Sujet : Making Jobs Rotationally Easier in 4.0

It's not a matter of acceptable or not, but it's possible to carry 1-2 bad dps if we're talking about normal mode. I often end up in the top 3 when I parse a normal alex run despite being a tank, so that means there are at least two underperforming dps in those parties.

Envoyé par aleph_null

It's not a matter of acceptable or not, but it's possible to carry 1-2 bad dps if we're talking about normal mode.

That's exactly what I think shouldn't happen. Of course, I'm not talking about not being able to "carry" someone would only do "slightly less than optimal", but someone doing "blatantly obviously bad without needing a parser to tell what's wrong".

For DPS, it's kind of easy to set. You put some DPS checks, and even personal DPS check so that someone could not compensate for you. But how would you do that for a tank ? If your DPS are great, your contribution is actually not that important, so you're easy to carry. The only thing where tank are on their own is generating enmity and surviving...which are both trivial before Savage.

We can argue all day about whether carrying a few bad players (regardless of role) in normal mode is acceptable or not, but currently it's possible and I don't think it'll change anytime soon, considering the gap between the good players and the average people you find in DF, which leads to what this thread is (supposedly) discussing about: trimming the skill set of jobs and making their rotations simpler.

I believe no one likes having to take on the burden of carrying underperforming players, especially those who refuse to listen to advices to get better, but as of now I think it'd be bad if you tune normal mode raids to something that would be impossible to clear with 1-2 bad players since the success rate of DFing those would plummet significantly. Fyi even in a9s and a10s you can clear while carrying a dead dps as long as the other 7 people are competent, meanwhile my friends who RF these fights every week still see wipes due to enrage or failing dps checks (adds). I don't think tuning a9n-a12n to something of that level would be something SE would do. Personally I have accepted that normal mode raids are places where you can carry a few bad players.

Envoyé par Ultimatecalibur

You need two identical comps for the numbers to be mostly comparable and to also check and see how "lucky" each group got.

You don't need to compare two different attempts to understand the defensive trade-off. You just need one. The vast majority of the damage in A9S comes from autos, with intermittent damage spikes from Scrap cleaves.

Autos hit for about 6k without Grit. You would have mitigated about 1.2k by being in Grit. The numbers are actually quite similar between the two runs if you calculate what Grit's contribution is in each case. The big difference between the runs is the number of autos received: 63 in the faster run, and 95 in the slower one.

It's a similar story for Scrap: you take 8k without Grit, and mitigate 2k by being in Grit. The difference again is that the faster attempt only sees half the number of cleaves as the slow one. It's worth noting that the upper bound damage for Scrap is about 10k without Grit, which is important for both you and your healer to gauge what the minimum safe threshold is before intervening.

I'm not really sure where you got the idea of the "probable"(?) soaked Double Scrapline from. You can directly check.

Damage Taken Per Second is not particularly useful here. Damage occurs at very discrete intervals, as does healing. The difference in damage taken with or without Grit is only relevant if it forces an additional heal or if it results in a death that could not have otherwise occurred. A 2k difference gets swept up in a Regen tick, and we're looking at max HP in the 32-34k range.

Turning off Grit is more than a 25% dps increase for a couple reasons. First, you have access to BW, which influences both CPM and available MP, both of which directly translate into dps. Second, as the raid encounter time drops, you spend more time in your opening burst. Even if you could keep all other external factors constant (i.e. party comp, buff/debuff uptime, proc and crit rates for all players), this second point is something that you can't control for. Clearing more efficiently will "inflate" your dps significantly. And even if it was only a 25% boost ("only" being more powerful than a permanent RS or Balance): why give yourself a dps penalty needlessly?

Envoyé par Lyth

The big difference between the runs is the number of autos received: 63 in the faster run, and 95 in the slower one.
[...]
The difference again is that the faster attempt only sees half the number of cleaves as the slow one.

But you can't realistically put that only on the tank stance dancing. The whole party did better, wether it's from experience, gear or composition, and they all contributed to the faster kill.

That's why a test should be made within two runs of the same team where the only (player-induced) variable is "stance dancing or not".
If you end with 63 autos in the fastest run and 70 in the slowest, it's not that big. (Numbers only here as an example)

Envoyé par Lyth

And even if it was only a 25% boost ("only" being more powerful than a permanent RS or Balance): why give yourself a dps penalty needlessly?

The point is not really why you should give yourself a penalty, but instead "Why you should put that much pressure on DPS optimization for tanks if the difference is not that important ?" Unless of course, you're doing a Savage race...

Yes, the difference isn't entirely due to tanking out of Grit. But realistically, you can't really construct the sort of experiment that you're suggesting, because of the reasons that I outlined above. Nor do you need to, to understand the benefits.

Envoyé par Reynhart

The point is not really why you should give yourself a penalty

Actually, that was pretty much exactly my point.

Envoyé par Instrumentality

Unequivocally yes.

For you...
In the end, the impact you have is far smaller than the effort you agreed to make.

Envoyé par Reynhart

For you...
In the end, the impact you have is far smaller than the effort you agreed to make.

Isn't that the way tapering skill cap works? As with quality vs. cost, effort tends to increase more exponentially for more linear gains in output.

For better or worse, that's been a targeted design methodology in many a game, right alongside "easy to learn, difficult to master".