T/H/D/D vs T/3DPS - By the numbers

[Renathras]

EDIT: Looked into this a bit more. Apparently ForbiddenWebsite has a weird glitch where, if two party members in a 4 man are the same Job (e.g. two BLMs) in a 3DPS party it will...randomly substitute one with a random Healer Job. If there's a rhyme or reason to it, I haven't see it. Regardless, the top compositions for speed runs do appear to be T/3D parties, with the traditional T/H/D/D party being approximately 13.5-22.5% slower, depending on dungeon. This doesn't change much since the deviation is relatively small (usually less than 2 minutes; for 5 Extreme runs in a week, this would save a total of around 10 minutes per week) and the speed running community for 4 mans is also very niche (meaning not a thing that the community as a whole cares about nor that the Devs balance around). Moreover, these are the top/best speed run results in the entire data set, not an average/what most players would experience attempting the same thing. But putting it here for the sake of completeness/awareness. Whether the difference is noticeable or worthwhile to you probably depends on what your reliable dungeon running group consists of and how much effort or coordination you do or don't wish to use. :ENDEDIT

.

While pre-made speed runs don't matter to the majority of the community and very likely don't matter to the Devs, the claim is often made that the Tank/3DPS party is outright superior for clearing 4 man content at the fastest rate. So, what do the numbers say?


Overall: (done by rating, not by time, though I think it's a proxy for time)

#1 is WAR/NIN/SAM/DNC - 710.74

#2 is WAR/SCH/BLM/SAM - 701.38 (a difference of -1.3169...% from #1, so less than 1.5% difference)


Dead Ends:

#1 is PLD/SCH/BRD/BLM - 10:48

#2 is PLD/SCH/BLM/DRG - 11:09

#3 is GNB/AST/NIN/BLM - 11:12

#4 is first no-healer party WAR/MCH/DRG/BLM - 11:27 (39 sec longer than #1)


Smiletown:

#1 is GNB/AST/NIN/BLM - 8:32

#2 is WAR/MCH/DRG/BLM - 9:22 (50 sec longer)


Stigma Dreamscape:

#1 is WAR/MCH/DRG/BLM - 9:07

#2 is WAR/WHM/NIN/BLM - 9:11 (4 sec slower)


Alzadaal's Legacy:

#1 is WAR/MNK/DRG/BLM - 9:13

#2 is GNB/AST/NIN/BLM - 9:17 (4 sec slower)


Fell Court of Troia:

#1 is GNB/AST/NIN/BLM - 9:30

#2 is WAR/WHM/NIN/BLM - 9:38

#3 is PLD/BLM/RPR/DNC - 10:57 (87 sec or 1 min 27 sec slower)


Lapis Manalis:

#1 is PLD/BLM/RPR/DNC - 10:55

#2 is WAR/WHM/NIN/BLM - 10:57 (3 sec slower)


In many cases, the standard party is the fastest clear of these 4 mans. In the cases it ISN'T, the Tank/3DPS clears are literally only a few seconds (3 or 4) faster.


This indicates that, in a coordinated party speedrunning content AND reporting on it (the forbidden website), there seems to be no real advantage to running T/3DPS over the standard T/H/D/D comp. And in several cases, the T/3D comp is actually slower. In one case by more than a minute.


Moreover, looking at Lapis Manalis specifically (since we can't "overgear" the content yet like we can the lower ones), the difference between the #1 and #50 party is 10:55 (655 sec) vs 12:01 (721 sec), a difference of 1 minute and 6 seconds. So we're not exactly talking massive differences here.


While it is absolutely true that a lot of the fast times ARE the T/3D comps, the fact remains that they are not the FASTEST in all cases, and where they are, are only so by a smidge. This means, when well played by a coordinated group, the time gap is relatively insignificant. And in some cases, the fastest clears ARE with the standard T/H/D/D, meaning for some dungeons, T/H/D/D is faster than T/3D.

[Curisu]

You compare completely different group compositions and player's.
By pure number's a T/3D group will be better, as long the Tank is a WAR.
Because a WAR does not care about heals. He does WtW pulls ajd can keep himself and others alive without wasting GCDs for like clemency.

What are your groups made of anyway?
Is everyone with BiS gear and on their main role?

The few times I tanked for 3Dps where by far the fastest dungeon clears I ever had.
Because the DPS had BiS gear and knew how to play. They did not get hit by anything and survived.
While the random standard DF group always has at least one player who is not that good.
Healer's who want to GCD heal a WAR, DPS who get hit by stuff and needs to be healed and player who just don't know how to play their class.

[ForsakenRoe]

I don't really remember that many people saying it's faster, but I do remember a lot of people saying it's pretty sad that it's possible in the first place

[WaxSw]

What you're looking is wrong data, while in the ranking they seem to have healers when you open the log of the run the roles change, a quick example:

Remember the names and the run time, then we click, open the log and...

War, Nin, blm blm no Sch in sight

Same happens in the rest for example the super fast smileton run:

with an ast but...

Ast evolves into Blm

and the same happens for the rest of the top 10 runs that show a healer, there are no top 10 runs with a healer in them

[Sebazy]

I hate to be the bearer of bad news, but you've read the logs incorrectly sir. Granted the overall list is actually kind of confusing on this one so I do sympathise.

On the team speed rankings overview page that shows the list, for whatever reason the job icon shows that person's 'main' class. I'm not sure how it's decided, maybe by whatever's currently equipped on lodestone or the job that they have the most uploads on. Eitherway it's not actually the job they ran that dungeon on in that example.

However, If you click into the 10:48 run you'll immediately see that Pori was actually on War, the DPS were BLM BLM and Ninja, for the second run they switched out the NIN for DRG.

So for dead ends, the fastest 3 DPS is 10:48

The fastest run with a healer is 12:16

For Smiletown, the 8:32 run you've linked is actually BLM BLM NIN WAR again.

Add a healer and the fastest time is 10:02, another run with Si'lvinna in a group called ORCA Speedrun, so they clearly aren't hanging about.

For Dreamscape, it's 9:07 for 3 DPS vs 10:27 with a healer.

In Alzadaal's Legacy it's 9:13 vs 11:07 with heals.

In Troia it's 9:30 vs 11:37

And lastly, for Manalis it's 10:55 vs 12:26

For reference the easiest way I've found to get the fastest healer run is to select speed fast the sorting then select all healers from jobs. For whatever reason not all runs end up on the overall team speed rankings, no idea why but the times seem valid.

The takeaway is that depending on the dungeon, taking a healer costs you somewhere between a minute and a half to 2 minutes and that'll likely scale up relative to the overall dungeon time with less optimised play or jobs.

(Beaten to it! The healer vs 3 dps numbers are still useful though!)

[GrimGale]

[Sebazy]

I know the above is in jest but before anyone actually jumps on Rena over the mistake, remember that he has openly stated in the past that whilst he has some Savage experience, it's not a significant amount nor does it go back that far. Chances are he's not really used logs for much if any real in-depth analysis before. This is fine, but it does also demonstrate the lesson that it's always wise to dig deeper before drawing any conclusions from Logs.

Building theories purely at a surface level is wildly consistent and might as well be a coin toss as to if you're accurate or not.

TLDR: Delve deeper before you jump to a conclusion sirs++

[ASkellington]

it's pretty sad that it's possible in the first place

This is the reason why its a point of contention. I'm sorry, no trinity game (healer/tank/dps) should allow one of the trinity to be irrelevant, in this case healers, because the outgoing damage is that friggin low. Even if we discard W2W I'm pretty sure on my GNB I can solo an EW boss. In fact, I have! From above 50% hp.

That's not a sign of good skill expression. That's a sign the devs have made healing far too low for w/e reason they claim to be a good one.

[Risvertasashi]

So even if 3DPS runs were slower by some nominally small amount. I'd still say it just feels bad to play a job that's only included because the duty finder requires it for matched parties, and that people are all too happy to drop the moment a coded system isn't enforcing it.

[fulminating]

I mean I’ve had the misfortune to carry out nigh solo runs of expert dungeons both this and last expansion - on some level I like that I’m able to but at the same time it doesn’t feel healthy for the game. In fairness I’ve also tanked as scholar, but I wasn’t exactly wall to walling there.