[X] My Assistant

@BlueWaterSplash

Do not trust the numbers on my data.
I tested each 1h setup once.

Just say that Cold strike is 'a little' better than Dark, not 'xx.xx %'
Use 'about 50%' instead of '48.52%'

Whenever you do a test, the numbers will change.

Optimal Infusion Strategy for Demonic/Hallowed Weapons

In the past the elemental relationships were roughly Dark/Holy being 10% stronger than Cold/Electric/Wind and another 10% stronger than Fire. Spike shield adds +15% to Fire, and +10% to Cold and Electric/Wind (plus speed penalty). Except for Imperil usage, Dark/Holy were superior or almost equal in all aspects; no further analysis required.

Given recent evidence I will guess all elements stayed the same except Cold improved 5% due to changed monster species. I'll guess spike shield adds +20% to Fire, and +15% to Cold and Electric/Wind due to PL increase. With day boost I think spike shield returns to before. Cold with fire shield is now almost 10% stronger than Dark/Holy.

The optimal infusion strategy for an Arctic weapon is obvious. Let's try to compare it with infusion strategies for a Demonic weapon.

Friday (Void/All), Wednesday (Cold), Monday (Dark) - Arctic user will infuse Dark, and Demonic user will infuse Cold, so it will be the same.
Sunday (Holy) - Day mitigation bonus 10% means Holy damage +20% so it's best for Demonic user to infuse Holy. The day's winner is Arctic by 10%.
Tuesday (Fire) - Demonic can use electric shield so his fire does 15% better damage than Arctic user. The day's winner is Demonic by 5%.
Thursday (Wind), Saturday (Elec) - If the Demonic user believes flame shield is still best, then the days' winners are Arctic by 10%.

If Demonic user uses cold/wind shield he will get 10% more element damage. Overall this is 1% more damage to the main attack. I studied recently that if using Haste this is around 0.8% of all damage, and if not using Haste surprisingly just 0.5% of all damage. Meanwhile I think the shield creates a loss of 0.6% damage due to counter attacks.

It seems now that cold/wind shield has very little benefit. For players who use Haste it may provide 2% bonus damage to Wind/Elec Strike. Overall the days' winners are still Arctic by 8%.

Ultimately I guess Arctic is superior against today's regular monster population. But a Demonic user can reduce the difference to 2 or 3 days by using infusions and changing spike shield on Tuesday.

Demonic/Hallowed are still quite superior to Shocking/Tempestuous/Fiery, perhaps even more so than years ago.

This post has been edited by BlueWaterSplash: Nov 5 2019, 00:05

I have been experimeting with Zebra mage (black, with white stripes)

Hoveraction:
Strongest([Cast('Paradise Lost'), Cast('Ragnarok'), Cast('Disintegrate'), Cast('Corruption')])

This works quite well in IW100 - same number of rounds, free holy proficiency. Not good for SG, with Imperil:200rounds extra.

For PFFEst, I would like to set it up like
start of round
1st: Cast Paradise Lost,
2nd: cast Ragnarok
3rd - start hover with Disintegrate & corruption until end of round

Would that be possible with monsterbation? I can't figure out that 'n' in the settings.... Sicken has made it a bit complicated by making so much possible! Does anyone know how this would be possible?

I have been experimeting with Zebra mage (black, with white stripes)

Hoveraction:
Strongest([Cast('Paradise Lost'), Cast('Ragnarok'), Cast('Disintegrate'), Cast('Corruption')])

This works quite well in IW100 - same number of rounds, free holy proficiency. Not good for SG, with Imperil:200rounds extra.

For PFFEst, I would like to set it up like
start of round
1st: Cast Paradise Lost,
2nd: cast Ragnarok
3rd - start hover with Disintegrate & corruption until end of round

Would that be possible with monsterbation? I can't figure out that 'n' in the settings.... Sicken has made it a bit complicated by making so much possible! Does anyone know how this would be possible?

move the paradise lost and ragnarok on the imperil routine?
instead of 123 as imperil monster 258 (347), it's 12 as paradise lost monster 1 (0) ragnarok monster 1 (0)
if you still use imperil, move it to 4 and 5.
while the hover is only disintegrate and corruption

This post has been edited by Fudo Masamune: Oct 27 2019, 12:06

move the paradise lost and ragnarok on the imperil routine?
instead of 123 as imperil monster 258 (347), it's 12 as paradise lost monster 1 (0) ragnarok monster 1 (0)
if you still use imperil, move it to 4 and 5.
while the hover is only disintegrate and corruption

that could work as I don't use impreril! Zebra mage uses Paradise Lost as Imperil-on-all-targets, with added bonus that the monsters are weaker against my Almighty Ragnarok after that. True, breached defense does explode, but the numbers show that breached defense is active in the turn that it explodes - much more dark damage afterwards!

Fuck me - this is just too complicated!

So I start with hover off, right? I would want the ragnarok to also toggle hover afterwards:

CODE

          Bind(KEY_1, Strongest([ToggleHover, HoverAction(Cast('Paradise Lost'), true]));\
          Bind(KEY_2, Strongest([TargetMonster(0), Cast('Ragnarok'), true]));\

this does not work?! Help would be appreciated!

This post has been edited by DJNoni: Oct 27 2019, 12:37

I've actually recently thought about doing that as a mage-warrior as well but hadn't gotten around to describing it yet; I will later. Especially since ravenfrost123 brought it up that he casts Smite and Imperil at the beginning of rounds as 1H warrior. (That would ruin any benefit of Haste as I mentioned earlier but it doesn't invalidate it as a possibly effective strategy). Aside from that, there are kind of obvious variants that would also work with Haste, which I'll bring up again later.

As for a stripe mage, maybe it could work better with holy-elemental then so that it doesn't explode Breached Defense. But from what you guys said a dark mage does higher damage so maybe the zebra mage is better.

Breached Defense won't stack well with Imperil, that's probably why it's worthless against School Girls. It's only a tiny 10% MMit reduction, it just happens to be very significant (40% more damage) when it is the first and only thing inflicted against player-created monsters with very high MMit and low HP (compared to School Girls).

Sorry I can't help on the scripting that's not my thing (for now, at least).

this does not work?! Help would be appreciated!

try this, with hover disabled at start of round and disintegrate and corruption as your hover action:

CODE

          Bind(KEY_1, Strongest([TargetMonster(0), Cast('Paradise Lost'), Cast('Banishment'), Cast('Smite')]));\
          Bind(KEY_2, Strongest([ToggleHover, TargetMonster(0), Cast('Ragnarok'), Cast('Disintegrate'), Cast('Corruption')]));\

I forgot that Breached Defense not only inflicts -10% PMit/MMit but also -25% Dark Resistance. I guess the zebra mage is best. This could also help 1H melee so we might as well start discussing evolved 1H styles now.

Holy/Dark/Cold 1H mage-warrior / Paladin

Cast T3 Holy "Paradise Lost" at the beginning of each round. Weapon is Dark+Cold, maybe even ignoring Day of the Week. Can use either Shortsword or Rapier. Breached Defense can replace Imperil and I believe T3 Holy mana cost is equal to 2 Imperils so it's actually cheaper if you only need to cast it once per round.

Because T3 Holy has a long 1.8 turn cast time, you will get hit twice as much by monsters in the first turn. Double of them will be Burned than usual and you won't Stun them all due to the 3 counters limit, enabling you to Burn even more next turn. So flame spike shield can inflict 50% of monsters instead of 25% for a short while, and Cold Strike may receive 30% damage boost instead of 15%.

Also for a short while you will get 1.4x damage from PMit reduction, and Dark Strike will further receive 70%/45% = 1.55 = 55% damage boost to 75% of monsters, so 42% damage boost.

If you have DD9 and a fully forged Peerless Slaughter set to clear enemies the fastest, and also Proficiency as a Holy mage, maybe this style can be faster than Imperil x3 Rapier or Shortsword. Maybe sssss2 is the only player who can try this style.

I think Haste won't be as good with this style because casting T3 Holy will cause Overwhelming Strikes to disappear, though Overwhelming Strikes will still be active more often, and you can get much more hits on the Burn and Breached Defense enemies.

Half-Imp or One-Imp 1H Rapier

At the beginning of each round instead of going all out with Imperil x3 or more to stick enemies, just cast Imperil once on the top group. Afterwards spread only the bottom enemies, leaving the top to die by counter attack since Imperil won't wear off. If Imperil missed the 1st or 2nd enemy just ignore them until the end. Rapier is required.

This will save a ton of mana and could be performed with just a minor use of resources. It may be just as fast or even faster than normal Imperil style: because you spread a smaller group the rapier's Penetrated Armor won't wear off as easily. Haste is a fairly compatible option with this style because casting Imperil once doesn't take much ticks.

Most 1H players would be able to test this style as long as their Deprecating Proficiency isn't too terrible. I can't though because I don't want to train my Faster Imperil to multiple targets until I'm much higher level.

try this, with hover disabled at start of round and disintegrate and corruption as your hover action:

CODE

          Bind(KEY_1, Strongest([TargetMonster(0), Cast('Paradise Lost'), Cast('Banishment'), Cast('Smite')]));\
          Bind(KEY_2, Strongest([ToggleHover, TargetMonster(0), Cast('Ragnarok'), Cast('Disintegrate'), Cast('Corruption')]));\

worked, but I had to first delete the bind:"..." from the dump screen.

Let's see how this works for Zebra Mage PFFEST!

Let's see how this works for Zebra Mage PFFEST!

wouldn't that really taxing on late round survivability?
I mean, you cast TWO T3 spell in a row, well, the ragnarok shouldn't be a problem as it'll kill a lot of them, but the paradise lost surely bring you lots of attention due to it's long casting time and it didn't kill a lot of enemies since you don't boost any holy damage?
or do you?

wouldn't that really taxing on late round survivability?
I mean, you cast TWO T3 spell in a row, well, the ragnarok shouldn't be a problem as it'll kill a lot of them, but the paradise lost surely bring you lots of attention due to it's long casting time and it didn't kill a lot of enemies since you don't boost any holy damage?
or do you?

CODE

hoverCtrlAction: Strongest([ToggleHover, Cast('Paradise Lost'), Cast('Ragnarok')]), // alternate hover action when holding ctrl
combinded wiht Bind Ctrl X to ToggleHover

Haven't tested it yet - I'll keep you posted when I do. Late round survivability sucks anyway, as I only have 19% cast speed bonus. My cotton is uncharged. I could use my 3 charged 3+2 build as well, but for now, I want to see if I can survive 2 T3's in a row


Okay - with no charged and only 3 spelweavers, this is not gonna fly. Died around round 700. But that happens all the time with this build, so not worse. However, last 200 rounds were Spark of life every single round. That, is not good.

I'm going to try again now, but with my 3+2 high cast speed build.

Hmmm... Defeated again, at round 800. This does not work as an alternative to imperil, it's too slow. Well, it does for IW and for normal arena's, and for the first 500 rounds of PFFEST. But then when it gets tough, it just doesn't work.

Is there any way to reduce the holy cast speed? Extra proficiency?

This post has been edited by DJNoni: Oct 27 2019, 21:17

I doubt there is anything you can do, but you can just use zebra mage on the first 500 rounds then switch to playing normally after that.

I discovered a flaw in the approach to use average damage to perform elemental comparisons. It's not obvious, but it ignores PMit after all. Imagine there are just two kinds of monsters:

Peasant
100,000 HP (dies in one or two hits)
0 PMit / 50 Resistance to Element A / 75 Resistance to Element B

Superman
100,000,000,000 HP
0 PMit / 75 Resistance to Element A / 75 Resistance to Element B

Peasant is 100 times more common than Superman. Calculating the average damage will result in the conclusion that Element A = Element B. This is correct; it means this approach correctly takes into account monster HP.

Superman
100,000 HP
99.9999 PMit / 75 Resistance to Element A / 75 Resistance to Element B

Change the Superman to this guy. The result of combat is exactly the same. The average damage will change, so it seems as if this method takes into account PMit. But if you compare the average damage you will find Element A = 2 * Element B. This is the wrong answer.

What about Penetrated Armor? It drastically affects average damage and I originally suspected it would cause further problems. But it makes no difference because the average damage comparison ignores PMit. In fact it's beneficial: because PMit drops quickly and most damage occurs afterward, PA makes the average damage method close to reality. Thus I don't think we need to change it.

If insisting on perfection, the method can be fixed. Instead of summing and averaging the damage, we should sum and average the resistances with each hit. This will correctly take into account HP, PMit, and PA. It requires the monster scan data to work.

Is there any way to reduce the holy cast speed? Extra proficiency?

Scrolls!

Scrolls!

aye, I always use them. I even use infusions on my armor!

Although Zebra Mage doesn't work well against SG, it works surprisingly well against FSM and TT&T RoB!

When did sssss2 collect that monster population data? I built a monster population spreadsheet this week so I can quickly calculate all the elemental damage relationships, but I can't get it to match his recent results with greatly superior Cold Strike using any variant of the population in his data.

CODE

=====================================
  Mitigation |  Average |  Imperiled
-------------------------------------
        fire |  59.91 % |    21.83 %
        cold |  56.01 % |    19.01 %
        elec |  59.11 % |    20.64 %
        wind |  57.38 % |    18.75 %
        holy |  53.35 % |    28.66 %
        dark |  51.51 % |    26.83 %
=====================================

Then I noticed in this data he provided with the monster population, the results neither match Research for 1H nor does it match his recent ones. However this data matches what my spreadsheet calculates for his population.

(IMG:[i.imgur.com] http://i.imgur.com/ATekfb9.png)

I made the spreadsheet look similar to the Monster Lab wiki. The first page has base monster stats. The primary purpose is to compare the overall average damage of each element. Real world data gathered by sssss2 should correspond to HP weighted averages. This method can additionally calculate Pmit weighted averages but with a rapier I'm not sure which is better.

(IMG:[i.imgur.com] http://i.imgur.com/DwAsEY2.png)

In reality the base monster stats never apply at any player level. The closest I could get was around level 69. The variation in scaled monster stats is less than the base.

(IMG:[i.imgur.com] http://i.imgur.com/tmLMFV5.png)

It turns out that player level barely affects the final elemental comparison so from now on I'll just enter level 500. Monster PL matters more as I described earlier. I start out with an equal monster population and PL 2250 even though in reality the game won't allow the average to be that high.

(IMG:[i.imgur.com] http://i.imgur.com/5AQC8Em.png)

In general spike shield simply adds 25% * 25 = 6.25 as long as monster PL is high enough that elemental resistance stays above the lower cap of 0. That's why with higher PL the four elements get increasingly worse than holy/dark, but better with spike shields.

(IMG:[i.imgur.com] http://i.imgur.com/wu9Jrzm.png)

Holy/dark are not as good with imperil, we all knew this. With mages there is no difference since it all drops to 0. I will show the results for sssss2 monster population, varying PL, and non-imperil mages later.

Holy/dark are not as good with imperil, we all knew this. With mages there is no difference since it all drops to 0. I will show the results for sssss2 monster population, varying PL, and non-imperil mages later.

I'd like to clarify that while holy/dark do not get as much benefit out of imperil as others (in part because it reduces mitigations by less than their elemental counterparts, because holy/dark imperil abilities cap out at 40 rather than 50) they still do get very significant benefit out of using imperil.

There is not significant data about non-imperil released at the moment, but it is likely it performs similarly, while costing slightly less real-time (because targeting imperil is slightly more tedious). However the difference in gear and DD perks required to achieve those heights is ridiculous. Most holy/dark mages should still be playing imperil.

This post has been edited by lestion: Nov 4 2019, 09:12

After playing with my spreadsheet I realized the improvement in flame shield is only partly due to PL increase tweaking the numbers. Monster resistances were low enough years ago that flame shield also hit the bottom cap of 0 resistance sometimes. That contributed to shock shield doing better back then. These days they should be more equal.

Optimal Infusion Strategy for Demonic/Hallowed Weapons (Revised)

In the past the elemental relationships were roughly Dark/Holy being 10% stronger than Cold/Electric/Wind and another 10% stronger than Fire. Spike shield adds +15% to Fire, and +10% to Cold and Electric/Wind (plus speed penalty). Except for Imperil usage, Dark/Holy were superior or almost equal in all aspects; no further analysis required.

Given recent evidence I will guess all elements weakened equally, except Cold improved 5% due to changed monster species. I'll guess spike shield adds +15% to Fire/Cold/Electric/Wind after PL increase. With day boost I think spike shield returns to before. Cold with fire shield is now almost 10% stronger than Dark/Holy.

Friday (Void/All), Wednesday (Cold), Monday (Dark) - Arctic user will infuse Dark, and Demonic user will infuse Cold, so it will be the same.
Sunday (Holy) - Day mitigation bonus 10% means Holy damage +20% so it's best for Demonic user to infuse Holy. The day's winner is Arctic by 10%.
Tuesday (Fire) - Demonic can use electric shield so his fire does 15% better damage than Arctic user. The day's winner is Demonic by 5%.
Thursday (Wind), Saturday (Elec) - Demonic can switch to cold/wind shield for 5% improvement if using Haste. The days' winners are still Arctic by 5%.

If Demonic user uses cold/wind shield he will get 15% more element damage. Overall this is 1.5% more damage to the main attack. I studied recently that if using Haste this is around 1.1% of all damage, and if not using Haste surprisingly just 0.6% of all damage. Meanwhile I think the shield creates a loss of 0.6% damage due to counter attacks.

Arctic is superior against today's regular monster population, but a Demonic user can reduce the difference to 1 or 2 days by using infusions and changing spike shield to match.

This post has been edited by BlueWaterSplash: Nov 5 2019, 00:02

I revisited some staff calculations with some requests from Paruri for certain parameters to be adjusted for accuracy. This is the first part of that data.

This is for imperil mages, and therefore should not be considered a very good evaluation of clear speed. Obviously, the deprecating CR makes a very big difference (somewhere to the tune of about 450 turns per grindfest, according to the imperil simulator I posted about before). But these numbers are interesting because they show the actual damage difference to be minimal in cases of both worst-case and average monster resists.

Additionally, we do know that redwood pulls ahead on raw damage against schoolgirls because they have lower resist. Do not misconstrue these numbers to be true in absolutely every case: they apply against the average set of player monsters, but not against boss monsters.

Now let's talk about the data itself and how it was constructed.

I started with a few assumptions about reasonable base proficiency (I will detail these cases shortly) and used that, combined with various levels of gear, to calculate proficiency factors for various cases. Then I simulated 10,000,000 spell casts against two levels of monster resist and used that and the equipment data to provide some average damage values.

I want to concede that I had to make some assumptions about two things: MDB is very difficult to calculate and will vary largely between gear levels, and so I chose to use static amounts based on real numbers gathered from some players in the Discord. These are approximations, but the relationship between MDB, EDB and counter-resist is essentially linear, and different levels of MDB still produce exactly the same % difference between staffs. These inaccuracies can be forgiven.

Additionally, there is no data on the average monster resist at lower levels. This is impossible to calculate without real play data. I chose to use the same resist rates level 500 players face, and I do not believe the real data would be significantly lower (400 players face essentially the same monsters, with just some slightly lower wisdom, which will cap for some monsters even then).

Now, to explain the cases:
1) Willow & Shoes, at 100% rolls and full forging, can reach 0.79 at 579 base proficiency. Using the same % rolls and base prof, a Redwood user reaches 0.876 prof factor. The EDB is also adjusted.
2) With 0% rolls and 0 forging, neither Willow nor Redwood can reach 0.79. We use the same base as above - 579 - but with this new gear, they reach 0.704 and 0.75 respectively.
3) Now I begin cases for average users - with an easily achievable 520 prof that someone who only just reached level 500 might have, 50% rolls on their gear, and forge level 13 (where ~50% of the maximum forge factor is reached). Again, they cannot reach 0.79 - Willow gets 0.682, and Redwood gets 0.748.
4) Finally, we'll test a level 400 case, using the same 50% rolls, forge level 13, and a reasonably 420 proficiency. Willow gets 0.7, Redwood gets 0.765.

First, the data against the current average monster resist (about 21.6%, provided by sssss2's tests).



Second, the data against maximum resist monsters (27.1%).



Accounting for the res reduction from prof factor as well as imperil, we can see that for pure damage, Redwood is basically the same as Willow at optimal gear levels, and marginally better in other cases.

Obviously, the difference is so small that it can be completely thrown out as irrelevant to your actual game speed, because imperil resists will slow you down far more than the small amount of damage here. lol

This post has been edited by lestion: Nov 13 2019, 04:48