109 réponses à ce sujet

[maegi46]

So dex is practically useless to a mage then?

[Musou1776]

Not trying to make things overly complicated, I did not know how glancing works. Like I said at the beginning looked like you had the right answer and then people carried on like you didn't post the math.

I find it annoying to have to sort through all the BS in a thread like this for the facts.

[Musou1776]

Rumination what kind of +crit items are superior in performance to the
better +elemental items? If you compare say duras blue flame I need to come up with 20% chance to do a 50% crit to equal it. What amulet gives +20 crit chance?

[rumination888]

Musou1776 wrote...

Rumination what kind of +crit items are superior in performance to the
better +elemental items? If you compare say duras blue flame I need to come up with 20% chance to do a 50% crit to equal it. What amulet gives +20 crit chance?

There isn't an item that gives +20% crit chance, but you don't need +20% crit chance to beat Dura's Blue Flame.

It has everything to do with your current stats.

For example, if you over +100% elemental damage, over +70% crit damage, and under 20% crit chance, then even an item with +10% crit chance(Eddie's Lucky Talisman) will be better than Dura's Blue Flame. The more bonus damage or critical damage you have, the better crit chance becomes. The more crit chance/crit damage you have, the better bonus damage becomes. Etc, etc.

Modifié par rumination888, 03 avril 2011 - 01:22 .

[Gloxgasm]

Lets say you do D damage a shot. Item 1 adds 20%, so you now do (1.2)D. Item B adds X% crit chance. 50% crit increase.

You have 1.2(D) = 1(D) + X(.5D)
X = 40% crit.


As for the above 2D + (.2 +.1)(1.7)(2D) vs (2.2)D + (.2)(1.7)(2.2D)

leads to 3.02 > 2.948 so indeed 10 extra crit > 20% extra spell damage.

Modifié par Gloxgasm, 03 avril 2011 - 02:23 .

[Lemen]

Grumpy Old Wizard wrote...

You are going to have to lay out your assumptions if you are going to try to do a realistic calculation. How many different spell trees will the mage have and what are they?

Things that affect all damage type spells:
1) the base dps of the equipped staff
2) the Magic stat
3) difficulty mode:some spells can't be used to full effect without damaging your party. You have have to include fewer enemies in the area of effect, delay casting it,  or skip the spell altogether
4) +%crit damage if a critical occurs

NO armor in the game helps ALL damaging spells. The Robe of Cleanliness helps fire, ice, and electricity based spells. It is pretty much impossible to get both the Robe and Final Thought. The most common bonus on mage armor is +%fire damage. +%Spirit gear can be bought in the Gallows.

Assuming you stop investing in Magic at 42 and Willpower at 32 you will have 40 or so points to invest in other stats. Investing in dex and cunning would allow you wear some rogue gear to boost your +%to crit/crit damage, helping ALL spells.

Is that the best way to go?  I certainly can't say but I am willing to try it out. What I do know since I have played a NM mage who pumped Magic up to 89 is that with pumped Magic your damage is still going to suck.

Now, can anyone propose a formula based on all the different factors (including a decrease in the effectiveness of spells that do friendly fire?)

I did some testing and calculations, so I'll have a go. If you don't want to read the boring math, I summarize the results in the end.

Base damage for a mage is given by the formula BaseDmg = StaffDmg + (Magic-10)/2. For example, if your staff damage is 40 and you have 80 magic, your base damage will be 40 + (80-10)/2 = 75. In other words, each point in magic increases your base damage by .5. There doesn't seem to be any diminishing returns on this. All spell damage is a multiplier of this base damage (except for spells like walking bomb which is based on the hp of the target).

With 50% crit dmg, each point in dexterity increases your DPS by .5%. In this case, dexterity is only more effective than magic if your base damage is more than 100 (which is unrealistic, unless you use xp glitch). So if we don't take into account equipment that gives +crit dmg, magic will always be better than dexterity.

But what if you include +crit dmg from equipment?  First, let's assume no rogue gear. I haven't found any belts or amulets that have +crit dmg. The best rings I've found are Etched Ring of the Twins (+9% crit dmg, +4% crit chance) and Three Wolf Boon (+5% crit dmg, +2% crit chance). If equipping both, each point in dex will increase DPS by 0.64% of base damage. Each point in magic will increase base dmg by 0.5, but since you get some crit chance from the two rings, it will actually increase the damage by 0.5*(1+0.06*0.64) = 0.52. Dexterity will now be better than magic if your base dmg is 82 or greater. This can be achieved by Final Thought + 78 magic after gear, which should be possible.

I know the Tiger's Tail staff also increases crit chance/dmg, but the DPS always comes out worse than Final Thought in my calculations.

But what if you do the same calculations, but include rune of valiance (+7 to all attributes)? The DPS increase for each point in dex is now 0.71% of base damage, while damage increase for each point in magic is 0.5*(1+0.13*0.71) = 0.546. In other words, dexterity is more effective than magic if your base damage is 77 or more. This can be achieved by Final Thought + 68 magic after gear.

Now what if you invest in both dex and cun, so that you can use rogue gear? This introduces a LOT of variables, but let's see what we can do. Let's assume you have Final Thought staff, Etched ring of Twins, Three Wolf boon and Four Fingered talisman. The Puzzle Ring might be better than the Three Wolf here, because of all the +crit dmg from other equipment, but there's no way you could afford that and all the other stuff. I also assume Rune of Valiance.
I will assume that you will have base 32 willpower, at least 42 magic, and 40 points to spare. Gear will give +7 to all attributes from rune, +2 to dex and cun from talisman and +4 to magic from staff.

Pure magic build: 82 magic 10 dex 11 cun before gear, 93 magic 19 dex 20 cun after gear.
Dmg: 89 * (1+0.23*0.74) = 104.15

Unfortunately, I have no idea of what +crit chance/dmg a mage can get from rogue armor, especially since the rogue sets will not be available. I know about the Dragonhide Mantle of the Predator, a cuirass with a nice 6/15 chance/dmg, but that will only be available for the last few fights anyway. Can randomly generated rogue gear even get +crit chance/dmg?
Anyways, I'll just go for a guesstimate of crit chance/dmg from armor. The rogue champion set gives 13/36 chance/dmg, so let's imagine you can get up to 10/30 with standard armor:

Dex/cun build: 42 magic 30 dex 31 cun before gear, 53 magic 39 dex 40 cun after gear.
Dmg: 69*(1+0.53*1.24)=114.3

One can even boost the damage a little bit by only getting 27 dex 27 cun (which should be enough to wear the good rogue stuff).

Dex/cun build 2: 49 magic 27 dex 27 cun before gear, 60 magic 36 dex 36 cun after gear.
Dmg: 73*(1+0.50*1.2)=116.8

Remember that these are not certain numbers. I don't know what +crit chance/dmg you can get on rogue gear as a mage, this is just a rough guesstimate.

While the dex/cun gets a little more defense, the pure magi build gets a lot more attack and magic resistance Also, the pure magic build can use whatever gear he wants. With +elemental dmg gear, a pure magic build should be able to outdamage a dex/cun build.

Summary
- Without +crit dmg gear, magic is best.
- With +crit dmg on rings, but using mage armor, mostly magic is best, but it may be worth putting a few points in dex at high levels. Won't be a big difference anyway.
- Using a lot of +crit chance/dmg items and rogue armor, a dex/cun build may outdamage a pure magic build, but only by 10-12%. The pure magic build is still advisable, since they can use whatever gear they want and will outdamage the dex/cun build with +elemental dmg gear.

There may be errors in the calculations (feel free to point out), or variables I haven't thought of , or other wrong stuff but this was the best I could do in 1 hours work

[Diego Vargas]

A fact in DA2 is mages are not as strong as they were in DA1, they are
not going to be clearing the field like they used to in the previous
game. 

One of the problems with the comparisons being made here is people are only comparing synthetic damage averages over time and not the practicality of the spiker playstyle. It is a much harder proof to quantify situational combos of setting up your melees to cleave right at the moment you chain a spike nuke. Anyone who has actually played nightmare DA2 (opposed to excel) as a spike nuker knows the value of burst damage. 

The hardest enemies (elite rogue etc) often use potions which restore a massive chunk of their large HP. This means the slow but steady constant damage of high magic builds has a hard time overcoming this simple enemy AI tactic to turtle heal/stealth. Playing many competitive MMO's we learn that often the best way to overcome a strong healing defense is to go for a chain of crits preventing your enemy from having enough time to react and tank.

Sometimes what might seem more attractive on paper doesnt work out in practice.

Modifié par Diego Vargas, 04 avril 2011 - 03:22 .

[Gloxgasm]

Do you reload each time your spike has failed?

Also, I one up you. Anyone who has played nightmare on a solo mage knows that higher average damage is better.

Crit only seems to be viable after Act 2 when your talent progression starts to fail at raising your damage.

[ezrafetch]

I think to spec solely towards spike damage is kind of foolhardy, as Glox said if your spike comes up short, where's your overall DPS? Having an average of 1000 damage, say, is pretty useless if you're coming up all 1s except for the one spike of 100000 that pushes your average damage up (that is an example).

I think it's one of those things your start pumping at the endgame: once your Magic (and attack) are up to snuff, your Con is up to where you feel comfortable, then it'd be fair to start pumping Dex/Cun for some extra spike damage. The spike damage is really the bonus, at least. Once you're hitting the really bad portions diminishing marginal returns for magic, then I see Dex/Cun being useful.

[Diego Vargas]

Gloxgasm wrote...

Do you reload each time your spike has failed?

Also, I one up you. Anyone who has played nightmare on a solo mage knows that higher average damage is better.

Crit only seems to be viable after Act 2 when your talent progression starts to fail at raising your damage.

Right so you agree you are not making a realistic comparison to what I am talking about.

I am talking about spike team chains, you are tlaking about self flagellation.  The point I don't see you guys getting is mages are absolute crap in DA2 for damage in general.  They will never plow the field like a strong warrior or rogue build, however if you want to do "useful" damage you can pair them with a strong warrior/rogue and get some really effective spikes off.

I've played through the game a few times on nightmare and the crit build with decent magic was by far the most effective.

More practical play, less excel.