[X] My Assistant

This thread is for all research, experiments, and hypotheses testing on how HV really works.

if you want specific posts moved to this thread, please provide the link to that post and I will do my magic.

This post has been edited by DJNoni: Oct 4 2019, 02:04

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, 14:43

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, 14:48

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, 16:04

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, 22:02

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.

So it's pretty close, at least for now.
It's gonna be nice to see more samples, I'd prefer data from a different level player than different hovering patterns. As you're pretty much the same level as me.

As for hovering patterns, take note that heavy OFC usage somewhat equalizes them. Around let's say a third of the time you're not even attacking.
My pattern is to sweep around trying to make everyone die nearly at the same time, potentially from counter-attacks. This is to not waste potential counter-attacks at an already dead monster. But the difference in pattern effectiveness is probably miniscule.

-snip-

Interesting results. But I would like to know a few things.

1) From what you have described, you have used supportive spells a few times. You basically does most of the early rounds only using Regen. When do you start using others? In my case I am used to using everything (except Haste and Shadow Veil) in all rounds (ending the effect I trigger again).

2) If you used Haste constantly (for example, putting in an Innate Arcana slot), what exactly would happen? Obviously the number of turns would be higher, but what about the total time? Since you only used it one time in most tests, I think its influence on the results was small.

3) Regarding the absence of OFC, why haven't you bought the missing Figurines yet? Are you trying to get them all on your own? It will take a long time. It took me over 3 years to achieve this. Imagine all this time without OFC? Good thing Figurines are tradeable items.