Picto painting potency should be readjusted

[Aco505]

But the balancing metric remains rdps anyway

The balancing metric is whatever SE decides. While they might check the player-created metrics, they likely use their own internal system for this, so all we can do is speculate on the specifics.

The reason why players look at rDPS, aDPS and cDPS and then get annoyed when job A or B is not where it should be is because the game cannot be balanced around high percentiles. This is because following such a model would mean that buffing jobs would always be worse unless played at a high level and with an appropriate party that can make use of such buffs, because some jobs simply do not buff feed as well as others.

So what the developers probably do is balance around what we'd call the 50-60 percentile ranges, which then translates into buffing jobs being inevitably "favored" when buff windows and kill times are exploited. Many 99 or 100 logs are simply so because the boss dies right after a 2-minute window either through manipulation or chance, and so are not necessarily fully representative of overall job or comp balance.

Additionally, all this mostly applies to organized 8-people content, mainly Savage raids. Ultimates were not the focus when balancing even though recently the developers did increase cleave damage of some jobs due to FRU P4 and later M6S, as well as the nerfs to PCT.

However, outside of 8-people content, non-buffing jobs and some other jobs like PCT reign supreme because of their superior personal damage, that does not rely on the party, and burst power. Just look at how the two new DT jobs have wrecked the level 100 dungeons damage-wise or which jobs besides NIN/SCH have been at the top of the CODCAR rankings.

[Carighan]

Importantly two things inherently change how resistant to messing up you are:

* Group buffs
* DoTs

The more of these you have (think Gunbreaker with its two DoTs, or for extreme cases Bard with its 1-2 buffs + 2 DoTs) the lower the risk of reducing your output as more of it is either automated, or does not depend on yourself. This is why Astros used to be ahead while people were learning fights: So long as you spent your buffs, you could "overheal" and spend too many GCDs on that, and still do good rDPS as the DPS you buff are, well, DPSing.

And I readily agree with Aco505 that even those two things alone make looking at xDPS for class balance problematic. It's one thing if you want to argue "Hey, I'm more than 10% (or whatever) away from the average, that seems excessive!". That's fine. But going "Yo in fight X class A is ahead of class B by N% but in fight Y class B is behind by M% and that needs a change to ability 145 by 15 potency to fix" that's faaaaaaar too specific an issue to look at. In fact the moment you compare rDPS vs cDPS or something you not only misuse what those numbers are meant to be used for (cDPS for comparing how well you perform when the rest of the group around you changes but nothing else does, for example, personal improvement and all that), but also try to invent a problem where there likely is none (if you have to look that specific).

Because, like Aco505 says, the devs are trying to balance the game for the actual playerbase, not just the few angry tryhards we represent sitting here on the forum rambling about marginal difference in output in highest-end fights.

[Kishin12345]

[QUOTE=Aco505;6792708]The balancing metric is whatever SE decides. While they might check the player-created metrics, they likely use their own internal system for this, so all we can do is speculate on the specifics. [QUOTE]

What i meant was the playersbase's balancing metric. When rdps is off players complain. That is what people look at.
If you ask me, the devs don't have any balancing metric and if at all very bad inhouse testing.
All we got from them was "we didn't expect picto to do so much damage"

[Carighan]

That's not unusual though. Every dev company of every game and most non-games runs into that problem, you can't test internally at scale, and nobody in your team not even professional testers test remotely as close to real-world conditions as your actual users will.

Get it "good enough" and fix it later. Or do the GW2 thing and throw 1.35 million options at a wall, see which stick, then forget about the others (but also never remove them).

[Aco505]

"we didn't expect picto to do so much damage"

A hilarious statement considering there were threads in this very forum on week 1-2 after 7.0 released warning about the job's power.