Why people still seem to be disagreeing about SAM/Kaiten

[cjbeagle]

Let's consider 6.08 crit variance of Midare using some hypothetical, but reasonable stats.

Let's say a fight takes 10 minutes.
Let's say you Midare once every 12s on average (taking skill speed, Meikyo Shisui, and Tsubame-gaeshi into account)
Let's say you have a 30% chance to crit.
Let's say your non-crit Midares hit for 25k.
Let's say your crit Midares hit for 40k. (+60% modifier)

If you Midare every 12s on average for 10m, that's 50 Midares for the entire fight.

Worst case scenario, you get 0 crits with Midare and do 1,250,000 damage with Midare over the course of 10m, which is 2,083 dps.

Best case scenario, you get 50 crits with Midare and do 2,000,000 damage with Midare over the course of 10m, which is 3,333 dps.

Now, 1.3k dps is nothing to scoff at, but:

- the chance of actually getting 0 crits out of 50 swings with a 30% crit chance is .0000000179846504%.

- the chance of actually getting 50 crits out of 50 swings with a 30% crit chance is .00000000000000000000000000717897988%.



Now let's look some possibilities that are more reasonable:

Let's say for a "lucky" clear, you crit 40% of the time despite only having a 30% crit chance - I think we'd all agree that'd be pretty splendid.

Likewise, let's say for an "unlucky" clear, you crit 20% of the time despite having a 30% crit chance - I think we'd all agree that'd be pretty dismal.

The fight you crit 40% of the time you'd do 1,550,000 damage with Midare over the course of 10m, which is 2,583 dps.

The fight you crit 20% of the time, you'd do 1,400,000 damage with Midare over the course of 10m, which is 2,333 dps.

Call me crazy, but imo their obsession with mitigating crit variance is completely unnecessary if great luck vs terrible luck only amounts to 250 dps.

[AsiTsurugi]

I think they meant 4k damage (per Midare), not dps. What I'd be interested to know is what the dps difference is between an "unlucky" clear and a "lucky" clear. Over the course of an entire fight I find it hard to believe that crit variance matters that much basically ever.

Anecdotally, as someone cursed with the most horrible crit rates on midare and ogi, in P4S P2 in patch 6.08, the difference between a "good RNG" run reflected by uploaded logs vs a typical run from me with crap like 0% ogi crits, 15% midare crits was about 4%. It was miniscule and barely even mattered.

[YukiB]

Call me crazy, but imo their obsession with mitigating crit variance is completely unnecessary if great luck vs terrible luck only amounts to 250 dps.

Yes, it is.

To their credit, it's more of an issue in short bursts like the opener and every 2 minutes after, where you have 2 Midare, a Kaeshi, and Ogi+Kaeshi for 5 big hits total. The chance of getting 0 out of 5 with 30% crit is a much more reasonable 16.807%.

To their detriment, this wouldn't be a problem if they didn't design all the jobs and all the fights to operate in these short burst windows interspersed with a "1-2-3 and 1-2-3 and follow the flashing symbols" scripted dance. It is, once again, a problem of their own making, and rather than address their job and fight design they decide to remove crits from an RPG. Insane.

[cjbeagle]

Yes, it is.

To their credit, it's more of an issue in short bursts like the opener and every 2 minutes after, where you have 2 Midare, a Kaeshi, and Ogi+Kaeshi for 5 big hits total. The chance of getting 0 out of 5 with 30% crit is a much more reasonable 16.807%.

To their detriment, this wouldn't be a problem if they didn't design all the jobs and all the fights to operate in these short burst windows interspersed with a "1-2-3 and 1-2-3 and follow the flashing symbols" scripted dance. It is, once again, a problem of their own making, and rather than address their job and fight design they decide to remove crits from an RPG. Insane.

I won't get into Ogi, but for Midare specifically (and I'm counting Kaeshi as a Midare fwiw, so 3 during burst), that's still only 3 out of the 10 you're doing per 2m, so even if those did double damage from buffs, it'd still only bump the dps difference between the 20% and 40% crit rate clears from 250 to 325.

I don't know how much more damage you do during burst windows, but I know it isn't double. =P

[Ruminous]

I won't get into Ogi, but for Midare specifically (and I'm counting Kaeshi as a Midare fwiw, so 3 during burst), that's still only 3 out of the 10 you're doing per 2m, so even if those did double damage from buffs, it'd still only bump the dps difference between the 20% and 40% crit rate clears from 250 to 325.

I don't know how much more damage you do during burst windows, but I know it isn't double. =P

Hard to get any real numbers now with the changes in, but after seeing Dragonsong War, I'm definitely in the thoughts of them trying to get ahead of any "balance" related problems samurai's 2m window might have. Still is a pretty crapass excuse to nuke the job and not compensate with another mechanic for it, but it's all I can think of.

[Shurrikhan]

I think they meant 4k damage (per Midare), not dps. What I'd be interested to know is what the dps difference is between an "unlucky" clear and a "lucky" clear. Over the course of an entire fight I find it hard to believe that crit variance matters that much basically ever.

About 3%, iirc?

[AsiTsurugi]

Hard to get any real numbers now with the changes in, but after seeing Dragonsong War, I'm definitely in the thoughts of them trying to get ahead of any "balance" related problems samurai's 2m window might have. Still is a pretty crapass excuse to nuke the job and not compensate with another mechanic for it, but it's all I can think of.

Here's why I think their excuse in this regard is completely nonsense: they changed NIN to have even higher burst than before (while, again, gutting the core identity of the job), which is perfect for a fight with tons of short burst phases and mechanics-only phases approximately 60s apart like Dragonsong ultimate. And yet they want us to think only samurai has this "issue" with respect to "crit variance", when NIN routinely invalidates these DPS checks they thought were so sacrosanct that they saw fit to effectively remove an entire job from the game and replace it with as much of a nothing as Yoshida's apology posts.

[Alex1337]

Here's why I think their excuse in this regard is completely nonsense: they changed NIN to have even higher burst than before (while, again, gutting the core identity of the job), which is perfect for a fight with tons of short burst phases and mechanics-only phases approximately 60s apart like Dragonsong ultimate. And yet they want us to think only samurai has this "issue" with respect to "crit variance", when NIN routinely invalidates these DPS checks they thought were so sacrosanct that they saw fit to effectively remove an entire job from the game and replace it with as much of a nothing as Yoshida's apology posts.

i fee like the ultimate is the reason they did this to sam when they did; nin and 6.08 Sam burst on dsu might been busted with all the breaks and they probably thought it’s easier to gut Sam burst than nin. Still shouldn’t be an excuse but whatever.

[xAFROx]

I'm a casual who still used Meikyo to build up for Midare instead of Higanbana and I still think it felt terrible without Kaiten, and that's not even taking into account that you're using Midare way more than Meikyo, so even if it did feel good, it'd only feel good like once a minute - that's...not great.

I miss Kaiten more than anything, but what I'd really like to see is a full reversion back to 6.08. Give back Kaiten, undo the potency redistribution, make Tanka Goken a cone again - undo everything.

Hear hear! I use Meikyo for whatever might be needed. Or rather, my opener is just not optimal.

Back in ShB, I would build my buffs and then use Higanbana, alongside Ikishoten and Senei, then go from there, or I'd build to Midare, use Senei, and then do the Yukikaze combo to put Higanbana, or some weird jumble of actions so long as I wasn't breaking a combo.

Then in EW, with 2 charges, I just keep one available for correction or getting he attack quickly. I didn't really try any of the current EX fights yet due to lack of playtime.

6.1 specifically, I Meikyo, Kasha and Gekko, Ikishoten, Senei, Ogi/kaeshi, shinten-Yukikaze-shinten-Midare/Kaeshi.

No idea if anything like that was optimal, probably not since it didn't involve Higanbana.

I've tried to fill the Kaiten-less void with Shinten and it's way too quick of an action to mentally buffer for Midare. It really isn't fun. I can make use of the kit without kaiten, but there's always going to be a feeling that it isn't right.

[Kabooa]

It's to remove damage variance. If you do any log reading, you'll notice that the variance between midare setsugekka now on most is within about 4k damage difference if it's at the same ilvl. The only thing it can do is crit or direct crit, so it keeps the average between multiple samples very close.

That being said, while it's good for balancing, it's less fun for the player. The excitement of building up to a big hit is all gone. Midare setsugekka is just "another skill" that you do. It just saddens me that critical hits are the problem and the system they have for it is flawed. Seeing big numbers, no matter what role you are is definitely fun and exciting. Making it common just rips a lot out of it.

I'm personally in the camp where "Criticals" should be the result of completing a conditional instead of random or auto applied.

That's a little incompatible with stats as they are currently, but outside some notable RNG situations, that's how PVP works. Every class has a few critical conditions.

Just spitballing, imagine if any Draw Attack gaining its critical bonus was connected to successfully defending with Third Eye*. Now I'm not going to pretend that's a great example, but consider how you might have to bend and change to individual encounters to land as many of those as possible.