; PLD had 20 less crit and 21 less DET; 10 STR natural disparity preserved), over a 90 sec interval.

A 90 second test is really weak, as this is the "burst phase" of anyone's damage, and as PLD has more frequent OGCD's it wouldn't surprise me that PLD can pull ahead here. Please do a longer test. I mean a 90 second test is just not enough. It's like claiming that BRD does 500+ DPS because they can burst over that for the first 60-90 seconds of the fight. You have to test long enough for the numbers to normalize, and 3 minutes is a good marker in my opinion as that's when both PLD and WAR will run out of TP with no assistance. Actually they run out around 2:50 so there is a slight delay in the last 3 attacks.

Every single time I've tested it, the numbers turn out exactly the same: PLD in SwO just *barely* edges out WAR outside of Defiance, and I've been testing it since 2.1. The fact that I have theory backing up my numbers and yours are only backed up by supposition that the theory is completely and totally wrong (not to mention that your numbers would be mediocre output even back when the gear you're talking about was actually good) leads me to place a helluva lot more credence on my own.

Then please post data and stop with the hypothesizing. This happens in every thread about this; people claim this or that but never provide any actual testing to back it up. You expect me to believe your numbers and theories but you won't even trust actual parses that I put up? We're not working against each other here, as you seem to believe. Also, no reason for the condescending tone of your post, let's try and be civil here.

Please see THIS post for updated info.

[Kyros]

A 90 second test is really weak, as this is the "burst phase" of anyone's damage, and as PLD has more frequent OGCD's it wouldn't surprise me that PLD can pull ahead here. Please do a longer test. I mean a 90 second test is just not enough. It's like claiming that BRD does 500+ DPS because they can burst over that for the first 60-90 seconds of the fight. You have to test long enough for the numbers to normalize, and 3 minutes is a good marker in my opinion as that's when both PLD and WAR will run out of TP with no assistance. Actually they run out around 2:50 so there is a slight delay in the last 3 attacks.

Actually WAR has more burst than PLD because of how berserk works. Short parses help WAR more immensely than longer ones.

Regardless, I just went ahead and parsed my PLD vs my WAR and here are the results:

Methodology:

- Used Identical i94 sets (http://ffxiv.ariyala.com/MYK2) with i90 STR accesories. Only difference being Curtana/Shield Zenith vs Dual Haken in the weapon slot (Ideally I'd have used allagan shield, but I don't have it).
- Total stat difference was the 10 native STR for WAR and 7 SS on PLD from shield.
- Total time for tests was 2:30. I was having TP problems at 3:00 and 2:45 marks so I set them lower to 2:30.
- Repeated tests until I got the cleanest parse I could get.
- PLD had SwO up and used the Halone > Repeat rotation with FoF, SW, and CoS used as soon as they where up with no afterthought.
- WAR had Defiance Off, Used BS/Berserk every time it was up and no fracture was used. Tested SP > BB > BB > Repeat and SE > SP > BB > Repeat rotation.
- Tests where done on a lvl 50 Dummy from the side, making sure not to hit the adjacent dummy with CoS.

Cleanest (And highest) Parse for each was:

PLD: 236.62
WAR (SP>BB>BB): 218.40
WAR (SE>SP>BB): 226.15

PLD consistently did more DPS vs both setups. I'm kind of baffled that the SE combo didn't get that much of a boost compared to the SP combo, so there's that. Probably would get more damage off fracture combo'd in.

So there is my own personal data. It seems to line up with the math.

[Skull_Angel]

Sure you do, read the posts.

I've read them, there's no parse log only a listed end result. My point is that you need to post more conclusive data if you intend to contest what is accepted to be true.

I encourage everyone to challenge my findings, but do so by posting actual results and numbers from identical tests; 3 minute parses in identical gear on a Training Dummy. Or post your own methodology and I'll test it. But please stop just saying you don't believe this or that, or that the math doesn't add up. Post your own tests. Otherwise you're just contributing to the confusion and not helping the cause.

There is no need to challenge your data, as it would be a waste of time for most at this point because of your lack of data.

Yet you challenge data with no data?

It's unnecessary to contest your data when the math that's been presented and known to be mostly accurate lines up with the parse information of many others [and myself]. Again, it's up to you to present more information if it goes against what's been widely accepted.

A 90 second test is really weak, as this is the "burst phase" of anyone's damage, and as PLD has more frequent OGCD's it wouldn't surprise me that PLD can pull ahead here. Please do a longer test. I mean a 90 second test is just not enough. It's like claiming that BRD does 500+ DPS because they can burst over that for the first 60-90 seconds of the fight. You have to test long enough for the numbers to normalize, and 3 minutes is a good marker in my opinion as that's when both PLD and WAR will run out of TP with no assistance. Actually they run out around 2:50 so there is a slight delay in the last 3 attacks.

A 90 sec. test lines up with the main damage CD of each WAR (Berserk) and PLD (Fight or Flight), either way, one data point is inconclusive due to RNG and ~10,000 points are necessary to bring the level of confidence to within 1%, iirc. But, there are limitations to the 90 sec. test as well, yes; it can not account for other OGCD abilities that boost DPS properly, more accurate results would most likely come from much longer parses in which a brd would be required (is it possible to spam TP abilities for ~15 min. even with a brd?).

Then please post data and stop with the hypothesizing. This happens in every thread about this; people claim this or that but never provide any actual testing to back it up. You expect me to believe your numbers and theories but you won't even trust actual parses that I put up? We're not working against each other here, as you seem to believe. Also, no reason for the condescending tone of your post, let's try and be civil here.

This happens in every thread because the lack of data that's being used to go against math that's been tested by multiple members of the theorycrafting community as being accurate enough for real world use.

[The_Badger]

Snip...

Thanks for running some tests instead of just coming back with "but math and theory!". I do appreciate it.

Interesting results you have and I can't seem to replicate it at all. I just ran 2:30 tests on both PLD and WAR in nearly identical gear, had to swap two accessories to STR for PLD to make up for the 30 points I have in STR on my WAR as I don't have 10k seals atm, and WAR still comes out ahead every single time. My lowest WAR test is still higher than my highest PLD test.

I find this odd that our results would vary so wildly. So I made a video and took a ton of screenshots. I've also uploaded the .act file of my parse. This should show exactly how I'm testing and allow for critique. If I'm doing something inherently wrong on PLD please let me know, as again PLD falls short of WAR.

HERE is a link to Dropbox which has the parses, labeled "+New PLD" and "+New WAR". There's some other tests from before in that folder as well if you want to view those.

I'm also rendering the video while I sleep and will upload it tomorrow, as it's an hour long. I won't be able to upload a video of that length to Youtube but I'll upload the file to be watched in something like VLC so you can speed through it. I'm simply uploading it to show my exact methods of replicating a 2:30 test.

Please see THIS post for updated info.

[Fue]

snip

Well there's your problem. that parser doesnt count SwO.

[Gildarts]

Well there's your problem. that parser doesnt count SwO.

and how you know it doesnt count it?

[Fue]

and how you know it doesnt count it?

Check the numbers its not listed anywhere.
The other one does calculate it though reason why theres such a big difference between the two.