[X] My Assistant

Up until now I've ignored 1H stun duration because I used to think it was synced to monsters' timelines, but now I realize it works similarly to your other ailments. Stuns inflicted by 1H counters are set to have 1 less duration than those inflicted by a club. Spike shield ailments have the same base duration of 3 turns as 1H stuns. Here is the performance of stun with haste.

CODE

  turn   Stun or Burn   "turns remaining" counter
  -----------------------------------------------
   01      5 turns               3 to 0
   02      4 turns               2 to 0
   03      5 turns               2 to 0
   04      5 turns               3 to 0
   05      4 turns               2 to 0
   06      5 turns               2 to 0

Haste increases stun duration to 4.66 turns following the same pattern as PA. Next comes stun with haste and 1.5% attack speed bonus. It likewise follows a similar pattern.

CODE

  turn   Stun or Burn   "turns remaining" counter
  -----------------------------------------------
   01      6 turns               3 to 0
   02      5 turns               2 to 0
   03      5 turns               3 to 0
   04      4 turns               2 to 0
   05      5 turns               2 to 0
   06      5 turns               3 to 0
   ...
   43      4 turns               2 to 0
   44      5 turns               2 to 0
   45      5 turns               3 to 0
   46      4 turns               2 to 0
   47      5 turns               2 to 0
   48      6 turns               3 to 0
   49      5 turns               2 to 0
   50      5 turns               2 to 0
   51      6 turns               3 to 0
   52      5 turns               2 to 0
   53      5 turns               3 to 0

I recently realized evade is not nearly as hurtful as I previously thought. This is because when an enemy evades an attack he can't get stunned, so he soon attacks again and the counter that was lost will quickly be made back. Haste is the same as before because it increases the length of stuns.

Low amounts of attack speed are decoupled from the increase in stun duration, so we now need to consider them separately. This new viewpoint can also be applied to haste. The drop in counters is only 1/4 due to action speed itself, while the remaining 3/4 is due to the increase in stun duration. This is because monsters don't move for 3 turns after being countered and stunned.

Let's break down haste more carefully. In arenas, the blame for (2.4 + 1.17 * 0.75) / (2.4 + 0.85 * 0.75) ≈ 8% loss in damage due to counters is reassigned: 2% for the +50% action speed, and 6% for increasing stun duration. In grindfest, the blame for (2.4 + 1.75 * 0.75) / (2.4 + 1.25 * 0.75) ≈ 11% damage loss is reassigned to 2.75% from speed and 8.25% from stun.

Simplifying my haste comparison table, let's say haste loses 3% turns and makes back 3% in turns/second boost, for zero loss in arenas. Of the 5% turn gains, I'll approximate that 1.66% comes from more reliable Overwhelming Strikes damage boost, 1.66% from Overwhelming Strikes counter-parry, and 1.66% from better Penetrated Armor duration.

In pfudor grindfest, let's say haste loses 6% turns and makes back 6% in turns/second boost, for zero loss. I've tweaked the numbers to be more skeptical of haste than my personal results. Of the 5% turn gains, I'll again guess that 1.66% comes from Overwhelming Strikes damage, 1.66% from Overwhelming Strikes counter-parry, and 1.66% from Penetrated Armor.

Stun also has positive effects because stunned enemies can't parry. Assuming a 6% parry rate overall, 1 turn of added stun is worth 2.4 * 0.06 ≈ 0.15 more damage, yet also loses 0.75 * 1/5 = 0.15 damage from counters. The chance you target that stunned enemy is divided by the number of enemies though, so the damage loss from counters is much greater.

This positive effect of stun doesn't enter the haste breakdown for spread styles because unless you target enemies intelligently, increased action speed causes you to hit unstunned enemies more often, too. It could play a minor role for swift strike since its stun durations increase via timing glitches rather than your attack speed.

This post has been edited by BlueWaterSplash: Jun 3 2020, 21:40

Monsters are getting stronger and stronger.
- Players are feeding their monsters crystals (and chaos tokens), and thus the average power level of monsters in battle is increasing.
https://forums.e-hentai.org/index.php?s=&am...t&p=5614254

And they are also giving more and more experience.
- This is because the monster's power level (and stats) affects the amount of experience.
https://ehwiki.org/wiki/Experience_Points#Monsters


The table below shows the experience of PFUDOR grindfest that I have recorded for four years.
- base_exp is the exp for which star bonus and cookie bonus are removed: base_exp = average_exp / (1 + star_bonus) / (1 + 20% * cookies)
- Other conditions are the same: PFUDOR difficulty, no stamina bonus, maxed out Adept Learner training and hath perks, forum posting bonus, no toplist bonus
* I found that 5th cookie, Different Chocolate Chip Cookie, had not given any exp bonus in 2018.

- The current base_exp is 20% higher than four years ago. (985m > 1184m)
- The overall exp that now I gain is more than four times as much as four years ago. (1379m > 5685m)

CODE

================================================================
          |    average_exp |  stars |  cookies |       base_exp
================================================================
  2016_07 |  LEVEL 500 WITHOUT A STAR
----------------------------------------------------------------
  2016_08 |  1,379,660,490 |        |        2 |    985,471,778
  2016_09 |  1,384,525,069 |        |        2 |    988,946,478
  2016_10 |  1,397,139,877 |        |        2 |    997,957,055
  2016_11 |  1,408,365,646 |        |        2 |  1,005,975,461
  2016_12 |  1,416,848,652 |        |        2 |  1,012,034,752
----------------------------------------------------------------
  2017_01 |  1,627,002,021 |        |        3 |  1,016,876,263
  2017_02 |  1,634,430,681 |        |        3 |  1,021,519,176
----------------------------------------------------------------
  2017_02 |  GOT A GOLDSTAR AND ONE MORE COOKIE
----------------------------------------------------------------
  2017_02 |  3,686,392,872 |      1 |        4 |  1,023,998,020
  2017_03 |  3,692,923,641 |      1 |        4 |  1,025,812,123
  2017_04 |  3,703,871,332 |      1 |        4 |  1,028,853,148
  2017_05 |  3,717,243,971 |      1 |        4 |  1,032,567,770
  2017_06 |  3,739,391,365 |      1 |        4 |  1,038,719,824
  2017_07 |  3,775,338,647 |      1 |        4 |  1,048,705,180
  2017_08 |  3,795,221,950 |      1 |        4 |  1,054,228,319
  2017_09 |  3,814,906,013 |      1 |        4 |  1,059,696,115
  2017_10 |  3,831,057,441 |      1 |        4 |  1,064,182,622
  2017_11 |  3,861,367,307 |      1 |        4 |  1,072,602,030
  2017_12 |  3,880,677,376 |      1 |        4 |  1,077,965,938
----------------------------------------------------------------
  2018_01 |  5TH COOKIE HAD NOT GIVEN EXP BONUS IN 2018
----------------------------------------------------------------
  2018_01 |  3,913,920,971 |      1 |      * 4 |  1,087,200,270
  2018_02 |  3,921,225,680 |      1 |      * 4 |  1,089,229,356
  2018_03 |  3,934,943,612 |      1 |      * 4 |  1,093,039,892
  2018_04 |  3,955,456,544 |      1 |      * 4 |  1,098,737,929
  2018_05 |  3,974,427,712 |      1 |      * 4 |  1,104,007,698
  2018_06 |  3,987,403,332 |      1 |      * 4 |  1,107,612,037
  2018_07 |  3,989,709,727 |      1 |      * 4 |  1,108,252,702
  2018_08 |  3,987,794,706 |      1 |      * 4 |  1,107,720,752
  2018_09 |  3,992,777,858 |      1 |      * 4 |  1,109,104,961
  2018_10 |  4,003,081,334 |      1 |      * 4 |  1,111,967,037
  2018_11 |  4,008,024,387 |      1 |      * 4 |  1,113,340,108
  2018_12 |  4,019,531,469 |      1 |      * 4 |  1,116,536,519
----------------------------------------------------------------
  2019_01 |  4,908,824,976 |      1 |        6 |  1,115,642,040
  2019_02 |  4,927,276,997 |      1 |        6 |  1,119,835,681
  2019_03 |  4,954,923,268 |      1 |        6 |  1,126,118,925
  2019_04 |  4,968,082,933 |      1 |        6 |  1,129,109,758
  2019_05 |  4,974,514,483 |      1 |        6 |  1,130,571,473
  2019_06 |  4,984,790,876 |      1 |        6 |  1,132,907,017
  2019_07 |  5,002,526,553 |      1 |        6 |  1,136,937,853
  2019_08 |  5,005,667,782 |      1 |        6 |  1,137,651,769
  2019_09 |  5,009,355,425 |      1 |        6 |  1,138,489,869
  2019_10 |  5,015,413,716 |      1 |        6 |  1,139,866,754
  2019_11 |  5,036,449,820 |      1 |        6 |  1,144,647,686
  2019_12 |  5,056,036,591 |      1 |        6 |  1,149,099,225
----------------------------------------------------------------
  2020_01 |  5,533,755,376 |      1 |        7 |  1,152,865,703
  2020_02 |  5,580,294,887 |      1 |        7 |  1,162,561,435
  2020_03 |  5,610,535,723 |      1 |        7 |  1,168,861,609
  2020_04 |  5,658,872,609 |      1 |        7 |  1,178,931,794
  2020_05 |  5,685,646,936 |      1 |        7 |  1,184,509,778
================================================================

This post has been edited by sssss2: Jun 1 2020, 17:46

...

Was the 2018 cookie fixed in 2019?

Monsters are getting stronger and stronger.
- Players are feeding their monsters crystals (and chaos tokens), and thus the average power level of monsters in battle is increasing.
https://forums.e-hentai.org/index.php?s=&am...t&p=5614254

And they are also giving more and more experience.
- This is because the monster's power level (and stats) affects the amount of experience.
https://ehwiki.org/wiki/Experience_Points#Monsters
The table below shows the experience of PFUDOR grindfest that I have recorded for four years.
- base_exp is the exp for which star bonus and cookie bonus are removed: base_exp = average_exp / (1 + star_bonus) / (1 + 20% * cookies)
- Other conditions are the same: PFUDOR difficulty, no stamina bonus, maxed out Adept Learner training and hath perks, forum posting bonus, no toplist bonus
* I found that 5th cookie, Different Chocolate Chip Cookie, had not given any exp bonus in 2018.

- The current base_exp is 20% higher than four years ago. (985m > 1184m)
- The overall exp that now I gain is more than four times as much as four years ago. (1379m > 5685m)

(IMG:[invalid] style_emoticons/default/huh.gif) So it is not just my idea, I am really taking longer to finish certain arenas despite the fact that I am continually improving my equipment.

I think it already shows too much the fact that the Monsters are not balanced, because it has happened to me sometimes that a common monster needs more attacks to be defeated than one of the school girls, which is already ridiculous, , that they shouldn't are those monsters on a higher level than I am currently??

I don't know how it works for those who are at level 500, but for those who are still below level 400-410, having to spend millions on better equipment to face monsters that already feel too unbalanced is a bit unfair. Not that I'm complaining; But I wonder how will those who have not had the luck to earn 20 million credits thanks to a drop be overcoming the situation?

This post has been edited by Arkoniusx: Jun 2 2020, 12:02

(IMG:[invalid] style_emoticons/default/huh.gif) So it is not just my idea, I am really taking longer to finish certain arenas despite the fact that I am continually improving my equipment.

I think it already shows too much the fact that the Monsters are not balanced, because it has happened to me sometimes that a common monster needs more attacks to be defeated than one of the school girls, which is already ridiculous, , that they shouldn't are those monsters on a higher level than I am currently??

I don't know how it works for those who are at level 500, but for those who are still below level 400-410, having to spend millions on better equipment to face monsters that already feel too unbalanced is a bit unfair. Not that I'm complaining; But I wonder how will those who have not had the luck to earn 20 million credits thanks to a drop be overcoming the situation?

What you are observing is probably nothing to do with chaos tokens and monster PL, which is a very long term, slow effect (on the scale of years for any noticeable effect). In your case, it's just because you're levelling up, and monsters get increasingly sized health pools. The game slows down as you level, and level 500 players feel it the worst.

However, it's not really that expensive to forge melee gear over time... almost nobody gets that one lucky drop that funds a huge upgrade for them. Income for arena-only melee players right now is relatively high compared to expenses; trophy prices are high, scrolls are sellable, as are most consumables, gum & vase & amnesia shards are sellable... etc. 20m is not a very large sum of money anymore.

Also - your weird build insisting on shitty weapons probably doesn't help you at all. As long as you're still having fun, it's okay, but... most people don't realise how much better the game can feel, and burn out trying to be special. I wonder how many players switched to mage - geared and forged enough to be worth it, of course - and then switched back to melee? (IMG:[invalid] style_emoticons/default/tongue.gif)

This post has been edited by lestion: Jun 2 2020, 12:09

I'll try to estimate how swift strike and attack speed reflect on offense by studying the turn-by-turn ailment durations, scaling those parts individually to haste, then using the previously explained breakdown for haste. We first study attack speed 1.0% with the case of ordinary attacking in 1H non-imperil rapier style.

For arenas we simplify by considering them as mobs of 5, with each enemy taking 3 hits to kill. Inflicted PA duration is 8 on the first turn, then 7 for the next six, for an average of 7.142857 which is exactly 2 times what is deserved. PA durations of the last 8 turns are irrelevant since the round will end anyway. Scaling to the 10.66 duration of haste gives 0.142857 * 1.66% / 3.66 = 0.064935% boost to offense.

I'll take the added turns of Overwhelming Strikes duration to be 3 over the course of the 15 turn round, so 4.2 average (which is 5 times the deserved boost). Scaling to the 6.33 duration of haste gives 0.2 * 3.33% / 2.33 = 0.285714% estimated boost to offense from both the added damage and counter-parry.

I'll guess 3 counter attacks and 2 lengthened stuns are inflicted on the first turn, compared to 1.3 counters per turn for 19.5 total and 13 stuns overall. Average stun duration is 3.153846 which is over 5 times the deserved penalty. Scaling to the 4.66 duration of haste gives 0.153846 * 6% / 1.66 = 0.553846% which I'll reduce by 1/5 to 0.443077% penalty after considering the side benefits of stun.

The direct impact of a 1% loss in counter attacks is 2% / 50 = 0.04% penalty, similar to evade. Actually this component should be marginally better than evade, but it's not worth considering the difference. For small amounts of attack speed this effect is less than extrapolated due to the counters limit, so it's feasible to choose either 0.04% or 0% penalty.

In arenas, haste displayed a turns/second boost of 3% of which I'll guess 1.5% comes from spread wasting and 1.5% from minimal casts. In the current case we have (0.553846% + 0.04%) * 1.5% / 8% = 0.111346% time bonus from spread wasting, and 1/15 * 100% * 1.5% / 50% = 0.2% from casts. Recall that low amounts of attack speed have a huge effect on mana savings.

The overall impact on offense for 1% attack speed is thus 0.064935% + 0.285714% - 0.443077% - 0.04% = -0.132428% or about a third of a butcher in penalty by turns. Scaled to haste this is over twice as bad. However we might have 0.2% + 0.111346% - 0.132428% = 0.178918% benefit to time, which is half a butcher.

In summary, this type of comparison is quite difficult but perhaps the only way to go about it. Errors in our approach can change the result. If you are skeptical, it's safest to consider attack speed to be offensively neutral in this case. It's not exactly a scaled down haste, but rather a collection of timing glitches which range from 2~5 times as potent as they should be.

The above calculation also applies to any amount of attack speed up to 7% because any added speed beyond the initial 1% has a minimal effect similar to evade, as none of the ailment durations are affected. The penalty is the aforementioned 0%~0.04% lost turns per added point of attack speed.

This post has been edited by BlueWaterSplash: Jun 4 2020, 09:00

I repeat the attack speed 1.0% calculation for grindfest, then with haste active at ~1.5% attack speed in arenas and grindfest.

Grindfest is treated as mobs of 9, with each enemy taking 3 hits to kill. Inflicted PA duration is 8 on the first turn, then 7 for the next eighteen, for an average of 7.05263, which corresponds to 0.05263 * 1.66% / 3.66 = 0.023923% boost to offense. Added turns of Overwhelming Strikes are taken to be 3 over the course of the 27 turn round, so 4.11 average, which becomes 0.11 * 3.33% / 2.33 = 0.15873% estimated boost to offense.

I'll guess 3 counter attacks and 2 lengthened stuns are inflicted on the first turn, compared to 1.75 counters per turn for 31.5 stuns overall. Average stun duration is 3.063492 which gives 0.063492 * 8.25% / 1.66 = 0.314286% which I'll reduce by 1/9 to 0.279365% penalty due to side benefits of stun. The direct impact of a 1% loss in counter attacks is 2.75% / 50 = 0.055% penalty or less.

For grindfest, haste is expected to have a turns/second boost of 6% of which I'll guess 4% comes from spread wasting and 2% from casts. In the current case we have (0.314286% + 0.055%) * 4% / 11% = 0.134286% time bonus from spread wasting, and 1/27 * 100% * 2% / 50% = 0.148148% from casts.

The overall impact for 1% attack speed is 0.023923% + 0.15873% - 0.279365% - 0.055% = -0.151712% penalty in turns and 0.148148% + 0.134286% - 0.151712% = 0.130722% benefit to time. This calculation also applies to 2% and 3% attack speed.

With haste in arenas, I saw 1.66 extra turns of PA duration at the beginning of rounds. Only the first 5 turns matter since the remaining have enough hasted PA duration to reach the end, for 1.66 / 5 = 0.333 average added PA duration. This corresponds to 0.333 * 1.66% / 3.66 = 0.151515% boost to offense.

Overwhelming Strikes has more subdued performance with haste, with just 1.5 added turns including the carryover effect, for 4.1 duration average. This gives 0.1 * 3.33% / 2.33 = 0.142857% boost to offense.

Enemies inflicted in each of the first 2 turns receive more stun, which I'll total to 3.5 added turns. Subsequent turns give back 0.333 so it becomes 3.166 added duration. Overall hasted counters are 0.95 per turn for 15 turns, totaling 9.5 stuns. The average stun duration increases by 3.166 / 9.5 = 0.333 turns, for a 0.333 * 6% / 1.66 = 1.2% loss which I'll reduce by 1/5 to a 0.96% penalty due to side benefits of stun.

The direct impact of a 1.515% loss in counter attacks is 1.515 * 2% / 50 = 0.0606% penalty. Though since there are 1.5 times fewer counters with haste on to begin with, I'll reduce this to 0.0606% / 1.5 ≈ 0.04% again. Likewise I think it's appropriate to reduce the stun penalty with haste on by a factor of 1.5, changing the prior results to 0.8% and 0.64% turn losses due to stun.

Turns/second boost is (0.8% + 0.04%) * 1.5% / 8% = 0.1575% time bonus from spread wasting, and 0.33/15 * 100% * 1.5% / 50% = 0.066% from casts.

The overall impact for 1.5% attack speed with haste on is 0.151515% + 0.142857% - 0.64% - 0.04% = -0.385628% penalty in turns, and 0.066% + 0.1575% - 0.385628% = -0.162128% penalty to time. This result also holds for attack speed 3.1% with haste, while attack speed 4.7% will move beyond the small timing glitches and function similarly to a scaled down haste.

With haste in grindfest, average added PA duration is 1.66 / 17 = 0.098 which corresponds to 0.098 * 1.66% / 3.66 = 0.044563% boost to offense. Overwhelming Strikes gives 0.0555 * 3.33% / 2.33 = 0.079365% boost to offense.

Enemies are again stunned for a total of 3.166 added duration. Overall hasted counters are 1.25 per turn for 27 turns, totaling 22.5 stuns. The average stun duration increases by 3.166 / 22.5 = 0.14074 turns, for 0.14074 * 8.25% / 1.66 = 0.69666% loss, which I'll reduce by 1/9 to a 0.61926% penalty. The direct impact of a 1.515% loss in counter attacks is 1.515 * 2.75% / 50 / 1.5 ≈ 0.055% penalty.

Turns/second boost is (0.46444% + 0.055%) * 4% / 11% = 0.18888% time bonus from spread wasting, and 0.33/27 * 100% * 2% / 50% = 0.04888% from casts.

The overall impact for 1.5% attack speed with haste on is 0.044563% + 0.079365% - 0.41284% - 0.055% = -0.343912% penalty in turns, and 0.04888% + 0.18888% - 0.343912% = -0.106134% penalty to time. I don't think there are any deep meanings behind these results as they are just the result of underlying small number timing glitches and luck with how they applied.

This post has been edited by BlueWaterSplash: Jun 11 2020, 21:30

Unless you care about exact formulas, you can disregard this post. The only interesting things I learned were that cast speed is affected by proficiency (which is on the wiki, albeit a bit well hidden, but nobody I spoke to knew about it) and that Scroll of Swiftness is actually the 60% it says it is, even though the last patch note that mentioned it claimed it was raised to 75%.

So, basically I learned nothing. But we have a formula for it now!

Many months ago I set out to find out whether the time resets to 0 when we enter new rounds. My thinking was that if it didn't reset, cast speed suddenly becomes very valuable for mages.

But to work that out, I needed to model the exact time units for every action. The formula I'd been using previously, to work out breakpoints for mage cast speed (where you have enough speed to avoid a whole monster attack), was very similar to BlueWaterSplash's approach, and didn't work in reality.

The more research I did, the more confused I got. It didn't seem to be multiplicative - the multiplier I would've been missing varied by test. It took a guest appearance from Skillchip to finally find the factor I was missing.

It was actually noted in the wiki all along. And in a patch note I'd missed, too (my searches for 'time units', 'action time', 'action speed', 'cast speed' and 'attack speed' didn't uncover it, because it was 'cast time'). Proficiency affects cast speed.

But I can now update the action speed page on the wiki to properly reflect what's going on internally, and it might be a little easier to digest.

The exact formula is as follows:

CODE

Time units = minmax(20, 10000/action speed, 500)
Action speed = 100 / (1 - cast speed) / specific action speed / prof factor speed mult * (1 + haste)

The specific action speed multipliers are mentioned in a few ability tooltips (faster imperil changes cast time to 0.6, for example) and others are derived in the past and found on the wiki action speed page currently.

The prof factor is a separate calculation. It uses a scale between 0-25% based on the spell's minimum requirement to use, a spell-specific cap, and your proficiency with that magic school. These requirements and caps are found on the wiki too - the spells page has that 'min/max prof' column. (By the way, this also affects mana costs by the same amount.)

The formula for that goes like this:

CODE

Prof factor = min( (effective prof - requirement) / (cap - requirement), 1 )
Prof factor speed multiplier = 1 - (0.25 * prof factor)

The following is an example for how to work out the time units spent on a tier 1 spell, using some variables from one of my tests:

CODE

Sheet speed: 14.7%
Elemental prof: 643
Specific action speed: 1
Haste: 50% (IA)
Prof factor: min((643 - 0) / (200 - 0), 1) = 1
Prof factor speed mult: 1 - (0.25*1) = 0.75

Action speed: 100 / (1 - 0.147) / 1 / 0.75 * 1.5 = 234.467~
Time units = minmax(20, 10000/234.467~, 500) = 42.65~

Finally, time units are always rounded up (not to the nearest whole - I checked). So this gives 43 time units per tier 1 spell.

How did I test this? Go into battle, put on a scroll buff, and watch the buff duration left after each cast. We know that buffs tick down after every 100 time units (this is beneficial when you start a new round and the time resets to 0 - which is what BWS was trying to explain with the 'lengthening' thing). After not many casts, you would see the calculated time units desync from when the buff actually ticks down, if it's even off by 1.

Incidentally, trying this with different variables showed that scroll of swiftness, as mentioned at the top, is actually still 60% speed. The last time it was mentioned in a patch, Tenboro said he was buffing it to 75%, but either that never happened, or it got changed back.

What does this mean about cast speed?
For imperil mages: imperil hits the speed cap of 20 time units pretty easily, so don't worry too much about it.

For non-imperil mages: T1,2,3 spells do not hit the cap (impossible - you would need 58% cast speed in addition to prof and scroll of swiftness). So it's important to keep your prof higher than 900 to speed up your T3s. This is not weighted by level, so, I guess fuck you to low-level holy/dark mages (IMG:[invalid] style_emoticons/default/tongue.gif)

For me: I need to rewrite the wiki to make this formula accessible, and the action speed page clearer in general. It's a relatively simple system internally and f4tal's tick notation seems to have confused scholars about how to calculate it.

And finally, for what it's worth, I've read over BWS's posts and they're more or less correct, in a really confusing way.

Now that we know proficiency affects the cast time of spells I can post my experimental data from the log file on the previous page. At level 388 my relevant current proficiencies are: 414.549 deprecating, 429.111 supportive, and 351.853 elemental/divine. Some increased as I performed these tests but probably not enough to make a noticeable difference.

CODE

technique                                     tested action time            base action time          minimum action time
-------------------------------------------------------------------------------------------------------------------------
Attack/Focus/Defend/Spirit Stance                    1.00                        1.00                         1.00
Cure/Protection/Absorb/Haste                         0.38                        0.50                         0.38
Shadow Veil/Spark of Life                            0.42                        0.50                         0.38
Regen/Spirit Shield                                  0.45                        0.50                         0.42
Weaken Lv1                                           0.53                        0.70                         0.53
Imperil Lv3                                          0.48                        0.60                         0.45
Imperil Lv2                                          0.52                        0.65                         0.49
Imperil Lv1                                          0.56                        0.70                         0.53
Imperil Lv0                                          0.64                        0.80                         0.60
Drain Lv5                                            0.75                        1.00                         0.75
Heartseeker                                          1.76                        2.00                         1.66
Full Cure                                            0.90                        1.00                         0.85
Freeze                                               0.90                        1.20                         0.90
Blizzard                                             1.16                        1.40                         1.05
Fimbulvetr                                           1.48                        1.60                         1.33
Smite                                                1.16                        1.40                         1.05
Banishment                                           1.44                        1.60                         1.27
Paradise Lost                                        1.70                        1.80                         1.55

Nearly all my action times were carefully tested to the exact 0.01 units with no margin of error. The wiki isn't sure about some of the min/max proficiency data but I reverse calculated the base action time of each technique, with guesswork in a few cases. In particular my data may not match the wiki proficiency range for Spark of Life.

Finally I projected the minimum action time for each technique with 600 proficiency, not counting mage equipment. The exact values may be worth verifying by a level 500 player, especially for Spark of Life, Regen, and Spirit Shield, as there are questions with their wiki maximum proficiency values.

and this is why i always recommend the supportive perk, to reduce cast time on full-cure

Now that I use a script (just a month ago), I decided to do some simple tests, to observe several things. The data are in the tables below. Sorry for the mess in the order of presentation of the results, but I was thinking about the type of test to be done only on the day itself (it was not planned in advance).

(IMG:[imgur.com] https://imgur.com/Vjnb4Ef.jpg)

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

Arena chosen: 80 rounds (because it has a good amount of monsters, minimum of 5, and also because it is quick to do).
Equipment: Power Slaughter Set (full forge)
Hath Perk: Dæmon Duality V (no other "game perk" more)
Supportive Spells in continuous use: Regen, Protection, Spirit Shield, Spark of Life, Heartseeker
Shock Spike Shield
Damage dealt: only physical
Damage taken: total

Many of the results I got are probably obvious to most of you, but since I have been using scripts for a short time, they are interesting to me. I also know that the number of tests is insufficient to affirm something with absolute certainty, therefore being just simple conclusions, and as I said, some very obvious ones.

[ A ] Imperil x NO Imperil

Comparing the tests done with Imperil (1, 2, 4, 6, 10, 13 and 14) and those without Imperil (3, 5, 7, 8, 9, 11 and 12), it is clear that the use of Imperil helps to reduce the number of turns and time (Imperil: 1191~1385 turns / 7m51s~9m26s) (NO Imperil: 1464~1807 turns / 9m52s~11m28s). In addition, the use of Imperil also increases the damage done per turn. Comparing columns 2 and 5 we have: 75.924 (Imperil) vs 61.869 (NO Imperil).

Personal note: Even though the constant activation of Imperil is more laborious and tiring, its use proved to be very beneficial to the player, with a clear reduction in turns and time. And it still gains deprecating proficiency.

[ B ] Infusion x NO Infusion

Comparing columns 1 and 2 (test with Imperil), and also columns 5 and 7 (test without Imperil), we can see that the use of infusion does not change the amount of total damage done. The use of infusion only redistributes the damage percentage for the weapon element and void damage. However, we can see that the use of the infusion helps to slightly reduce the number of turns and time, and increases the amount of damage done per turn.

Personal note: BlueWaterSplash was right. The use of infusion does not have such a noticeable effect. The gains were minimal. I always imagined it was the other way around. However, in the tests I did for the last 3 arenas (SG arenas), the use of infusion showed a more notable difference (later I present the results obtained in DwD).

[ C ] OFC x NO OFC

Comparing columns 3 and 5, we see that the use of OFC presents slightly better results than not using it (permanent spirit stance).

Personal note: Even though I had to recharge the overcharge several times, and trigger spirit repeatedly, I still saved turns. Therefore, I consider the consecutive use of OFC to be very beneficial.

[ D ] Haste x NO Haste (The test was performed without Imperil, OFC and Infusion to verify only the effect of the supportive spell.)

Comparing columns 7 and 8, we see that the use of Haste led to a small increase in the number of turns and time, but nothing very remarkable. The biggest effects were the large reduction in the number of counter attacks (2194 vs 1680), and the amount of damage taken (764.246 vs 455.577) or damage taken per turn (450 vs 265).

Personal note: The "individual use" of Haste proved to be little harmful to the 1H style. I imagined that the results would be much worse. In addition, it helped a lot in reducing damage taken. For those with defensive problems (especially low-level players), this is a very useful feature.

[ E ] Shadow Veil x NO Shadow Veil (The test was performed without Imperil, OFC and Infusion to verify only the effect of the supportive spell.)

Comparing columns 7 and 9, we see that the use of Shadow Veil practically does not cause differences in the number of turns and time (it is within a certain margin of error). There was only a small reduction in the amount of counter attacks (2194 vs 2069) and the amount of damage taken (764.246 vs 706.526) or damage taken per turn (450 vs 418).

Personal note: In the same way as in the previous test, the "individual use" of Shadow Veil also proved to be little harmful to the 1H style. However, the defensive benefits were much less than I expected. Its use is only recommended for those who have a huge defensive problem, and therefore needs all possible help.

[ F ] Combined use of Haste and Shadow Veil (The test was performed without Imperil, OFC and Infusion to verify only the effect of the supportive spell.)

Comparing columns 7 and 11, we can see that the combined use of Haste and Shadow Veil led to the worst possible result. There was a increase in the number of turns and time and a reduction in the amount of counter attacks (2194 vs 1589). Regarding the reduction in the amount of damage taken, I expected a reduction greater than that obtained, as I was using two defense spells. The displayed value was slightly worse than that obtained when only Haste was used.

Personal note: For some reason, which I can't explain, the "combined use" of Haste and Shadow Veil led to a horrible result. Therefore, it is not advisable (unless, of course, for players with defensive problems, as mentioned earlier).

[ G ] Combined use of Haste and Shadow Veil with Imperil, OFC and Infusion

Comparing columns 6 and 13, we can see that the results followed the same line as that observed in the previous comparison (item [ F ]). The benefits from the use of Imperil, OFC and Infusion have been reduced by the use of Haste and Shadow Veil.

Personal note: I'm glad I haven't used Haste and Shadow Veil in a long time. For me, defensive gains do not compensate for offensive losses.

[ H ] Use of Aether Shard

Magic Hit Chance: 104,6% -> 154,6%
Mana Cost Modifier: 162,3% -> 146,1%

Comparing columns 4 and 10, we see that the use of Aether Shard led to a slight reduction in the number of turns and time. This is probably due to an increase in the efficiency of Imperil.

Personal note: This was my best result to date for arena 80 rounds. In higher arenas I had similar (and with proportionately greater) results, that is, all of my best results were obtained using Aether Shard. Unfortunately it is not possible to repeat this every day; it would be economically bad.

[ I ] Use of Featherweight Shard (on all equipment)

Attack Speed: 3,7% -> 6,3%
Evade: 5,0% -> 28,9%
Mana Cost Modifier: 162,3% -> 129,9%
Interference: 124,7 -> 59,9
Burden: 67,6 -> 22,8

Comparing columns 7 and 12, we see that the use of Featherweight Shard resulted in a small increase in the number of turns and time, as well as a small reduction in the amount of damage done per turn. Even with an increase in attack speed and evasion, there was not a huge reduction in the amount of counter attacks and damage taken.

Personal note: I thought there would be a greater reduction in the amount of counter attacks. I also thought I would take less damage. The only advantage appears to be a reduction in mana consumption, but in terms of cost it doesn't pay (Featherweight Shard is approximately 10 times more expensive than Mana Draught).

[ J ] Using everything at the same time (Imperil, OFC, Infusion, Haste, Shadow Veil, Aether Shard, Featherweight Shard)

Attack Speed: 3,7% -> 6,3%
Evade: 5,0% -> 28,9%
Mana Cost Modifier: 162,3% -> 116,9%
Interference: 124,7 -> 59,9
Burden: 67,6 -> 22,8

Looking at the results in column 14, and comparing with the others, the main highlight is due to the large reduction in the amount of damage taken. All other parameters were in intermediate positions, depending on the comparison made.

Personal note: If someone is having problems with survival in the game, the advice I give is: use whatever is available. It is better to play comfortably (and preferably without being defeated), than to save items, turns and time. As you get stronger, you gradually become less dependent on items and supportive spells. In my case, as I don't have major defensive problems, I will continue to use only Imperil, OFC and Infusion.

PS: I do not intend to make tests without the use of Spirit Shield, because this spell is the one with the greatest defensive power (the one that holds the heaviest attacks). It would be suicide. (IMG:[invalid] style_emoticons/default/heh.gif) It doesn't make up for the time I would spend healing.

I have extensive Imperil tests done since a month ago but I'm waiting to post them until I have time, so I won't talk about it for now.

Thanks for the OFC data, I find it useful since I don't have OFC. It's a significant improvement to turns ~9% but not so much that a player can't live without it. So I don't mind waiting to get it, and I think it's best to leave OFC out of other tests, as we have been doing. OFC only had 2.5% improvement to your time so some players may not bother with it. Maybe it would have no improvement in small mob arenas.

Haste only caused a 0.5%~1% loss of turns (if you put it on IA then you would save 8 more turns) so actually it was much better for you at level 500 than it even was for me. Recall that my tests and theory suggest that infusion strongly helps Haste. And you didn't use it! So Haste with a neutral infusion might reduce your turns compared to no Haste with neutral infusion.

I think that Haste will save significant time for you in Arena 80 if you put Haste on IA. It wastes a lot of time to cast Haste, so only with it on IA will you notice its theorized turns/second boost due to less casts and also less spread wasting effect. You can leave Spirit Shield on, the important thing is to have enough IA slots to put Haste on it.

Your Shadow Veil tests are as I (recently) would have expected and exactly matched my results in offensive performance. But I didn't expect the defensive benefit would be so little. I never bothered to check my data on that, although I still have it.

I am laughing at the "horrible result" of Shadow Veil and Haste together. I also don't have any explanation, and I do not expect that. I would have said that your results so far don't match sssss2 comparison in grindfest. But actually it matches, because he also compared only Shadow Veil and Haste together.

Using full Featherweight Shard is basically a similar effect to using Shadow Veil, therefore it's correct that your number of counter attacks barely drops.

Maybe I don't fully understand the reason, but I've never used Spirit Shield in arenas, except for testing purposes. And I never get sparked and almost never heal. So either the high level player does not scale and improve his defenses as much as I expect, or maybe it's just that my continuous usage of Haste is enough.

Regarding the defense granted by Haste, it should be 50% improvement which is what we roughly see. But in practice, it depends on the quality of our weapon and DD level. If we deal a certain level of damage then Haste allows us to kill most of the enemies before they can launch spirit attacks, which appears as a disproportionate amount of damage savings.

Your turns are ~13% worse than mine even though your equipment and forging is much better, so I guess that is what I can look forward to as I approach level 500. I suppose going from DD5 to DD9 will make back that difference.

After further thought, I understand why your Shadow Veil saves so little damage, about ~8%. It's for the same reason Shadow Veil reduces counters by much less than expected, just ~6% for both of us.

Even though we expect damage and counters to be reduced by ~25% according to the amount of evade granted, because evaded enemies don't get stunned in 1H style, when you successfully evade something the enemy quickly recovers to hurt you again on the next turn and also possibly be counter attacked.

So this would seem to make Shadow Veil bad again like I used to think. However I still think Shadow Veil is very good. The reason is that when I analyze the amount of Cures saved with Shadow Veil, it's indeed about the same as Haste. I tested this in Item World and it would be hard to test in arenas, but Cures saved are what should count.

This happens because Shadow Veil reduces damage by the 25% it deserves when it really matters: when the enemy launches their spirit attack. The fact that the enemy may recover to damage you with many other regular attacks doesn't really matter. What's important is that 25% of the time their spirit attack will miss.

This analysis is also related to the misconception that Protection equipment is better than Warding. I believe Warding is much better than Protection for 1H style. Some people before analyzed their script logs and saw that Protection reduces way more total damage. It's true, but so what? That's the wrong metric for measuring damage.

The correct way to measure damage is the amount of Cures saved, and I believe Warding would save way more Cures than Protection equipment. No one has tested it, but we can see in this Shadow Veil tested example that the damage taken and Cures saved gave entirely different results.

This analysis is also related to the misconception that Protection equipment is better than Warding. I believe Warding is much better than Protection for 1H style. Some people before analyzed their script logs and saw that Protection reduces way more total damage. It's true, but so what? That's the wrong metric for measuring damage.

Both Decon and Uncle Stu have been saying this for a pretty long time - that you should be wearing a mix, at the very least. (Force Shields in particular benefit more from warding than they do from protection.)

The correct way to measure damage is the amount of Cures saved, and I believe Warding would save way more Cures than Protection equipment. No one has tested it, but we can see in this Shadow Veil tested example that the damage taken and Cures saved gave entirely different results.

It's worth considering that not many cures are saved in the first place in most content - you can see that from the data you posted before from Pope Killdragon doing PF IWs - when using a decent gear setup, even at low level without all the mitigation & avoidance from PABs. SVeil is exceptionally useful in very late grindfest, though.

Here's another reason they're significantly worse when used together: the counter cap. Let's approach this with a few numbers.

Imagine you have 0% speed and 0% evasion, 30% block, and for the sake of simplification, 0% parry. The counter cap is 3, so you need to take an average of 10 attacks per turn to hit this.

Let's say you're, in reality, being hit 12 times per turn.

If we use shadow veil, you're now evading 25% of those, so you're being hit 9 times out of your required 10. You lose a little, but since you were already above what you needed, it feels like you didn't lose much damage. It's 2.7 counters average.

The same goes for haste, although it's a bigger drop - if we're 50% faster, than we're now taking an average of 6 attacks per turn. That's 1.8 counters average.

Now combine the two: the 12 becomes 4.5 attacks per turn, and you're suddenly only getting 1.35 counters average.

Obviously in reality it's a lot muddier than this, but the simple matter is you're being hit enough that one buff alone is only just greater than your excess, but adding the second one is then affecting the amount at its full value.

I demand an altar to be erected as well as sacrifices to be made, in the name of yours truly.

I demand an altar to be erected as well as sacrifices to be made, in the name of yours truly.

Follower of Decondelite Perk - costs 50k hath to be paid to DJNoni. The effects are great! It really, really helps*.


*really helps DJNoni, that is (IMG:[invalid] style_emoticons/default/ph34r.gif)

Follower of Decondelite Perk - costs 50k hath to be paid to DJNoni. The effects are great! It really, really helps*.
*really helps DJNoni, that is (IMG:[invalid] style_emoticons/default/ph34r.gif)

Hmmm.... what do I get?

Both Decon and Uncle Stu have been saying this for a pretty long time

I've been saying that for even longer. That's why basically everyone knows it's basic knowledge...(IMG:[invalid] style_emoticons/default/heh.gif)

- snip -

Thank you very much for the explanations. (IMG:[invalid] style_emoticons/default/wink.gif) Despite the simplifications made, it was possible to understand why the negative effect of using Haste and Shadow Veil for the 1H style. Considering all the parameters involved, the mechanism is quite complex. (IMG:[invalid] style_emoticons/default/wacko.gif)

As Juggernaut Santa just said, this is all basic knowledge. (IMG:[invalid] style_emoticons/default/heh.gif) But I decided to do these simple tests just to see, in numbers, all these concepts that have been said over the years. A lot was just said (in text), and justified in most cases only on the basis of theory (equations and considerations). And since I'm not very good with these equations (in fact, I don't even know what to do with them (IMG:[invalid] style_emoticons/default/blush.gif) ), I preferred to see them in practice. I think that the presentation of real numbers always helps to understand the complex mechanisms of the game (especially for those who do not understand this just with theory, or because they do not know it).

Follower of Decondelite Perk - costs 50k hath to be paid to DJNoni. The effects are great! It really, really helps*.
*really helps DJNoni, that is (IMG:[invalid] style_emoticons/default/ph34r.gif)

Not even for free. (IMG:[invalid] style_emoticons/default/tongue.gif) (IMG:[invalid] style_emoticons/default/laugh.gif)

This post has been edited by Basara Nekki: Jun 14 2020, 18:57

I'd already thought of lestion's explanation for why shadow veil and haste might be extra bad together, but I'm not sure that's the correct reason, so I didn't mention it. It helps, but there might be something else more important that I haven't figured out.

There's some other interesting issues with haste and counter attacks that I noticed while testing out my action speed theories (the log file). Once any enemy recovers from stun, it no longer matters whether you have haste or not in terms of the number of counter attacks you perform (except for increasing the stun duration from 4 to 6~7 turns, which is the biggest effect on loss of counters anyway).

If I interpret what I experienced correctly, it seems that ~95% of all enemies we encounter are chaosed enough to have over 25% attack speed bonus (more than half chaosed, so 10+ levels to attack speed). Combined with the 25% attack speed bonus of PFUDOR difficulty this means that whether we have haste or not, most monsters will attack 1H players twice every time they recover from Stun.

And if we approximate that 1H style deals 3 counter attacks and 2 stuns every turn, then this creates a perpetual loop. Because in many turns, 2 monsters will recover from stun, and those 2 monsters will attack 4 times, for an average of 3 counter attacks (75% chance) and once again 2 stuns (~70% chance of stun with a counter attack).

Note the wiki says "monsters receive a free action when a stun expires on them" which is as I just explained. But I think that's only true for 1H style. I haven't used DW club in a long time but I think it can restun a monster with your main hit without the monster making any attack at all.

Both Decon and Uncle Stu have been saying this for a pretty long time - that you should be wearing a mix, at the very least. Force Shields in particular benefit more from warding than they do from protection.

I've also said and done the same thing since I started playing, but End of All Hope might have said it before me, I don't know. It's not just force shields but all power and plate armor that have a bigger boost with the warding suffix than protection. I think it's the same difference in boost if you look at it additively.

This post has been edited by BlueWaterSplash: Jun 15 2020, 01:04