[X] My Assistant

Holy and Dark Imperil is capped at 25% while Imperil mitigation reduction for fire/wind/elec/cold is capped at 40%.
Unless you are up against SGs Imperil works better with the 4 elements.

*nods*

since you can bind OFC to a single keypress I don't see how it can take up to 5 seconds (which are a whopping 12.5 turns at his speed).

ah, ok. sure, that's another matter then. i don't know if it's cryosite's case but keybind lowers cast time indeed, yep.

1. Using Imperil on every SG.
2. Casting OFC (almost at the beginning of the round) - this makes all regular enemies die at single blow and SG stay half damaged.
3. Then I am attacking SG with my regular non-spiritStance attacks (if spiritStance hasn't depleted - that even better).
4. Few rounds later my OC is regenerated enough to make me activate spiritStance back (if it had been depleted).
5. Then I just killing impireled SG with my spiritStance attacks as usual.
6. At new round, if my OFC has been recharged - I am starting from step #1, if not just cast Imperial and attack enemies as usual.

Since when do axes crit more often?

Also how can anyone lose 3-5 seconds when casting a single OFC?
Isn't that too much?
Even considering the cases where you get out of spirit stance I cannot see anyone "wasting" more than 2 seconds.

I forgot to change difficulty before starting today's DWD. So I decided to give more focus on the OFC technique.

I based my usage on F4tal's description above. Some of why it takes longer for me is due to not having it trained into muscle memory to do smoothly yet. Keep that in mind.

I found I could use OFC 25 times total in the run; I didn't bother using it prior to round 113 with fewer than 3 SGs. Since I was in a counting mood, I noticed I had to heal 7 times; twice to spark, five just due to getting into the redzone of HP. Thank you Monsterbation for helpful warnings.

Out of those 25 uses of OFC, once I kept enough OC to stay in Spirit Stance. It took my clearing two rounds normally to build up enough OC to use OFC again.

The other 24 times I lost Spirit Stance due to OFC use. Sometimes I built up 10 OC pips out of Spirit Stance finishing off the SGs wich meant I was at or nearly at 50t cooldown for the next round. Other times I had to deplete cooldown/rebuild OC in the following round. So sometimes I could use OFC "every round" and sometimes I could use it every other round.

So my initial estimate of 12 uses per DWD was pretty off (had guessed 1 every three rounds). So the time savings from OFC can be double what I originally estimated.

As for the "use a hotkey binding" not everyone uses Monsterbation. It might take moving the mouse over to your quickbar, clicking, then moving your mouse back to the enemies. I do have it hotkeyed, so I do shave off a little of that time, even without it being a practiced technique.

What really slows it down is that you have to stop doing your hoverplay "normal" attacking for a moment. Look at OC, Look at OFC for cooldown, make a quick decision if it is ok to do OFC this round (some of which can be estimated the round prior with practice, so this can be reduced too).

Then there is the "setup." Casting imperil (even using hotkey) can be a single cast if lucky (SG's are next to each other and Imperil lands on all three). It may take 2-3 casts due to resists on a stubborn SG. Part of this "setup" is paying attention and having to recognize how successful Imperil was, and continue with more Imperil or hit OFC as appropriate.

Another annoying part of Setup is that you're expending OC without landing melee attacks to maintain it. So the setup process can sometimes lower your available OC to less than needed for OFC and cause a misfire (you earn cooldown, but get no damage).

One thing I mixed into that setup is to make some regular attacks on one of the SG's, especially if I still had 5 or less cooldown on OFC. Once I landed Penetrated Armor, I'd cast Imperil on one of the other SG's (especially if the Penetrated Armor one is by herself and the other two are next to each other).

All of that takes a little bit of time to move your eyeballs around, think a moment, then react. It will get faster as it becomes practiced, but it can be slow-going while learning it. Even when practiced, it can be many hotkeys/clicks, not just one. A guess of closer to 2 seconds seems pretty fair, but I don't think a slower time of 3 or even more seconds is unrealistic.

This will vary by person obviously a whole lot.

Which leads me to the other bit:

I don't think the general metric most people have adopted of "turns = fair speed estimate" is as fair or useful as people claim it to be. Yes, latency will impact your t/s, especially for normal stretches of hammering away at hoverplay. It will also inflict lag on hotkey uses as you wait for the game to respond. But I am not too sure if things like using a potion increase your t. Not every t used is equal either. Hoverplay normal attacks are "very fast t" while "use a hotkey that you're very used to using and takes very little thought" is a "slower, but still fast t" and "I have to pay attention to what is happening and make decisions" may not increase t, but takes time.

Part of what speeds up clear is how fast your overall style is. How much of your style lets you coast on very fast hoverplay t, vs how smooth your techniques are and how much time they save, and how little time you spend on slow decisions.

I think instead of trying to compare t counts, people really should still compare time. But it's a value you need to translate. Does your arena clear include getting up in the middle and getting a snack? Are you lazily playing while watching hentai? Or is this a run you're actively trying to push yourself to go as fast as you can? What does your latency look like compared to what other people have? Is your latency slowing you down by a lot and how much?

t/s helps us figure out that last one, but both lag and a very lazy/inefficient style can both worsen t/s. Very slow and poorly executed techniques can reduce total t, but increase your clear time.

I think it's reasonable to use turns to measure your speed because everyone has vastly different internet connections and setups. Time is the most accurate way to measure speed but there is no universal way to quantify it that applies equally to everyone.

Also, speed is not everything. It is not unreasonable to optimize your playstyle for turns/efficiency/resources rather than speed, especially for players with slow connections.

I am not sure about decondelite's calculations for penetrated armor. I was under the impression that 3 stacks would reduce a monster's mitigation multiplicatively, not additively, meaning a little bit of mitigation always remains. This doesn't invalidate his related arguments, just pointing it out. I did test earlier that setting elemental resistance aside, Imperil does not provide any further stacking with Penetrated Armor x3 though. So it's either capped at 75% or additive like decondelite said.

I think it's reasonable to use turns to measure your speed because everyone has vastly different internet connections and setups. Time is the most accurate way to measure speed but there is no universal way to quantify it that applies equally to everyone.

Also, speed is not everything. It is not unreasonable to optimize your playstyle for turns/efficiency/resources rather than speed, especially for players with slow connections.

Speed isn't everything, but we're discussing it currently because it is important to a lot of people. No one suggested it was everything. Do you go into threads that don't interest you and chide people for discussing something that isn't of interest to you?

But, since you insist:

How to play more quickly and efficiently is more important to players with a slow connection than those with fast. Because they stand to gain the most by making faster choices which can have bigger impacts on how long their runs take. Latency causing a normal strike to take 5x as many milliseconds as it takes me just means that for you, the saved t portion of what we're working on here matters about 5x as much. Considerations like "move eyeball to different parts of screen, think about decision, and hit hotkey" become less impact because time spent waiting on pages to load due to lag can be spent doing those things or preparing for them (hovering finger over hotkey).

On the other hand, something that takes more t but involves less user inputs might be preferable to someone with a slow connection. Forgoing OFC, despite the fact that it reduces t used at the end is a valid choice, if the choice lets you pay less attention to the game and more attention to your hentai. That has nothing really to do with latency, though someone with a lot of latency might enjoy longer periods of time spent poking at Yuki with their rapier while they look up hentai of Yuki being poked with someone else's rapier.

Speed still matters, but it can be achieved in different ways. Discussion like what I'm suggesting reveals that and allows someone to make a more informed decision. This is why I'm blatantly ignoring suggestions to just give the tl;dr version. There isn't one. This is an attempt to build actual knowledge, not just tell people what to do.

I am not sure about decondelite's calculations for penetrated armor. I was under the impression that 3 stacks would reduce a monster's mitigation multiplicatively, not additively, meaning a little bit of mitigation always remains. This doesn't invalidate his related arguments, just pointing it out. I did test earlier that setting elemental resistance aside, Imperil does not provide any further stacking with Penetrated Armor x3 though. So it's either capped at 75% or additive like decondelite said.

From your link:

Monster's physical mitigation is reduced by 25% per stack, with maximum of 3 stacks (mitigation reduced to 25%.)

I'm not terribly concerned if the defense reaches true zero or not. Hitting bottom from one source of defense bypass and having a very tiny (due to multiplicative) further reduction of defense from using both sources results in a tiny gain. Not as impactful as the first source chosen.

Using one defense bypass (penetrated armor or Imperil) results in close to a doubling of damage. Very worth doing. Using both results in either no increase in damage, or at best another 12.5% reduction to pmit. (if you're correct, the equation would look like: base pmit * (1-0.75)*(1-0.5)=pmit. So an estimated 50% pmit would become 25% with Imperil only, 12.5% with three stacks of penetrated armor, and with both 6.25%. Decondelite's example of an athropod with 80% would go to 40%, 20%, and 10% respectively. Things with low pmit would see correspondingly less difference in mitigation.

All three "defense bypassed" configurations give pretty close to the same damage values, and we'd need precision like Decondolite expected to really show the difference. However, I don't see that being too important.

Spirit Stance, Heartseeker, no penetrated armor:
Void Strike hits Yuki Nagato for 2993 void damage.
Estimated normal strike: 5500 void damage.

550k hp/(3k+5.5k) = 65 swings to kill Yuki. For example, using a shortsword/axe/club that does comparable damage to my rapier, and not using Imperil. Also represents those t spent attacking when penetrated armor fails to proc promptly. Shortsword/Axe would add bleed damage to this and reduce swings to kill.

Spirit Stance, Heartseeker, no penetrated armor, Imperil:
Void Strike hits Yuki Nagato for 6176 void damage.
You hit Yuki Nagato for 11453 void damage.

550k hp/(6.1k+11.5k) = 32 swings to kill Yuki. For example, using a shortsword/axe/club that does comparable damage to my rapier, and using Imperil. Shortsword/axe would add bleed to this damage and reduce swings to kill.

Spirit Stance, Heartseeker and penetrated armor:
Void Strike hits Yuki Nagato for 5683 void damage.
You hit Yuki Nagato for 11921 void damage.

550k hp/ (5.7+11.9k) = 32 swings to kill Yuki. My typical rapier performance.

Spirit Stance, Heartseeker, Imperil, and penetrated armor
Void Strike hits Yuki Nagato for 7501 void damage.
You hit Yuki Nagato for 13962 void damage.

550k hp/(7.5+14k) = 26 swings to kill Yuki. My rapier plus imperil. Experience has shown this results in approximately 200 less t spent clearing DWD (4900-5100t).

Numbers off due to damage variance, Overwhelming Strikes bonuses, and crits. The fact that Imperil and penetrated armor results both indicate a matching number of turns is coincidental, not definitive. The fact that combining both gave more damage is also not conclusive. The imprecision Decondolite objected to can account for these variances or lack.

But what I am getting at is that all three configurations result in something substantially faster than no bypassed defense at all. Even if one works out to be slightly better on average, it doesn't result in an actual, meaningful increase to kill speed.

Bypassing defense, regardless of how you do it, seems to result in pretty close to (but not quite fully) a doubling in damage dealt. This means that Rapiers are nice because they hand out penetrated armor to accomplish this with no extra user input, while Imperil does so at the cost of more button mashing and mana use, but lets non-rapier weapons compete in damage performance. That part of this theory so far seems pretty sound.

The wiki itself says that the PA stacks and Imperil are additive between themselves, and it is highly suggested (in a very awkward way, I agree) that they are also additive to the monster PMit.
Someone told me so aswell some time ago, don't remember if it was tenbrony, super or jenga.

I'm planning on doing a benchmark with my favourite test weapons (a IW10 crude rapier, and I'll build an IW10 crude shortsword aswell) to check all of this soon. And more importantly: I'll design a test protocol to confirm if the values for schoolgirls are correct or not. Finally, I'll probably build a table of average monsters PMit depending on player level, Power Level and monster class. That should help us see at what extent exactly Penetrated Armor and Imperil are useful.

Not doing it right now, I gotta do other things before.

The wiki itself says that the PA stacks and Imperil are additive between themselves, and it is highly suggested (in a very awkward way, I agree) that they are also additive to the monster PMit.
Someone told me so aswell some time ago, don't remember if it was tenbrony, super or jenga.

I'm planning on doing a benchmark with my favourite test weapons (a IW10 crude rapier, and I'll build an IW10 crude shortsword aswell) to check all of this soon. And more importantly: I'll design a test protocol to confirm if the values for schoolgirls are correct or not. Finally, I'll probably build a table of average monsters PMit depending on player level, Power Level and monster class. That should help us see at what extent exactly Penetrated Armor and Imperil are useful.

Not doing it right now, I gotta do other things before.

I look forward to your findings.

I've comissioned IW for a legendary shortsword I had which is somewhat comparable to my rapier. Once it is done, I'll forge5 it, and start testing out with it on my end.

(in a very awkward way, I agree)

Nonono.
In a very useful™ way.