[X] My Assistant

In theory Swift Strike should be just as good as Haste but past tests suggested to me that maybe Swift Strike would never increase the duration of Penetrated Armor and Overwhelming Strikes because the amount is not enough to reach the difference threshold. Today I again tested Swift Strike (presumably, since I have 1.7% action speed bonus) without Haste to make things easier.

Swift Strike does seem to occasionally increase the duration of all ailments. When the duration of my inflicted ailments randomly increased at the start of rounds, they typically increase together. I've seen PA, Overwhelm, Stun, and Burn all simultaneously inflict with +1 duration. However if inflicted again later in the same round the duration would not be increased again. It appears there is some sort of timing jitter algorithm that is not described on the Action Speed section of the wiki, in order to keep speed bonuses fully functional at low amounts. Monsters were known to jitter, but not the player.

Swift Strike may be even better than Haste because the durations increased much more often than I would expect from the math, so there may be quirks in the jitter algorithm that favor low amounts of speed bonus. I also think I noticed that Overwhelming Strikes may get +1 duration, or perhaps a random fraction of a turn, each time a round ends (I am guessing this has nothing to do with Swift Strike).

Just for reference, the normal durations of ailments are below. I tested all these things with script turned off so I could count and analyze each turn carefully.

Penetrated Armor - 7 turns. You get to hit the enemy 7 times after inflicting PA. The "turns remaining" counter goes from 6 to 0, and you get to hit them again at 0.
Overwhelming Strikes - 4 turns. The "turns remaining" counter goes from 3 to 0. If Swift Strike gives an additional turn the counter will display 4 to 0.
Stun - 3 turns. The true duration is 4 turns just like DW club but in 1H it's inflicted when the monster hits you. The desync causes a turn to be lost in nearly all cases.
Burn - 3 turns. Same as above, the true duration is 4 turns but losing a fraction of a turn usually loses a whole turn. The counter typically displays 2 to 0.

I was surprised at how poorly my Overwhelming Strikes performed without Haste. I felt like it was inflicted only half the time, so I don't even get the +15% attack damage bonus! And on average I would only have 10% counter-parry from Overwhelming x1, this is pathetic compared to its power with Haste. All the other 1H players who don't use Haste should probably tell me I'm wrong: I only did the first two arenas today without Haste.

Though I won't claim I proved anything, right now all evidence suggests that Haste and Swift Strike are extremely beneficial for 1H style. They might be just as good as in Dual Wield or any other style!

We failed to appreciate before just how much better Overwhelming Strikes, Stun, and spread out Penetrated Armor get with Haste. If my crude estimations are accurate, using Haste gives us a 5~7% raw damage bonus just from Overwhelming Strikes being active way more often! That's already making up half the damage lost to counters in big arenas (in small arenas it fully makes up the counter damage).

Plus we get around 50% more counter-parry, roughly the equivalent of Overpower 12. Extrapolating the results from Research for 1H thread that's worth 1.3075% * 2.5 = 3.27% more hits. So altogether Haste unconditionally improves 1H by 8~9% from Overwhelming Strikes alone before considering the other factors like PA or difficulty.

If this is true then we've been doing so many things wrong. 1H can use the same stats as DW with nearly balanced Agility. SDA shields are as good as SDE shields, and we should be aiming for low burden armor. All 1H players should be using Haste in Grindfest, IW, and all arenas, except for perhaps Imperil styles (Penetrated Armor duration doesn't matter and Overwhelming Strikes is more likely to wear off anyway).

Though I won't claim I proved anything, right now all evidence suggests that Haste and Swift Strike are extremely beneficial for 1H style. They might be just as good as in Dual Wield or any other style!

Particularly in regards to this comparison: how does the increased attack speed affect your OC generation? Are you still capable of keeping up Spirit Stance infinitely, and is there an excess enough to use weapon skills in addition to this?

That question is easy to answer as I've been using Haste all my life, and so have a number of other 1H players. Unless going through a very bad stretch of luck with counter attacks, a Haste user can keep Spirit Stance up almost indefinitely with mobs of 3+ enemies. With 4+ enemies OC will accumulate enough to use skills, though of course at a lower rate compared to a non-Haste user.

For small arenas with lots of 1~2 enemy mobs your OC will drain, but if you build up enough OC before activating Spirit Stance, someone with high attack damage and Daemon Duality can usually make it through without losing Spirit Stance until the enemies increase in number. If I max my OC before activating Spirit Stance, sometimes I will run out once when doing the beginning parts of the SG arenas, but it's possible to make it through.

I repeated my previous tests again just now, doing the first two arenas again without script, this time with Haste. I wanted to check if my paltry 1.7% Swift Strike stacks with Haste.

Penetrated Armor - 10~11 turns, roughly half the time each. Though expected since Haste increases duration to 10.5 turns, I'm pleasantly surprised to discover it's not just 10 turns.
Overwhelming Strikes - 6~7 turns, with 7 turns occurring rarely. This proves Swift Strike still stacks with Haste.
Stun - 5~6 turns, with 6 turns being most common. This is mostly the result of Haste and a nice leap from 3 without it.

What bullshit is that...
While using haste you generate far less OC, that is far less OFC usage. In IW and GF, OFC makes you much faster. Without haste, you can ppl cast a OFC every 2~3 rounds while you are against 7~9 monsters.
Now with haste that's a huge loss. Will the counter-parry help you reduce turns as many as an OFC does, assuming half of the monsters are already stunned and unable to parry?

- snip-

Could you do a simple test? For example, play the 80 or 90 rounds arena, today without Haste and tomorrow with Haste, and compare the number of turns and the time.

I don't have Orbital Friendship Cannon and my logical thought process does not include that. My calculations and arguments are partially precise but parts are rough. I can't say whether Haste makes you faster or slower in the end, but what has become apparent is that Haste should or might not make you anywhere near as much slower as most people believed.

with haste :



without haste :

Fudo did the Haste comparison last year in Grindfest. His turn counts are exactly the same with or without Haste, and his clear time improved with Haste by 8.36%. Wow, looking at his data, Fudo used OFC.

My equipment isn't so great but I suppose it doesn't matter, I have become curious about the Haste comparison in arenas now. I may compare arena 80~100 tonight and tomorrow. Which regular arena has the biggest mobs of monster? I'll use the spread targeting style. Since I don't have OFC who knows what will happen, haha.

I believe things such as player level, equipment, Daemon Duality, playstyle, etc will affect the Haste comparisons, but it's complicated and I can't say which way things go. We can only try it.

He was using mag equipments, and there wasn't any information regarding his weapon, his forgings, his DDs. He was even using shadow veils (I don't know at what round he start to use shadow veils). Anyway, the record is only useful for drawing out his own conclusion, but never other player's conclusion. Nor for 1H specialists.

The clear time was due to more cures resulting in lower t/s, but not less turns. Would this ever apply to Arenas and IWs? Would this difference apply to players who have some degrees of investment on the armors?
The 8.36% is a lie that seems to be significant, but never helping giving us a valid conclusion for the vast majority. It's as misleading as some researches saying you have multiple times probability getting cancer while you are eating certain common food or having a certain habit. Please don't create these numbers to exaggerate the difference when they have no empirical meanings yet on studies and battles.

Checking Reseach for 1H in this topic, it appears that sssss2 always uses Haste at least in Grindfest. His number of counter attacks is characteristic of a Haste user (about 1:1 with his main attack) and his average counter-parry from Overwhelming Strikes appears to be about 50% like mine. I can't tell exactly because we don't know what monster's original parry is, especially those years ago.

CODE

Level 456
Butcher Lv.3 Fatality Lv.2 (Potency Tier 5)

======================================================================
        | hit   | crit  | parried | sum    | parried_rate | hit_rate  
----------------------------------------------------------------------
normal  | 3,382 | 3,641 |     716 |  7,739 |      9.2518% |  90.7486%
stunned | 6,098 | 6,327 |       0 | 12,425 |      0%      | 100.0%    
----------------------------------------------------------------------
overall | 9,480 | 9,968 |     716 | 20,164 |      3.5509% |  96.4491%

CODE

Level 456
Overpower Lv.5 (Potency Tier 5)

=======================================================================
        | hit   | crit   | parried | sum    | parried_rate | hit_rate  
-----------------------------------------------------------------------
normal  | 3,329 |  3,683 |     452 |  7,464 |      6.0557% |  93.9443%
stunned | 5,946 |  6,330 |       0 | 12,276 |      0%      | 100.0%    
-----------------------------------------------------------------------
overall | 9,275 | 10,013 |     452 | 19,740 |      2.2898% |  97.7102%

Overpower increases the real hit rate by 1.3075%. (97.7102 / 96.4491 = 1.013075)

It appears the counter-parry from Overpower stacks additively with Overwhelming Strikes. This means Overwhelming Strikes does not diminish the usefulness of Overpower aside from the times when Overwhelming x5 reaches 100% counter-parry. It's mainly Stuns that diminish the usefulness of Overpower, and it looks as though approximately 38% of spread strikes hit unstunned enemies.

We previously neglected to consider that monster Parry increases with player level. This means Overpower becomes better on a high level player. All monster species at PL 2250 and level 500 player should have exactly 19% parry without chaos upgrades on Pfudor difficulty. Using sssss2's data of 38% unstunned enemies, Overpower at level 500 should improve to 0.38 * 4.6914% = 1.7827% this is wrong, too optimistic. Edit: I made some mistakes, I wish Sapo84 were still around to help me catch these.

This post has been edited by BlueWaterSplash: Oct 8 2019, 08:43

vincento, I agree with you. That's why even though I liked Fudo's result when he posted it last year, and I did many calculations based from it (not here) I never claimed his results are for everyone, just interesting. I especially thought they would not apply in arena.

But maybe Fudo was right. Now I noticed sssss2 also used Haste. These last few days I realized more ideas and understand why Haste and Swift Strike are so much better than I previously thought. Maybe they are good in arena too. It's time for more tests while I'll hopefully try soon. My tests won't be definitive, either.

Proper research has to be done two ways. 1) the experimental approach: just try Haste vs no Haste and see, but very many people should try it, not just Fudo. 2) the logical math approach: you must use the game formulas and think of the reasons why Haste is good or bad. This is what I like, but even I cannot do it perfectly. I so far can calculate or estimate the amount of damage lost to counters (~12%) and the amounts gained back from Overwhelming Strikes (~10%) and spread PA (depends on player) but it's a rough approach. No calculation for OFC yet, maybe I could do it later after I have it. I only found 5 ponies.

Method 1) is like an Engineer's way. Method 2) is like a Mathematician's way. We must investigate both ways.

My anecdotal experience is that haste very greatly hurts my performance in arenas as 1H.

I run a forge-20 rapier with forge 20-something shield, forge 5 power armors, 3 slaughter, 1 balance, 1 savage warding. I imperil and also cast the t1-holy spell every round for proficiency gain purposes. I slowly decrease OC over time if there are 4 monsters, I stay about even with 5, and I gain OC if there are 6+.

I've always felt the biggest conclusion is pretty simple - if haste saves you from having to use cure, then you should use haste since it will both simplify your play and save you overall time since anything besides hovering or keybind spams of imperil will always make you slower. If you don't have to cure much or ever though, then haste should definitely be skipped.

My anecdotal experience is that haste very greatly hurts my performance in arenas as 1H. I imperil and also cast the t1-holy spell every round for proficiency gain purposes. I slowly decrease OC over time if there are 4 monsters, I stay about even with 5, and I gain OC if there are 6+.

Casting Imperil worsens the usage of Haste. Casting the Holy Spell will ruin it more. They use turns and cause Overwhelming Strikes to wear off. I mentioned it above: Imperil 1H styles probably shouldn't use Haste.

This also explains why you cannot keep up your OC. Casting those spells will waste 2 OC. It's great that you mentioned this though, as I didn't fully realize it before. Imperil 1H with Haste is worse than I previously thought. Might still be decent to cast Haste with large mobs though.

Come to think of it, I also cast Imperil at the beginning of rounds in the later half of arenas for proficiency gain purposes, after my mana is overflowing and in excess (I rarely or never use draughts). But I only cast 1 Imperil, and it only has 1 target, and by then there are big mobs so I never notice the OC drain.

I decided to postpone my Haste vs no Haste arena tests by one day. Today is Fire day: just to be safe I don't want to boost my Spike Shield Burn damage. The next 2 days I'll waste Holy infusions and compare. I never use infusions, but they will affect the test. I can test next week without infusions.

This post has been edited by BlueWaterSplash: Oct 8 2019, 08:48

I'll put some anecdotal evidence since I play 1H shade and that's a less often seen one. So it might add something.

At my level (~375) so far when I compare myself to other players 1H shade doesn't fall much behind in clear speed compared to 1H heavy.
Similarly I haven't felt a massive difference in clear speeds between 1H shade with haste and without haste.



I'll be lazy to collect big sample sizes of data but here is a super small one if anyone cares.
Again, this is for 1H shade.

[x30 arena, imperil spam, OFC spam]
without haste 498t 2m45s
with haste 528t 2m57s

[x50 arena, imperil spam, OFC spam]
without haste 796t 4m28s
with haste 819t 4m32s

[x60 IW, imperil spam, NO OFC used]
without haste 1360t 7m39s
with haste 1436t 8m05s

[x110 arena, imperil spam, OFC used when at 6 monsters]
without haste 2358t 13m32s
with haste 2403t 13m50s (haste usage started when the first SG appears)

-> altogether averaging 3-4% difference
-> element of the day: dark for non-haste test, fire for haste test
weapon used: cold+holy
spike shield: fire
so I think this is neutral as far as elemental bonuses goes?

About overcharge and 1H shade:
at around 4 monsters I can keep infinite spirit stance.
at around 5 monsters it goes up.
- using haste seems to need lets say 0.25 more monsters (completely ballpark estimate). At bigger monster count the overcharge generation difference seems to be more notable. Still not a massive difference though.
- imperil usage is actually more taxing on overcharge. I'd say you need 1 more monster to keep spirit stance permament when you use imperil.


Things to consider about haste:
- haste lets you heal less and healing also costs overcharge
- with haste you need to recast less draughts and spells -> they also cost overcharge
- against 3 schoolgirls (and I think it might also be true for 2 schoogirls) haste lets you stack penetrated armor way better than no-haste play. With haste you can have full penetrated armor stacks on all schoolgirls and without a need to be super attentive, too. OFC hits way better that way and this probably saves time. I personally start using haste at 2 schoolgirls. Similarly I use haste for other boss fights - the dragons or noodles.
- for ease-to-play purposes I personally start using haste at round 40 in IW (which would translate to around 300 in PFFEST); in that regard the potential clear speed is of no matter to me as I (as well as some other players) would value ease above clear speed. Whether haste will help the clear speed of IW and PFFEST or hamper it I don't know.

My two cents as well:

1. People wondering about perma-spirit stance in 1h with haste probably also use orbital friendship cannon. This has been proven multiple times to be extremely helpful for players, since it can basically gift 80-100% of a round every, uh... 2-4, depending on players' gears (lower amount for worse gears, since they will clean rounds slower).
Now, it's quite rare to be able to keep spirit stance on after having casted ofc, to the point that the time spared in casting ofc can be partly balanced by the loss in waiting for perma-spirit stance to reappear (I mean, a full OC bar) - no objective data here, just a gutfeeling. I also found that haste hinders the amount of OC that can be collected, hence the amount of cannons that can be fired

2. From my small tenure as 1h shader, shade (+ haste) hindered even more

3. Grindfest at high levels may require haste not much because it's useful, but because monsters hit so hard that even haste can help survive

4. In case of long posts, some boldings or tl;dr may help

This post has been edited by Scremaz: Oct 8 2019, 10:04

Thanks megawife, that was nice data. Though some people might not like 1H Shade, I don't believe it would invalidate your results for the Haste comparison, just shift it at most. Considering that you use Imperil, I concur your results are fairly close in speed, while the defensive benefit from Haste is 10x greater as expected.

Yup, all our results here are consistent and suggest that usage of Imperil requires you to have 1 extra monster around to keep up Spirit Stance, and 2 extra monsters if you want to cast Imperil 3+ times and/or Holy spell (!) etc.

Hey, if we end up deciding that Haste and attack speed are fine for 1H, then guess what another result would be: 1H Shade gets way better too! (perhaps with optimizations, experience, and tweaking)

isn't OC gain per turn capped, btw?

A successful counter will add OC, but only once per turn.

I forgot about this. Overcharge gain is limited to 1 from counter attacks. This means that when there are a lot of enemies, Haste usage doesn't hamper OC gain much. Even using Haste it's already more than 1 counter per turn average. So OFC might be able to be used almost as often as well, as long as the swarms are big.

Fudo's data actually reported this. His Haste run used 485 OFC and his no Haste used 482 OFC. It's exactly the same. He also did 4 runs in total and they used the exact same numbers of OFC. He's consistent. Since Grindfest is 1000 rounds he is pretty much using OFC every other round. I've never casted OFC so I cannot critique this.

every, uh... 2-4, depending on players' gears (lower amount for worse gears, since they will clean rounds slower).

For me, it is OFC every third round against bigger mobs. Haste or no haste doesn't seem to change that.
It would be really significant if haste usage for someone is a breakpoint between rounds - from 4 rounds to 3; or from 3 rounds to 2.

As for keeping perma stance after OFC, it is really easy if you meet 2 requirements: OFC at max overcharge and no straggler monsters remaining alive.
You probably don't want it to happen though as you might not get to full overcharge for your next OFC cooldown. It is awkward when that happens as you manually have to turn spirit stance off. Funnily, haste is helpful for this because it hinders your first requirement.

Casting Imperil worsens the usage of Haste. Casting the Holy Spell will ruin it more. They use turns and cause Overwhelming Strikes to wear off. I mentioned it above: Imperil 1H styles probably shouldn't use Haste.

This also explains why you cannot keep up your OC. Casting those spells will waste 2 OC. It's great that you mentioned this though, as I didn't fully realize it before. Imperil 1H with Haste is worse than I previously thought. Might still be decent to cast Haste with large mobs though.

Come to think of it, I also cast Imperil at the beginning of rounds in the later half of arenas for proficiency gain purposes, after my mana is overflowing and in excess (I rarely or never use draughts). But I only cast 1 Imperil, and it only has 1 target, and by then there are big mobs so I never notice the OC drain.

I decided to postpone my Haste vs no Haste arena tests by one day. Today is Fire day: just to be safe I don't want to boost my Spike Shield Burn damage. The next 2 days I'll waste Holy infusions and compare. I never use infusions, but they will affect the test. I can test next week without infusions.

Then according to you how to distribute Agi stat in Imperil 1H styles?

This post has been edited by acnx365: Oct 8 2019, 16:02

acnx365, for now you should probably distribute Agility lower as before in Imperil 1H styles.

I've completed the first half of my testing, with no Haste. I made a few occasional mistakes but for the most part I played well enough. I decided to do Arenas 55, 60, 65, 80, and 90. The monster distribution in different arenas isn't as complicated as it could have been. Arenas 60, 65, and 80 have the biggest swarms of monsters and are pretty much identical in distribution to each other except for being longer. Arenas 55, 90, and many others share the most common distribution.

I tweaked my normal script settings and item usage style from the usual for greater speed. I only hardcasted Protection, Spark, and Spirit Shield at the beginning of arenas and when my Heartseeker wore off. I always casted Regen manually and tried to Cure as little as possible, typically never.

I didn't even play like this for any of my DwD clear times since I'm not overly interested in speed or showing off. But I wanted to be more similar to other players for this test, and playing quicker will show more clearly if Haste makes any difference for me in turns/second during arenas. People who want to know my equipment can check that thread.

It felt to me like I had Overwhelming Strikes maybe two thirds of the time; that is still a significant time of inactivity and damage bonus loss, so I look forward to tomorrow's results. I slightly overestimated how many monsters I need to keep permanent Spirit Stance, it might be 4 monsters with Haste, but it also felt like 4 monsters without Haste. I forgot to consider that +50% counter-parry with Haste gives back some OC. In any case Spirit drains slowly enough with 3 monsters for me to rush through.

[imgur.com] http://imgur.com/XhgQLl1 [imgur.com] http://imgur.com/1P7UvVq
[imgur.com] http://imgur.com/WFxurkq [imgur.com] http://imgur.com/0jq6MBk
[imgur.com] http://imgur.com/u7uRaCx [imgur.com] http://imgur.com/d1R4qW3
[imgur.com] http://imgur.com/IDCLdma [imgur.com] http://imgur.com/urvuR2s
[imgur.com] http://imgur.com/3dDinc8 [imgur.com] http://imgur.com/ZJbjAi9

55 rounds - 840 turns 0:05:39 (2.478 t/s) +2.4% faster, +2.9% quicker (Attack: 818, Counter: 954, Protection: 2, Spark: 2, Shield: 2, Heartseeker: 2, Regen: 7)
60 rounds - 1,015 turns 0:07:14 (2.339 t/s) +2.6% faster (Attack: 981, Counter: 1279, Protection: 2, Spark: 2, Shield: 2, Heartseeker: 4, Regen: 9, Cure: 3, Mana Draught: 1, Mana Potion: 1)
65 rounds - 1,128 turns 0:08:03 (2.335 t/s) +1.0% faster (Attack: 1092, Counter: 1408, Protection: 4, Spark: 4, Shield: 4, Heartseeker: 3, Regen: 9, Cure: 2, Mana Draught: 2)
80 rounds - 1,386 turns 0:09:53 (2.337 t/s) +2.5% faster (Attack: 1339, Counter: 1759, Protection: 4, Spark: 5, Shield: 5, Heartseeker: 4, Regen: 12, Cure: 1, Mana Draught: 1, Spirit Draught: 1)
90 rounds - 1,331 turns 0:09:01 (2.46 t/s) +3.2% faster, +0.7% quicker (Attack: 1293, Counter: 1499, Protection: 3, Spark: 4, Shield: 4, Heartseeker: 4, Regen: 10, Cure: 2, Mana Draught: 1)

55 rounds - 860 turns 0:05:49 (2.464 t/s) (Attack: 844, Counter: 704, Protection: 1, Spark: 1, Haste: 1, Heartseeker: 2, Regen: 5)
60 rounds - 1,041 turns 0:06:55 (2.508 t/s) +4.6% quicker (Attack: 1018, Counter: 954, Protection: 1, Spark: 1, Haste: 1, Shield: 2, Heartseeker: 2, Regen: 7, Cure: 1)
65 rounds - 1,139 turns 0:07:38 (2.487 t/s) +5.5% quicker (Attack: 1121, Counter: 1031, Protection: 1, Spark: 1, Haste: 1, Heartseeker: 2, Regen: 8, Cure: 1)
80 rounds - 1,420 turns 0:09:20 (2.536 t/s) +5.9% quicker (Attack: 1392, Counter: 1319, Protection: 1, Spark: 1, Haste: 1, Shield: 1, Heartseeker: 3, Regen: 8, Cure: 3)
90 rounds - 1,373 turns 0:09:05 (2.519 t/s) (Attack: 1342, Counter: 1158, Protection: 2, Spark: 2, Haste: 2, Heartseeker: 3, Regen: 9, Cure: 2)

Overall for my 1H no-Imperil Rapier style with Infusion at level 378, without Haste was 5833 / 5700 = 2.3% faster in turns while using Haste was (339 + 541 + 434 + 483 + 593) / (349 + 545 + 415 + 458 + 560) = 2.7% quicker in real world time due to higher turns/second. Arenas with the largest swarms seem better suited to Haste styles; Grindfest and IW have bigger ones.

A level 500 player will encounter higher monster PMit thus using Haste may improve their offense even more with spread playstyles. Individual targeting styles should be worse but it won't be too bad because the Overwhelming Strikes difference is a constant for everyone.

Haste improved turns/second despite my poor internet connection. You don't have to cast Heartseeker and Regen as often and use less items. In Grindfest and IW you'll need to Cure less often. This difference will be less on level 500 players and greater on those with fast connections.

I believe hardcasting buffs to channel Heartseeker saves a bit of mana but at high level it eventually becomes pointless. If you only cast Heartseeker and Regen your turns/second won't change as much. It's possible that Haste improves turns/second through other means: perhaps with fewer counters enemies die more evenly which could save mouse cursor time.

It's possible that Haste improves turns/second through other means: perhaps with fewer counters enemies die more evenly which could save mouse cursor time.

Now that I think further, this is significant. Without Haste the only thing that improves is counters, which kill monsters a turn sooner once in a while. Assuming you spread, some premature deaths will occur in random locations near the end of a round. If the player fails to see and react to this, and sweeps over the dead monster without speeding up the mouse cursor, that dead monster may as well have still been alive.

The few gained turns from not using Haste are more apt not to be utilized by sweeping players in real world time.

This could be tested by having players compare Haste versus no Haste using individual targeting styles: any turns/second difference should arise purely from casting. I've seen End of All Hope and a few other high level 1H players say they target monsters individually.

We still don't know if Haste will speed up or slow down players who target individually. They should get less turns/second boost and won't benefit from improved PA duration, but can benefit from improved Stun duration instead, and will benefit from Overwhelming Strikes the same. Meanwhile, their counter attacks are nearly irrelevant.