[X] My Assistant

Penetrated Armor is 3.15 * 20 / 8 ≈ 7.8 times stronger than Bleeding Wound. With Imperil applied, PA stacks suboptimally (multiplicatively) thus becomes 3.9 times stronger. Unless there is still a logical error which is quite possible. Decondelite also pointed out that Bleeding Wound damage applies a turn earlier than Penetrated Armor.

This seems unfair and I figured out the problem. What I wrote is correct but only strictly applies in the long run. Both work over time so if enemies die quickly Penetrated Armor and Bleeding Wound become less important, and that extra first turn of Bleeding Wound gives it a big edge over PA.

Normally I estimate enemies to die in 3 hits from a strong rapier player, and that each PA applies on average for 1 hit. If PA sticks in turn 1 it applies on 2 following hits, if PA sticks in turn 3 it's useless, so (2 + 1 + 0) / 3 = 1. With imperil, let's say that enemies die in 2 hits. Now each PA applies to (1 + 0) / 2 = 0.5 hits while BW averages (2 + 1) / 2 = 1.5 bleeds thanks to that extra first one. 1.5 / 0.5 = 3 and makes up for that 3.15 damage factor with PA.

Imperil Shortsword vs Imperil Rapier (Theoretical Math Comparison)

To simplify I'll assume both players cast until all monsters have imperil. This eliminates any considerations of deprecating proficiency.

Here are calculations against a PMit = 0.8 monster. After imperil, PA drops PMit from 0.4 to 0.3, so the damage is worth 0.1 * 3.15 * 0.5 = 15.75% of "base damage" while BW is worth 8% * 1.5 = 12% of "base damage." The difference of these two is 3.75% of "base damage," or 3.75% / 0.6 / 3.15 ≈ 2% of actual damage each turn.

Here are calculations against a PMit = 0.7 monster. After imperil, PA drops PMit from 0.35 to 0.2625, so the damage is worth 0.0875 * 3.15 * 0.5 = 13.78% of "base damage" while BW is worth 8% * 1.5 = 12%. The difference of these two is 1.78% of "base damage," or 1.78% / 0.65 / 3.15 ≈ 0.87% of actual damage each turn.

The peerless warrior has 4.209% more attack base damage with a shortsword. Therefore shortsword (or maybe axe at peerless high level) wins by a respectable margin! Furthermore, that 2% boost from rapier PA needs to be multiplied by the chance to stick which is 1 - (1 - 57.88%) * (1 - 25%) ≈ 68% at peerless, so it's a mere 1.36% bonus at best.

There was some fudging regarding how quickly enemies die, and rapier will benefit if some enemies don't get imperiled, so this purely mathematical analysis may only show that the result is too close to be sure. Proficiency and playstyle issues will interfere with experimental comparisons so perhaps we can never have a definitive answer.

Just as a reminder, I've already done the practice test of rapier VS shortsword in a PFFEST.
And the shortsword was better by basically how much more ADB it provides.

That's also the reason why I got furious when basara said that basically "nothing was proven". That is, I did have put the shortsword at a disadvantage by not Imperil'ing absolutely everything, whih does force one to hack to death a decent amount of monsters.

This post has been edited by decondelite: May 8 2020, 01:44

While we are on the topic of grindfest
I've mostly mentioned about the advantages of BW in arenas

but in other battle modes (I've only done limited testing in IW's, so take it with a pinch of salt), I don't even rely on the proc, BW or PA. Spreading attacks as much as possible a fully imperilled round, to maximise counters, is my targeting strategy. Far from perfect, since I don't stop to decide on targets after every hit, just haphazardly aim at whatever has the highest hitpoints at any given moment. The closer everything is to dying at the same time, the better. Though, you have the drawback of it getting somewhat painful, as some monsters may build up their specials and survive long enough to shoot it, and you take more hits overall in the round but in less turns. With this strategy, better overall stats of the shortsword wins over rapier

That's basically how I play PFFEST once I've slashed to death unimp'ed monsters you know. And that's why it's important to use Protection/Warding.

Just as a reminder, I've already done the practice test of rapier VS shortsword in a PFFEST.
And the shortsword was better by basically how much more ADB it provides.

I considered your test "little" valid. (IMG:[invalid] style_emoticons/default/tongue.gif)

You cannot compare two devices with different IW results, different levels of upgrades and one soulfused and the other not. There are many variable parameters. (IMG:[invalid] style_emoticons/default/wink.gif)

So, if I don't consider your test "totally" valid, and you don't consider mine, and nobody else did it, then nobody proved the fact in question. (IMG:[invalid] style_emoticons/default/heh.gif)

Just to clarify, my recent math considerations suggest that even with imperil on, PA is about 3.9 times better than BW in the long run so it's not quite right to say they are equal. But given the reality that imperiled monsters die super fast and BW applies one turn earlier, they amazingly equalized with PA being just 1.3 times better. And comparing to the player's overall damage this difference is 1% of it.

This is consistent with decondelite's shortsword vs rapier test, as this 1% difference was not likely to be noticed by him. It's much less than the 4.2% greater attack damage when using a shortsword. But we still need to be careful because that 1% can grow big favoring either side if my 2 turns assumption is wrong.

When comparing shortsword vs rapier, a critical point is how strong/damaging the player is. When the player is peerless with DD9 this makes the shortsword or axe perform better. If someone could get DD20 and crush all imperiled enemies with one hit, then there is no use of the rapier as PA does nothing, and BW still does a little. We should all be prepared to acknowledge the superiority of shortsword under extreme limits.

But many players are not at that limit, and we are not quite sure what that limit is either. Do imperiled monsters really die in 2 hits average? And sometimes monsters don't get imperiled, so there is some wiggle room.

I considered your test "little" valid. (IMG:[invalid] style_emoticons/default/tongue.gif)

You cannot compare two devices with different IW results, different levels of upgrades and one soulfused and the other not. There are many variable parameters. (IMG:[invalid] style_emoticons/default/wink.gif)

So, if I don't consider your test "totally" valid, and you don't consider mine, and nobody else did it, then nobody proved the fact in question. (IMG:[invalid] style_emoticons/default/heh.gif)

OK. Go get two full forged Peerlesses with the exact same potencies. Now. (IMG:[invalid] style_emoticons/default/sleep.gif)

Penetrated Armor is 3.15 * 20 / 8 ≈ 7.8 times stronger than Bleeding Wound. With Imperil applied, PA stacks suboptimally (multiplicatively) thus becomes 3.9 times stronger. Unless there is still a logical error which is quite possible. Decondelite also pointed out that Bleeding Wound damage applies a turn earlier than Penetrated Armor.

I realized some huge logical errors, where is Sapo84 to fix my mistakes? Although I guess this particular mistake is shortsword specific, so decondelite was the best hope to fix me here.

Presuming you spread your attacks efficiently, Bleeding Wound is effectively multiplied by the number of monsters alive, so we could say it deserves a 5x ~ 8x multiplier in most cases, depending if in arena or grindfest/iw. Therefore even in the long run, Bleeding Wound is comparable to Penetrated Armor without imperil, and better than Penetrated Armor with imperil. (However it won't deserve this multiplier if you have to beat down an isolated monster that failed to imperil).

Against monsters with 0.7 PMit the rapier's PA worsens to 3.15 * 17.5 / 8 ≈ 6.9 times stronger than Bleeding Wound without imperil, and 3.4 times with. Around level 300 ~ 380 monsters will have 0.6 PMit so PA will only be 5.9 times or 3 times stronger than Bleeding Wound.

In consideration of this even non-imperil shortsword ought to be viable, and actually I suppose data gathered by various players did suggest that. It was worse than rapier but not anywhere near 2 times worse, like I might otherwise have expected. Thus Bleeding Wound is crucially important to shortsword play, as it acts like an alternative form of Penetrated Armor since its damage ignores enemy PMit.

One caveat of shortsword play: you absolutely must spread to get the full utilization out of Bleeding Wound, therefore it's important to make sure you are tough enough to survive with Power of Warding/Protection, especially if you do grindfest or item world. And you shouldn't say you won't do those (like some rapier players do) because grindfest and item world is when Bleeding Wound and shortswords perform best. Haste, Swift Strike, or shields with Agility should be avoided like the plague.

This also has interesting ramifications for other styles like Dual Wield. Axes, wakizashi, or shortswords can spread to attain greater performance like 1H. Club + Axe/Waki/Shortsword can be spread to stun all enemies and provide superior defense like 1H. Overpower would be a recommended potency in that case because the stun will often be wearing off.

Imperil Shortsword vs Imperil Rapier (Revised)

Assuming imperiled monsters die in 2 hits, our previous short term BW multiplier of 1.5 average bleeds is changed to 1.5 * 5 = 7.5, or 1.5 * 8 = 12 depending if in arena or grindfest. Rapier is getting crushed but I discovered a second logical error to help. PA works with counters on the same turn it is inflicted. And even if it didn't, we were supposed to have assumed it did anyway because most counters occur against non-targeted monsters.

So we should separate our 3.15 damage multiplier back into the 2.4 main hit and 0.75 per counter. PA once again only works on 0.5 main hits average, but now works on all counters. I'll increase the number of counters to be 1.2 per turn in arenas and 1.8 per turn in fests. Our new PA multipliers are 2.4 * 0.5 + 0.75 * 1.2 = 2.1 and 2.4 * 0.5 + 0.75 * 1.8 = 2.55

Here are calculations against a PMit = 0.8 monster. After imperil, PA drops PMit from 0.4 to 0.3, so the damage is worth 0.1 * 2.1 = 21% of "base damage" while BW is worth 8% * 7.5 = 60% of "base damage." The difference of these two is 39% of "base damage," or 39% / 0.6 / 3.3 ≈ 19.7% of actual damage each turn. In grindfest this changes to (96% - 25.5%) / 0.6 / 3.75 ≈ 31.3% of actual damage.

Here are calculations against a PMit = 0.7 monster. After imperil, PA drops PMit from 0.35 to 0.2625, so the damage is worth 0.0875 * 2.1 = 18.38% of "base damage" while BW is worth 8% * 7.5 = 60%. The difference of these two is 41.62% of "base damage," or 41.62% / 0.65 / 3.3 ≈ 19.4% of actual damage each turn. In grindfest this changes to (96% - 22.31%) / 0.65 / 3.75 ≈ 30.2% of actual damage.

The peerless warrior has 4.209% more attack base damage with a shortsword but that doesn't even really matter, it's dwarfed by spread Bleeding Wound being so much better than Penetrated Armor after imperil is applied, and especially in the short term. The 19% or 30% boost from shortsword BW does need to be multiplied by the 68% chance to inflict so it's more like a 13% or 20% damage boost.

Did I finally get it right this time? Maybe there is still a huge logical error somewhere.

I realized some huge logical errors, where is Sapo84 to fix my mistakes? Although I guess this particular mistake is shortsword specific, so decondelite was the best hope to fix me here.

Presuming you spread your attacks efficiently, Bleeding Wound is effectively multiplied by the number of monsters alive, so we could say it deserves a 5x ~ 8x multiplier in most cases, depending if in arena or grindfest/iw. Therefore even in the long run, Bleeding Wound is comparable to Penetrated Armor without imperil, and better than Penetrated Armor with imperil. (However it won't deserve this multiplier if you have to beat down an isolated monster that failed to imperil).

Against monsters with 0.7 PMit the rapier's PA worsens to 3.15 * 17.5 / 8 ≈ 6.9 times stronger than Bleeding Wound without imperil, and 3.4 times with. Around level 300 ~ 380 monsters will have 0.6 PMit so PA will only be 5.9 times or 3 times stronger than Bleeding Wound.

In consideration of this even non-imperil shortsword ought to be viable, and actually I suppose data gathered by various players did suggest that. It was worse than rapier but not anywhere near 2 times worse, like I might otherwise have expected. Thus Bleeding Wound is crucially important to shortsword play, as it acts like an alternative form of Penetrated Armor since its damage ignores enemy PMit.

One caveat of shortsword play: you absolutely must spread to get the full utilization out of Bleeding Wound, therefore it's important to make sure you are tough enough to survive with Power of Warding/Protection, especially if you do grindfest or item world. And you shouldn't say you won't do those (like some rapier players do) because grindfest and item world is when Bleeding Wound and shortswords perform best. Haste, Swift Strike, or shields with Agility should be avoided like the plague.

This also has interesting ramifications for other styles like Dual Wield. Axes, wakizashi, or shortswords can spread to attain greater performance like 1H. Club + Axe/Waki/Shortsword can be spread to stun all enemies and provide superior defense like 1H. Overpower would be a recommended potency in that case because the stun will often be wearing off.

Imperil Shortsword vs Imperil Rapier (Revised)

Assuming imperiled monsters die in 2 hits, our previous short term BW multiplier of 1.5 average bleeds is changed to 1.5 * 5 = 7.5, or 1.5 * 8 = 12 depending if in arena or grindfest. Rapier is getting crushed but I discovered a second logical error to help. PA works with counters on the same turn it is inflicted. And even if it didn't, we were supposed to have assumed it did anyway because most counters occur against non-targeted monsters.

So we should separate our 3.15 damage multiplier back into the 2.4 main hit and 0.75 per counter. PA once again only works on 0.5 main hits average, but now works on all counters. I'll increase the number of counters to be 1.2 per turn in arenas and 1.8 per turn in fests. Our new PA multipliers are 2.4 * 0.5 + 0.75 * 1.2 = 2.1 and 2.4 * 0.5 + 0.75 * 1.8 = 2.55

Here are calculations against a PMit = 0.8 monster. After imperil, PA drops PMit from 0.4 to 0.3, so the damage is worth 0.1 * 2.1 = 21% of "base damage" while BW is worth 8% * 7.5 = 60% of "base damage." The difference of these two is 39% of "base damage," or 39% / 0.6 / 3.3 ≈ 19.7% of actual damage each turn. In grindfest this changes to (96% - 25.5%) / 0.6 / 3.75 ≈ 31.3% of actual damage.

Here are calculations against a PMit = 0.7 monster. After imperil, PA drops PMit from 0.35 to 0.2625, so the damage is worth 0.0875 * 2.1 = 18.38% of "base damage" while BW is worth 8% * 7.5 = 60%. The difference of these two is 41.62% of "base damage," or 41.62% / 0.65 / 3.3 ≈ 19.4% of actual damage each turn. In grindfest this changes to (96% - 22.31%) / 0.65 / 3.75 ≈ 30.2% of actual damage.

The peerless warrior has 4.209% more attack base damage with a shortsword but that doesn't even really matter, it's dwarfed by spread Bleeding Wound being so much better than Penetrated Armor after imperil is applied, and especially in the short term. The 19% or 30% boost from shortsword BW does need to be multiplied by the 68% chance to inflict so it's more like a 13% or 20% damage boost.

Did I finally get it right this time? Maybe there is still a huge logical error somewhere.

I found an error: you used the wrong thread

I realized some huge logical errors, where is Sapo84 to fix my mistakes?

Enjoying his mage set I guess?
It's true that I probably was the first one daring to say that shortswords weren't that bad (if I remember correctly I had around 5% worse time with a shortsword, 1H no imperil) but I honestly wasn't that invested in that argument.

I also personally did not enjoy Imperil with 1H and I still don't enjoy it while maging, which is the main reason I went with Holy at the time.
It's fantastic that you, decondelite and others are analyzing 1H and trying to squeeze out even the slightest amount of damage, but what you end up with is a playstyle I honestly wouldn't enjoy too much (and in the end this is a game).
I think that in the end there should be a balance between the efficiency of a style and its complexity because it's that much easier to play HV while doing other things if the style is easy to play (I'm usually doing my ~50minutes of daily HV while listening to podcasts youtube videos and sometime farming in Azur Lane, no way I'm imperiling every round while doing that (IMG:[invalid] style_emoticons/default/tongue.gif)).

I think that in the end there should be a balance between the efficiency of a style and its complexity because it's that much easier to play HV while doing other things if the style is easy to play (I'm usually doing my ~50minutes of daily HV while listening to podcasts youtube videos and sometime farming in Azur Lane, no way I'm imperiling every round while doing that (IMG:[invalid] style_emoticons/default/tongue.gif)).

wow I am not the only one who consider not being bothered to use imperil every round is more convenient.
that's a surprise lol.

As I started using script a short time ago, I decided to make some more records. (IMG:[invalid] style_emoticons/default/sleep.gif)

I wrote down some data that I consider most relevant and put it in the table below:

(IMG:[imgur.com] https://imgur.com/wIbUwSb.jpg)

Legendary Hallowed Rapier of Slaughter + Infusion of Darkness, Slaughter Set, Full Forged, DD V

The results seem quite consistent. (IMG:[invalid] style_emoticons/default/happy.gif) However, this indicates that there was an unexpected error in the result I obtained earlier: https://forums.e-hentai.org/index.php?s=&am...t&p=5675159

I don't know exactly what happened, but certainly that number of turns below 3700 is not real. (IMG:[invalid] style_emoticons/default/huh.gif) But the elapsed time seems to be correct.

The tests performed on days 11 and 18 show the differences resulting from the continuous use of "Haste". The number of turns was within the average, but the elapsed time was the worst. (IMG:[invalid] style_emoticons/default/sad.gif) The big difference was in the amount of counterattacks, on average around 1000 less. (IMG:[invalid] style_emoticons/default/ohmy.gif)

Someday I do a test using less "Imperil" (or none). (IMG:[invalid] style_emoticons/default/tongue.gif)

I don't have any opinion on the usefulness of haste in school girl arenas because there are too many different ways to play them (targeting strategy, skill usage, imperil usage, etc).

However looking at your data, it suggests that Haste is neutral for you during school girl arenas. Time is what counts (using haste is supposed to boost turns/second, but not in school girl arenas) but your turns/second can vary with connection luck and the day, so in this case you should be comparing turns.

My haste test results are also subject to issues of luck which is why I don't insist that my turns/second results are definitive. I tried to reduce the randomness by having a wide array of different kinds of tests to cross check and confirm if turns/second boost exists on the average (which it did, maybe).

Also I can notice a gradual improvement in your turns/second on all the later days, as you get more used to using script. This is not related to haste usage.

4175/4208 = only 0.8% difference. You might as well therefore use Haste. Actually, because you don't have haste on IA, you wasted 17 turns to cast Haste. If you give back these 17 turns, the difference is only 0.4%

Someone earlier stated that for school girl arenas it is better to use haste only after the school girls appear. I have no opinion on this.

Also using haste is bad with imperil. I recently said that maybe it wasn't (shortsword test) but since then I learned the shortsword bleeding wound is very powerful, so now I once again believe imperil is quite damaging to haste usage.

However, there might be a way to reduce this incompatibility: do not cast Imperil on the first few turns of each round. This needs more investigation so I don't want to focus too much on it yet. Also for regular arenas, you must cast Imperil first anyway. But for school girl arenas, I'm not sure.

Your 3700 turns result might have been real too, who knows. It's a bit extreme but sometimes you can get a very different result in DwD depending how you play, or from luck. It could be luck on how many Konatas appeared. If you close your game window and open it again, your Monsterbation will forget those turns, so that could be another possibility?

edit: I figured out why your turns/second was poor with Haste. It's partially luck, but also because manually casting Haste wastes a ton of time and drops your turns/second. (It's the same reason why players who need to cure a lot in grindfest/IW will gain huge benefit from haste usage, because they don't need to cure and waste time).

If you put haste on IA you will save 17 turns and probably 50 seconds. Another option, because haste increases your defenses immensely, when using haste you can stop casting spirit shield (this is what I did in almost all my haste tests, if you check) and at your level, you should not need spark of life either. I recommend you test using protection+haste vs protection+spirit shield+spark

This post has been edited by BlueWaterSplash: May 20 2020, 11:00

I don't have any opinion on the usefulness of haste in school girl arenas because there are too many different ways to play them (targeting strategy, skill usage, imperil usage, etc).

However looking at your data, it suggests that Haste is neutral for you during school girl arenas. Time is what counts (using haste is supposed to boost turns/second, but not in school girl arenas) but your turns/second can vary with connection luck and the day, so in this case you should be comparing turns.

My haste test results are also subject to issues of luck which is why I don't insist that my turns/second results are definitive. I tried to reduce the randomness by having a wide array of different kinds of tests to cross check and confirm if turns/second boost exists on the average (which it did, maybe).

I thought about that too. In general, the variations were small. (IMG:[invalid] style_emoticons/default/sleep.gif)

About "targeting strategy", I kept using the same one as before, randomly hitting the monsters at the beginning, and in the later rounds focusing first on the SGs. About "skill usage", I am used to using OFC only (on average between 60 and 65 times; reload and shoot). I use Vital Strike only on dragons (I've tried to use it in SGs, but I fumble of the OC's recharge and end up wasting time thinking about what to do at the moment (IMG:[invalid] style_emoticons/default/heh.gif) ). And about "imperil usage", I keep using it all the time (you can see it by the large number of times that I triggered it).

Also I can notice a gradual improvement in your turns/second on all the later days, as you get more used to using script. This is not related to haste usage.

Seeing the table, after it was ready, I also thought about it. Although yesterday's result was almost the same as that shown in the table (day 12), very little better. All therefore within a certain margin of error. (IMG:[invalid] style_emoticons/default/sleep.gif)

4175/4208 = only 0.8% difference. You might as well therefore use Haste. Actually, because you don't have haste on IA, you wasted 17 turns to cast Haste. If you give back these 17 turns, the difference is only 0.4%

I think not because, using Haste, I used the other supportives less often. One thing made up for the other.

With Haste: 19 Regen, 12 Protection, 18 Spirit Shield, 16 Spark of Life, 6 Heartseeker
No Haste: 26 Regen, 15 Protection, 28 Spirit Shield, 22 Spark of Life, 7 Heartseeker

However, there might be a way to reduce this incompatibility: do not cast Imperil on the first few turns of each round. This needs more investigation so I don't want to focus too much on it yet. Also for regular arenas, you must cast Imperil first anyway. But for school girl arenas, I'm not sure.

Maybe I'll try to do that next Monday (Imperil only in SGs, and also a run without Imperil).

Your 3700 turns result might have been real too, who knows. It's a bit extreme but sometimes you can get a very different result in DwD depending how you play, or from luck. It could be luck on how many Konatas appeared. If you close your game window and open it again, your Monsterbation will forget those turns, so that could be another possibility?

I believe there was a mistake. There must have been a slight lag in the connection that caused a turn count error.

edit: I figured out why your turns/second was poor with Haste. It's partially luck, but also because manually casting Haste wastes a ton of time and drops your turns/second. (It's the same reason why players who need to cure a lot in grindfest/IW will gain huge benefit from haste usage, because they don't need to cure and waste time).

If you put haste on IA you will save 17 turns and probably 50 seconds.

Someday I buy IA. (IMG:[invalid] style_emoticons/default/tongue.gif)

Another option, because haste increases your defenses immensely, when using haste you can stop casting spirit shield (this is what I did in almost all my haste tests, if you check) and at your level, you should not need spark of life either. I recommend you test using protection+haste vs protection+spirit shield+spark

Indeed, the use of Haste decreases the amount of damage suffered.

Total damage taken
11/05/2020 (Mon): 　1.306.425 (with Haste)
12/05/2020 (Tue): 　1.692.542
13/05/2020 (Wed): 　1.865.286
14/05/2020 (Thu): 　1.669.300
15/05/2020 (Fri): 　1.914.719
16/05/2020 (Sat): 　1.811.344
17/05/2020 (Sun): 　1.888.760
18/05/2020 (Mon): 　1.736.032

Still, I don't think I'm going to take this test that you recommended. Without Spirit Shield the damage suffered is greatly increased.

This post has been edited by Basara Nekki: May 20 2020, 18:45

I didn't even notice that you used a full imperil style. That was horribly damaging to haste effectiveness. Even 1 imperil style would cause harm. Now my guess is that your haste may use less turns with a no imperil comparison.

Recent run.
Non bonus day, finally splurging for DD5 and the last IA slot.

For the first time doing in less than 25 minutes (05/06/2020):

(IMG:[imgur.com] https://imgur.com/fNlGA70.jpg)

And for the first time, playing under 4000 turns (12/06/2020):

(IMG:[imgur.com] https://imgur.com/BIRLGs5.jpg)

I think it will be difficult to improve these results. (IMG:[invalid] style_emoticons/default/heh.gif)

DW Shade (Club + Rapier) w/ DD2

1h dd1

One of my best for turns, but worse t/s than normal due to internet being crap during it:


A bit of an opposite to the other day, almost 3 t/s (so close), but considerably worse turns.


This post has been edited by Ubershank: Jun 14 2020, 01:34

Trying 1H Imperil, it uses more drought but significantly faster than plain 1H.