[X] My Assistant

NOTE: The info in this post is somewhat out of date as of HV 0.5.1.
Base action time is now determined by subtracting action speed from 500, not 300. I'm assuming that nothing else has changed, but I don't know for sure. I might look into it some time, or I might not.

I worked out some stuff that might be of interest.

So, what do we know about action time? Well, every player or monster action takes a certain number of "ticks" in game time. For the player, the number of ticks varies depending on the action, while for a monster all of its actions have the same duration. These times determine the order of player/monster actions.

The basic character attribute is action speed. Here's what Tenboro had to say about how that translates into action time.

action speed translates into how long a 1.0x action (like an attack or tier 1 elemental spell) by subtracting it from 300. So if you have an action speed is 112, that means a 1.0x action takes 188 time units to complete

And some info about things that affect it:

X-anything makes that particular action 25% faster.

Every point of burden above 20 reduces action speed by one "point".

- Haste has been tweaked as follows:
-- The base action speed increase has been reduced from 25 to 20
-- The effect of proficiency has been increased. Every ten points of proficiency increases the bonus by 4, for +100 action speed with 200 proficiency.
-- There is now a readout on the status effect for how much your action speed is increased.

- Rebalanced the Shadow Veil and Haste scrolls to equal a 200-proficiency cast.

We also know that various player actions have time multipliers greater or less than 1, though there's not much information on what the numbers actually are. I scraped some references together, but there's no guarantee that any of this is current:

- All AoE spells have been changed to five-point abilities. The first point will give you the spell like before. Each additional point reduces the spell's cast time by 0.1, increases the damage of the spell by 5%, and decreases the mana cost of the spell by 5%. With the full 5 points, the cast time of the AoE will be equal to the corresponding single-target spell.

- Rebalanced the cast time of offensive magic:
-- Reduced the cast time of Tier 1 (Fireball, etc) offensive spells by 0.4. The usage time of Tier 1 single-target spells is now equal to that of normal attacks.
-- Reduced the cast time of Tier 2 (Flare, etc) offensive spells by 0.2.
-- Reduced the cast time of Tier 3 (Nova, etc) offensive spells by 0.1
Tier 4 have the same cast time as before. This is mostly to differensiate between the different ties of spells, and make the lower tiers more useful on later levels.

- Smite and Banish are now properly categorized as Tier 2 spells, and have had their cast time reduced by 0.3 to reflect this. Damage/cost remains the same.

- The spell Cure II is now available at level 100. Compared to Cure, Cure II restores 250% health but costs 275% mana (90% efficiency) and has a longer cast time (0.6).

- The spell Cure III is now available at level 200. Compared to Cure, Cure III restores 400% health but costs 500% mana (80% efficiency) and has a longer cast time (0.7).

Also a little snippet of info regarding monsters. If you look at the start of a battle log, where it says "Spawned Monster A: ...", one of the parameters it outputs is called "NA". What's NA?

Debug output. It's short for Next Action, the time units until the monster has its first action in the battle.

Now, I had a theory that each monster's NA was equal to half of its action time, presumably to avoid either side getting a double hit straight off the bat (assuming you just charged in with a normal attack). Well, I finally got around to checking that (I was right, btw), and in the process I found out a bit more about how the whole thing works. It turns out some of what it says above is wrong...

The most significant discovery was how Burden works. It turns out the time penalty from Burden is applied equally to every action. It isn't affected by the action time multiplier, nor is it reduced by X-Attack ar Haste. The only case where it doesn't apply is when using a Powerup Gem, which takes zero time (you may remember Tenboro had to tweak it to specifically exclude the penalty for this). It's not all bad news, though: after disregarding the first 20 points of Burden, the penalty is equal to only half of the remainder.

I also confirmed that action speed is capped at 200. This has a number of consequences.
1. Since everyone's base action speed is at least 100, and Burden is treated separately, this means that the nominal 100 boost from a Scroll of Swiftness is actually rather meaningless: using one will always make your action speed equal to 200.
2. There's no point in using the spell version of Haste and a Scroll together, as your speed can't be increased any further.
3. From the point where your Hastened speed hits the cap, the spell becomes less and less useful as you level up. This shouldn't be a big problem though, as the monsters' speed also appears to be capped, and they don't get X-Magic at level 200.

I'll try to summarise all of this in a vaguely mathematical fashion.

Base action speed = 100 + AGI / 3

Boost from Haste (spell) = 20 + Supportive prof / 2.5
Boost from Haste (Scroll) = 100

Action speed = max (base action speed + boost from Haste, 200)

Base action time = 300 - action speed

Time multiplier = action specific multiplier * X-action multiplier
Some action specific values are:
Normal attack = 1.0
Defend = 1.0
Magic Missile = 1.1
I haven't looked at any others.

X-action multiplier = .75 if player has the appropriate X-* ability, 1.0 otherwise

Burden penalty = (Burden - 20) / 2 if Burden > 20, 0 otherwise

Action time = (base action time * time multiplier) + Burden penalty

This post has been edited by Sayo Aisaka: Mar 13 2011, 22:43

Actually this is better than I thought, even if X-action don't reduce 1/4 the effect of burden penalty, each point of burden above 20 only give 1/2 points of burden penalty (IMG:[invalid] style_emoticons/default/smile.gif)

The speed euqations have always been too complicated.

I worked out the numbers for Regen. The reason no one could fit it to a straight line graph was, of course, that it isn't a straight line graph...

The calculation uses base HP (i.e. including equipment and Shrine bonuses, but not Health Tank) and Curative proficiency:

Amount of HP restored per turn = base HP * 3 * (1 + prof / 100 + (prof / 200) ^ 2) / 100 , rounded up.

I worked out the numbers for Regen. The reason no one could fit it to a straight line graph was, of course, that it isn't a straight line graph...

The calculation uses base HP (i.e. including equipment and Shrine bonuses, but not Health Tank) and Curative proficiency:

Amount of HP restored per turn = base HP * 3 * (1 + prof / 100 + (prof / 200) ^ 2) / 100 , rounded up.

Did you send this equation to Sonic to add to the calculator?

I've put it in his calculator thread.

edit: Wrong thread.

This post has been edited by Anto of the Sand: Jul 23 2010, 00:06

Some more numbers, for those who like to know such things.

Because I was bored, and also because I recently completed my set of Elementalist gear, I thought I'd take another look at how Elemental proficiency affects spell damage. It didn't alleviate my boredom, but at least I was able to pin down the scaling factor:

Spell damage multiplier = 1 + prof / 200

I also looked at magic rating, since we're told damage depends on magic rating vs barrier rating. There are too many unknowns to be able to guess at the formula, but it was interesting to discover that a 5% difference in rating gave approximately a 10% difference in damage. Looks like I may have been underestimating the importance of INT bonuses when choosing my equipment...

I was going to include the details of what I did, but... it's really dull. I'll just say that it involved two sets of equipment, one with maxed Elemental bonus and one with none, chosen to give the same magic rating. Cast spells, record damage, compare results. They agree with a 200 scaling factor. (Give or take. It could be 203.405 for all I know... but it's definitely less than 250). And some similar comparisons using non-Elementalist sets that maximised and minimised magic rating.

SPELL DAMAGE

NOTE: This post is out of date as of HV 0.5.3.
Elemental ratings no longer increase spell damage, and the way Infusions work has been changed. (They give a 25% boost to their respective element, but it's not clear where that 25% actually fits into the equation. I'll probably work it out at some point.)
The Interference changes have been incorporated.

Since people seem to be interested in maximising their spell damage, I thought I'd try to collate a list of things that affect it, and figure out how they all work. So here it is:

1. Basic damage roll.
   1. Random variation: maximum damage is 1.5 times the minimum
   2. Critical hits ("blasts") multiply damage by 1.5
   3. Channeling also multiplies by 1.5
   
   
2. Magic Damage Multiplier. Comes from:
   1. HV title
   2. Indigo Aura (2% per Rank)
   3. MDM from equipment
   4. Arcane Focus. Applies a damage multiplier that depends on Supportive proficiency. Base is 10%, with a 1/250 scaling factor (so with 250 prof you'd get 20%)
   These all add together for a straight % boost.
   
   
3. Proficiency. Proficiencies that directly affect damage are:
   1. Elemental (affects Fire, Cold, Elec, Wind, Magic Missile)
   2. Divine (Holy)
   3. Forbidden (Soul, Dark)
   4. Deprecating (Poison)?
   Spell damage is multiplied by (1+prof/200).
   Note: I've only checked Elemental. It's probably safe to assume Divine and Forbidden work the same way, but as for Deprecating, who knows?
   
   
4. Elemental damage boosts. Two sources:
   1. Elemental ratings (Soul, Fire, Cold, Elec, Wind, Holy, Dark)
      These come from
      1. Auras, including those obtained from Hath Perks
      2. Points invested in Abilities
      3. Infusions. These add 50% to the respective Rating.
         (Infusion of Gaia boosts all elemental ratings)
   2. Elemental Damage Bonuses from equipment
   The bonuses for each element add together to give a % boost.
   
   
5. Interference (from equipment).
   Each point of interference over 25 reduces all magical proficiencies by 0.75%.
   As of HV 0.5.3, interference no longer reduces spell damage directly.
   
   
6. Monster elemental resistances/weaknesses.
   The Monster Lab offers some clues as to how they work: it appears that monsters have ratings for each element (in fact for all damage types), which presumably work in a similar way to player elemental ratings. Interesting snippet: some ratings can be negative, which presumably increases damage relative to the base value. In theory, we could test how it works by comparing the effects of various spell types against some player monsters, while they're still unbuffed and their stats can be determined from their class... but I'm (probably) not going to do it. After all, what we really want to know is "which spell hits monster X hardest?", not "what % of base damage does Fire do?"
   There's one thing that we know affects these: status effects from spells of one element reduce resistance to spells of another - hence the well known Elec+Fire strategy. My guess is that it would subtract a fixed amount from the respective rating, but I haven't really looked into it.
   
   
7. Caster's Magic Rating versus target's Barrier Rating. Lots of things affect this:
   1. Basic rating derived from your primary stats (INT and WIS).
      This can obviously be boosted by Auras, Shrine bonuses and equipment such as Fox or Owl gear.
   2. Monster barrier rating. Not that we know anything about how this is derived, but I'd expect it to scale in proportion to its level.
   3. Difficulty setting. Monsters take less damage on higher difficulty.
   4. Nerf lowers the target's barrier rating.
      The "strength" of Nerf depends on Deprecating proficiency, but the actual numbers are unknown, not can I see any way of working them out.
   5. Breached Defense (status effect from Holy spells) lowers barrier rating, though apparently not by much.
   6. Bewilder increases damage taken. From the way it works I'm guessing it reduces barrier rating. The effect is even smaller than that from Breached Defense, and almost unnoticeable for common monsters.
   7. Using a Flower Vase increases magic rating - quite significantly, judging by the damage increase.
   8. Bubble-Gum probably works the same way. Or it might reduce all your targets' barrier ratings. Or it could do something entirely different. I don't plan to waste mine finding out. (IMG:[invalid] style_emoticons/default/smile.gif)
   And some things that don't have any effect:
   1. Penetrated Armor (as of 0.5.0). (IMG:[invalid] style_emoticons/default/sad.gif)
   2. Item World equipment quality.
   
   Exactly how Magic Rating and Barrier Rating interact is a mystery. We don't know what the Barrier Ratings of any monsters are (and the Monster Lab doesn't give any clues). Nor do we know how the various debuffs affect the numbers, and their results are far from simple. But I've managed to work out the relative effectiveness of some of the factors listed above:
   
   Nerf > Breached Defense > Bewilder
   
   And I found something quite interesting. Remember how I said an increase of X% in Magic Rating seems to increase damage by 2X%? Well, it may be about that (or a bit less) for common monsters, but it turns out the effect of reducing Barrier Rating is greater, the higher it is to begin with. For example, Nerf is much more effective on a Manbearpig than it is on a Cockatrice. There's more information below, if you can bear to wade through my detailed results.

Summary

I'm not going to go into all the formulas for calculating totals that you can find on your equipment screen...

MDM = magical attack damage boost given on Equip screen, plus (10*(1+Supportive prof/250))% if using Arcane Focus
EDB = elemental rating given on Character screen, plus spell damage bonus on Equip screen, plus 50% if using appropriate Infusion
Prof = relevant proficiency from Character screen, +/- any modifier given on Equip screen
Interference = value given on Equip screen
Base damage = lumps together everything else, since we don't really have any exact numbers...

Spell damage = Base damage*(1.5 if a crit)*(1.5 if Channeled)*(1+MDM)*(1+EDB)*(1+prof/200)


I hope I haven't made too many mistakes, or left out anything important...


From the Horse's Mouth

Just for reference.
There's not a lot of information out there, but here's what I managed to find.

Your spell damage is determined by a) the damage roll from magic rating vs barrier rating, b ) your elemental rating, c) your proficiency, d) the target's elemental rating, e) the critical multiplier, f) the base damage rating of the spell, and g) whether the target was the primary target for an AoE. Simple. (IMG:[invalid] style_emoticons/default/rolleyes.gif)

Not that this tells us anything new.

Re Breached Defense:

It removes base 10 points shield/barrier rating, which is kinda meh I guess. But for a free proc, hey.

The "base" part implies that the amount could be increased somehow. By Holy proficiency? Who knows?

Question about nerf: can you have it display how much the enemy's ratings are reduced by, or at least tell us the base and how it increases with proficiency so we can figure it out? We really have no way of knowing if nerf can take the place of weaken/bewilder/blind etc in any situation :\

Same thing for arcane magic and heartseeker. It would be nice to know how much they actually affect us.

Nerf is mostly balanced against the X-Nerf you get on level 200, so its main function is increasing how much damage monsters take as well as reducing their chance to evade and resist. The effect on rating isn't very high. But either way, the numbers it uses aren't very useful for a tooltip, since you're just gonna go "ok, and that does what exactly?".

Arcane/Heart is a bit easier though. They currently increase the corresponding damage/hit chance/crit chance equipment stats with base +10, with a /250 proficiency scaler. The reason they don't display their actual effect is I haven't really settled on those numbers yet.

I'm still thinking that, even after looking into it...

- Deprecative magic proficiency now has a larger effect of the damage from Poison.

- Deprecative magic proficiency now has a larger effect of the effect of Nerf, particularly the defensive reduction.

Well, we don't know the numbers, but at least we know it has an effect. And it's larger than before.


Testing

Now for the boring part. How I tested this stuff, and what results I got. Read on only if you really want to check it for yourself.

The basic idea is to hit the same monster with the same spell as many times as possible, to ensure that I get as close as possible to the minimum and maximum damage for that spell. A ratio of 1:1.5 would mean that I've nailed the end points of the range. So for most tests I set the difficulty to IWBTH and use an element the monster is resistant to. To reduce the damage I take, I play in a Flimsy Item World unless stated otherwise.

Reference

First, I need to establish a baseline. Set title to Newbie, turn off Indigo Aura, wear Silk (no INT/WIS bonuses) and don't equip a Staff (or any other weapon). Again, this applies to all tests unless stated otherwise. This gives me

Elemental prof = 187 (capped at my level)
Supportive prof = about 162
Magic Rating = 429
Magic Damage Multiplier = zero
Relevant Elemental ratings: Fire 35%, Cold 36%, Elec 36%, Holy 25%

Reference encounters, giving monster, spell used, min/max damage, and the ratio of max to min (in brackets):

L182 Fire Fox, Fireball: 242/363 (1.5)
L184 Fire Fox, Fireball: 239/358 (1.498)

L184 Blue Slime (Bewildered), Lightning: 268/400 (1.493)

L184 Cockatrice, Icestrike: 274/406 (1.482)
L183 Cockatrice, Icestrike: 278/409 (1.471)

L182 Giant Panda, Icestrike: 272/405 (1.489)

I tried out various things on each of these monsters (the results of which I'll list below), i.e. many of my comparisons use data from a single monster. However, it would appear that all monsters of the same type and level have identical stats, so it doesn't really matter whther I fought one L182 Fire Fox or several.

One thing we can see from these initial values is the effect of monster level, which presumably affects barrier rating. Two levels difference (Fire Fox) reduced the damage by between 1.24% and 1.38%. (Results for the Cockatrices are inconclusive due to insufficient range of hits.) Luckily that's not a big enough difference to interfere with most of my tests.

We'll leave further investigation of magic/barrier ratings for later.

By the way, when comparing two sets of damage values, we compare the maximum of each set (divided by 1.5) with the minimum of the other. This gives upper and lower bounds on the true increase/decrease.

Note: Having said all this about establishing a reference and knowing exactly what my comparisons are based on, I realised only later that

1) I had somehow got it into my head that Arcane Focus affects magic rating, when in fact it's just a straight damage multiplier, and
2) I had completely failed to look at the effects of Supportive proficiency on Arcane Focus.

So I had to do a few more tests. Of course, I had now levelled up and raised some of my stats, and I have no idea how much my proficiencies had changed in the meantime (probably not significantly. I hope). Oh well. Changing my Supportive prof requires using gear which would change my Magic Rating and MDM too, so that pretty much means establishing a new baseline anyway...

Elemental Damage

First, confirming that an Infusion adds 50% to the damage from elemental ratings.

Using an Infusion of Flames:
L182 Fire Fox, Fireball: 338/493 (1.459)

Damage increase: 35.8% to 39.7%
Expected increase: (135+50)/135 = +37%

Using an Infusion of Frost:
L184 Cockatrice, Icestrike: 377/559 (1.483)

Damage increase: 36% to 39.3%
Expected increase: (136+50)/136 = +36.8%

That seems to confirm the 50% increase.

To see how elemental damage bonus from equipment combines with the bonus from ratings, I decided to use Holy, basically because I have some Heimdall gear... Holy damage boost: 34.3%

L186 Cookie Monster, Condemn (making sure not to cast with Breached Defense active): 706, 1052 (1.49)
With Infusion of Divinity: 931/1366 (1.467)

Expected damage increase:
If EDB adds to rating bonus: (125+34.3+50)/(125+34.3) = +31.4%
If they are separate multipliers: (125+50)/125 = +40%

Actual increase: 29% to 32.7%. This confirms that they add together.

Difficulty

On Nintendo:
L187 Fire Fox, Fireball: 257/382 (1.486)

That's an increase of 5.2% to 6.2% damage over the L182 "reference" Fire Fox. We'd expect less damage because of its higher level (about 3%?), so difficulty clearly makes a difference. I'm not going to investigate this further, because it would be quite tedious...

Item quality

Using an Average for IW instead of a Flimsy:
L185 Cockatrice, Icestrike: 272/404 (1.485)

Damage reduction versus L184 reference: Zero to 1.7%. If the quality did make a difference, I would expect it to be more noticeable.

Magic Damage Multiplier

Just to confirm that this multiplier is applied separately from all the others.
Set title to Destined, giving (in theory) an 8% multiplier.

L186 Blue Slime (Bewildered), Lightning: 287/427 (1.488)

Damage increase vs L186 reference: 6.2% to 7.6%.
Allowing for the difference in level, it looks like it does what it says on the tin.

Arcane Focus

With AF: L183 Cockatrice, Icestrike: 318/476 (1.497)

Expected damage increase (if the numbers given by Tenboro are still correct): (1+162/250)*10 = 16.48%.

Actual increase: 14.1% to 16.6%.

While I was still under the impression that AF worked by raising magic rating, I tried it a few times in combination with Nerf. Turns out that was fairly pointless, but it did give me a few more results that confirmed the numbers.

L184 Fire Fox, Fireball: 322/477 (1.481)

Damage increase over Nerf alone (see Nerf tests below): 15.2% to 18.7%.

L184 Manbearpig, Fireball: 792/1167 (1.473)

Damage increase over Nerf alone (see below): 13.0% to 18.7%.

All of these results are consistent with the expected value.

Is the damage boost independent of other magic damage multipliers, or do they add? Using my normal mage setup gives me +64% damage. (No boost to Supportive prof.) If AF adds to this, it should increase damage by

(164+16.48)/164 = 1.10 times, or a 10% increase.

It wasn't easy getting good results with gear optimised for killing things quickly... but I eventually got some data:

L185 Fire Fox, Icestrike: 669/992 (1.483)
With AF: 727/1076 (1.48)

Damage increase: 7.2% to 9.9%
Allowing for rounding errors, that's probably consistent with the expected value. (And I got some other results which, though not very precise, all excluded 16.48%.)

To check the effect of Supportive proficiency: Add an Earth-walker staff to the basic equipment setup. This gives:

Supportive prof: 162.54 + 49.55 = 212.09
Magic damage: +16%

Damage boost from AF should be 10*(1+212.09/250) = 18.48%. However, since it adds to MDM, the overall increase should be (116+18.48)/116 = 1.159, or +15.9%.

L182 Giant Panda, Icestrike: 324/486 (1.5)
With AF: 375/557 (1.485)

Damage increase: 14.6% to 15.7%. Well, that's probably within rounding error.

Interference

Load up with Heavy armour, axe and shield for 75.6 interference. Elemental proficiency is reduced by 70.94.
Expected damage relative to zero interference: (1-0.756) * (1+(187-70.94)/200)/(1+187/200) = 19.9%.

L182 Fire Fox, Fireball: 49/73 (1.49)

Actual damage: 20.1% to 20.2%. Again, probably within rounding error. (I know, I should probably do a rigorous analysis. Life's too short.)

I had a thought. Just to confirm the interference reduction is applied separately from the MDM boost, wear a Heavy set and set title to Destined. 75.4 interference, 187.22-70.76 elemental proficiency. (Yes, the numbers are different. The Destined title reduces interference by 3%, so I changed my helmet to bring it back up. Also, I had levelled up. None of this matters.)

Expected damage relative to zero interference and MDM:
If interference penalty adds to MDM: (1+0.08-0.754) * (1+(187.22-70.76)/200)/(1+187/200) = 26.7%
If they are separate multipliers: (1-0.754) * (1+(187.22-70.76)/200)/(1+187/200) * 1.08 = 21.7%.

L185 Cockatrice, Icestrike: 59.3(assumed, from 89 crit)/88 (1.48)

Actual damage vs L184 reference: 21.4% to 21.8%. So the multipliers are separate.

Status Effects

Although I didn't specifically look into this, I noted some numbers along the way (they're all somewhere below, if you're interested). Against Manbearpig, Fire spells did about 4x the damage of Cold, and the damage from Cold spells was increased roughly 3x when cast with Searing Skin in effect. That's consistent with MBP's ratings being 0% for Fire and 75% for Cold, ratings reducing damage in the same way for monsters as for players, and Searing Skin reducing Cold rating by 50%. That's a plausible looking number - but there are infinitely many possible solutions...


Magic Rating and Barrier Rating

Now we come to the complicated part... First, let's look at the effects of some things on common monsters.

Bewilder

L182 Giant Panda, Icestrike: 279/410 (1.47)

Damage increase: 0.5% to 3.3%. It's fairly pathetic...

Breached Defense

L187 Mutant Peacock, Condemn.
Without Breached Defense: 287/425 (1.481)
With Breached Defense: 295/442 (1.498)

Damage increase: 2.7% to 4.1%. Better than Bewilder, but probably not worth it unless the monster is weak against Holy.

Nerf

L184 Fire Fox, Fireball: 276/407 (1.475)

Damage increase: 13.5% to 15.6%.

In terms of damage increase, Nerf seems comparable to AF. I won't go into reasons you might choose one over the other, since that's a little off topic. However, read on for my findings re Bosses.

Bosses

So far, so good. But things are significantly different when we look at a monster with (presumably) higher Barrier Rating. Fought L182 Manbearpig in the RoB on Normal. Started off with no magical debuffs, but then realised I'd better use Bewilder to guard against his Spirit Attack, so Bewildered is effectively my baseline.

Used Icestrike.
Not Bewildered: 64/95 (1.48)
Bewildered: 69/104 (1.5)
Bewilder+Nerf: 108/158 (1.46)

Damage increases as follows.
Bewilder: 8.3% to 8.9%
Nerf: 52.7% to 55.8% over Bewilder

As you can see, the effect of reduced barrier rating is much higher for MBP than it is for common monsters.

(Also used Flame Spikes, which had the effect of sometimes giving him Searing skin. Noted a few values for Icestrike: 229/311)

Increased magic rating

By replacing my Silk with Gossamer (Fox and Owl, but avoiding Elementalist - and still not using a Staff) I managed to increase my magic rating to 494.

Effect on common monsters:
L183 Blue Slime (Bewildered), Lightning 342/507 (1.482)
L183 Cockatrice, Icestrike: 345/516 (1.496)

Damage increase: 26.1% to 28.2% (Slime), 23.7% to 26.5% (Cockatrice).
(Note that the Slime is a level below the reference)

I also tried an intermediate value of MR (462).

L185 Fire Fox, Fireball: 275/411 (1.495)

Damage increase over L184 reference: 14.6% to 15.2%.

What does that tell us? Well, comparing the % increase in damage against the % increase in Magic Rating for each set, we get

462 MR: 1.98 to 2.05 times
494 MR: 1.68 to 1.75 times

The increase in damage is greater than the increase in rating, which agrees with some rough testing I did before, but it seems to fall off at higher values. We are comparing different monsters, though, with possibly different barrier ratings, so I'm not sure if any firm conclusions can be drawn. Nor can I tell whether players at different levels will get similar results.

Back to max MR set.
Fought MBP in RoB again... and forgot to turn down the difficulty from IWBTH. Had to spend a lot of mana on supportive/curative spells... Still, I can be sure his barrier rating wasn't lowered. He was Bewildered all of the time.

L184 MBP, Icestrike: 132/196 (1.485)

Damage increase over L182 MBP on Normal: 89.4% to 90.4%. That's quite a lot, and it's despite two levels difference and the increase in difficulty.

I also looked at the effect of Nerf: 169/252 (1.491)

Damage increase: 27.3% to 29.3%. That's a lot less than what I got with my minimum MR set (though it's roughly similar in terms of points of damage). Clearly you get less benefit if you're starting from a stronger position.

Might as well look at his weakness (Fire) since there's no danger of killing him too quickly.

Fireball: 528/792 (1.5)
Nerfed: 683/993 (1.454)

Damage increase: 25.4% to 29.4%. Well, you'd expect it to be the same across different elements.

As it happens, I ran into a Manthra while testing with my normal mage setup (which gave me a magic rating of 497, not too different from 494 as it happens), so I tried Nerf on him too.

L185 Manthra, Fireball: 1030/1528 (1.483)
Nerfed: 1223/1721 (1.407)

Damage increase: 11.4% to 20.7%. Not very precise, unfortunately, but it's definintely less than what it gave me against the MBP. This tends to confirm that reducing barrier rating is more effective against more powerful opponents.


One final test

I had a quick look at the damage dealt with my normal mage setup, all auras on, Destined title set, etc.

Elemental prof: 187+93.07 = 280.07
Magic Rating: 497
Magic Damage Multiplier: 64%
Intereference: 0.3
Cold rating: 39%
Cold damage bonus: +3.1%

Expected damage boost to Cold spells: (1+280.07/200)/(1+187/200) * 164% * (100-0.3)% * (139+3.1)/136 = 2.12 times

...times some unknown quantity due to difference in Magic Rating. However, the value is quite close to the maximum (without staff) of 494 that I tested before.

L182 Cockatrice, Icestrike: 756/1127 (1.491)

Damage increase vs L183: 2.18 to 2.20 times, a little higher than expected. A difference of one level and 3 points of MR would account for some of that, but all of it? I really don't know. It's close enough that I'm not going to look into it any further, at any rate.

This post has been edited by Sayo Aisaka: May 31 2011, 21:30

Soo, I guess nothing is obviously broken at any rate?

Exactly how Magic Rating and Barrier Rating interact is a mystery

It really is, but I'm afraid to tinker with it. (IMG:[invalid] style_emoticons/default/smile.gif)

Soo, I guess nothing is obviously broken at any rate?

I couldn't find anything that looked especially weird, but feel free to check my figures if you like...