Some theory – weapon and shield selection

One of the early decisions for a lot of melee – heck, let’s remember I’ve got a shaman and stay with that tight a focus – is whether it’s better to go for one hand weapon plus shield or for two hand weapon.

I wrote, and just killed, a huge post.  I hope you’re happy (grin).  Let’s do this straight by starting with the conclusion.

The general rule is that you are better served by taking the two hand weapon over one hand and shield.  CAVEATS:  Talents and enchantments, of course.  And assuming we’re talking a pair of weapons of nominal equal level.  And if your shield has an armor value of 8.5% or better of your armor cap, you’re probably better off with the one hand and shield.  Finally, and not insignificant, I’m talking levels below 40.  Now let’s walk the numbers, ok?

Let’s start with a finding I made.  In general, when you compare sibling weapons (ie, a one vs two hand of equal level and type) the two-hander will have approximately 25% more DPS than the one hander.  That means (in reversal) that the one hander has approximately 2/3 the DPS of the two hander.  (tighter example.  If the one hander has a DPS of ~15, it’s sibling two hander’s DPS will be approximately 20.)

What this means is that we can kill things approximately 25% faster with the two hander than with the one hander.  BUT… we are giving up armor to do so.  Now, people get really confused about armor.  See, a given armor value adds less armor mitigation as it’s added to higher levels of preceding armor.  However, a given armor value adds approximately the same amount of bonus time to live (time till you die) regardless of whether it’s your first armor or it’s what takes you to the cap.  In this situation, the time to live is what matters.

If I kill things 25% faster, then this is good if my TTL decreases by less than 25%, and bad if my TTL decreases by more than 25%.  Fortunately this turns out to be an easy thing to check.

Assume your base time to live is at zero armor.  At your armor cap – 75% armor mitigation – your TTL is 400% your base TTL.  Now since the increase is 300%, and since armor points increase TTL linearly, we do not have to do any fancy calculations to figure out what reduces us by 25% TTL.  In fact…

Armor cap for levels 1-59 is simplified as 1455 * level (of attacker).  That is, a level 20 facing other level 20s needs 29100 armor points to have 75% damage reduction which is 400% base TTL.  Now for the magic trick…

29100 / 300 = 97.   So every 97 armor points is a 1% change in TTL for a level 20.

Using the two hander kills 25% faster.  25 * 97 means that unless my shield has 2425 armor points, I am better off with the two hander — My increase in killing the mob is greater than it’s increase in speed of killing me.  oh, by the way… 2425/29100 = 8.3333…%   And final by the way – as a level 20, the “best” armor value shield you can have is the crest of Darkshire, with a whopping 661 Armor.  That’s a 6.8% TTL change.

OK, let’s get into all those caveats.

Talents and enchantments, which include the bonuses on the weapon and the shield.  The biggest and most obvious is with a two hander you get one enchantment slot, while with one-hand and shield you potentially get two.  This might not be insignificant, though it mostly boils down to modifying one of two values – speed of killing and time to live.  The latter by increasing your stamina or your armor, the former by increasing your damage potential.  Pretty easy, right?  Talents… whoo, boy.  Pick up a talent that adds damage to one handers, and it definitely changes things.  Same deal is, for example, being a warrior who fights in defense stance.  The shield matters, suddenly.

Pretty obvious is that this doesn’t apply if your weapons are of equivalent level.  I mean, if you’ve got a twink dagger and a basic green two-handed axe, the dagger’s DPS may be a far and away winner.  But I expect you to be reasonable here.

Also, and coming back to the talents and such, once you break 40 your abilities — and more importantly the synergy of all the pieces — may make the base rule inapplicable.  Of course, once you’re level 60 or better the armor rules change again, too.  I’m going to leave that alone for now – I’m low level on this character, I’m just going to do the low level work.

Final caveat not mentioned above – all this is assuming solo work.  When you’re in a group, sometimes your GROUP will work better if you give up your uber ability.  That slight edge in TTL may mean the group blasts through mobs faster since there’s less stopping for the priest to regain mana, spent racing to keep you the tank alive…  yeah.  There is no hard and fast rule, here.  Recognize the exceptions and adjust accordingly.

But still… in general, for solo work two handed weapon beats one hand plus shield.  Just something to consider…

Go have fun.


~ by Kirk on May 13, 2008.

7 Responses to “Some theory – weapon and shield selection”

  1. But also consider in a group that if you aren’t the tank and you control your threat you shouldn’t be taking damage, so the more you can output the less the Tank will take, thus saving the healer mana. Interesting assessment but as a Warrior I would probably always stick with 2hand or dual wield. As a Shaman it becomes more interesting because you are in leather armor and thus have a much lower TTL then a plate wearer. Even though you stated TTL is linear the difference between 200% base TTL and 230% base TTL is greater then 300% to 330%. So while the time it takes to kill you increases linearly (that is how I read what you posted) the % increase in TTL is not linear.

  2. Bernard, reread. If you’re the tank, use the shield (probably). If not the tank, you do more dps without. You’re agreeing with me there.

    Dual wield is next post.

    And no, as I said – armor mitigation is confusing. The mitigation percentage improvement for a piece of armor is lower when adding to 250% than adding to 350%. But the TTL increase is linear for that piece, regardless of what amount of armor it’s added to. See for a basic discussion.

  3. I just leveled my enhancement shaman to 41 and I can say that an increase of 25% of the DPS of my weapon is not an increase of 25% of my DPS. Shocks, lightning shiled, searing totem…all of these things are part of my overall DPS. In fact a 25% increase in my weapon DPS is probably a lot closer to 10% increase in my DPS. The slower weapon speed also means less crits which reduce the mana cost of my shocks. Without giving it a ton of math I chose 1 handed and shield up to 40(dual weild) for these reasons.

  4. ooookay. Intro to theorycrafting…

    EVERYTHING MATTERS. BUT… you must take each piece in isolation to determine its optimal situation first, THEN combine them.

    By the way, the bit about crits… Let’s assume for a moment that you have a choice between two cousin weapons, a one-hand with weapons speed 2 that does ~100 dps, and a two hand with weapon speed 3 that does 133 dps. (numbers are nominally correct for cousin ratios, though there’s some wiggle in specific items.) You have a crit rate of 10% just for this example.

    Your one-handed weapon does an average of 200 points of damage, and a crit will do 400. Your two handed weapon will do an average of 399 – let’s round it to 400 – points of damage. It crits for 800 points of damage.

    Over one minute of fighting, you will swing your one handed weapon 30 times. That will give you an average of 27 normal and 3 critical hits. 5400 + 1200 = 6600 damage over the minute.

    Over one minute of fighting, you will swing your two handed weapon 20 times. That will give you an average of 18 normal and 2 critical hits. 7200 + 1600 is 8800 damage over the minute. The 1:1.33 ratio is sustained even when including crits.

    For weapon choice alone, BEFORE including supplemental and complementary effects, Dual does more damage per time period.

  5. Hmm, wowwiki claims that the armor formula is:

    DR% = Armor / (Armor + 400 + 85 * AttackerLvl)

    Which, for level 20, would work out to an armor cap of 6300. Wowwiki also claims the cap against level 70 is 31,672. So to live 25% longer (the difference in time to kill between *equivalent* one-handed and two-handed weapons), you’d need 1/12th of that amount of armor from a shield, or about 525 armor.

    Of course, this assumes that you’re equally happy killing a mob at 80% of the speed that you would with a 2H weapon, and that you have equivalent one-handed and two-handed weapons available.

    It also assumes dual-wield is not an option, since dual-wield generally does equal DPS with 2H without any damage-boost or to-hit talents. With to-hit or extra off-hand damage, it pulls ahead substantially in many cases.

  6. […] appropriate response to Calandris Rereading my response to Calandris in the 2h vs 1+Shield post, I realize I came off a bit brusk.  Worse, it looks like I ignored a big chunk of what he said.  […]

  7. Calandris, I can understand your reasoning for a 1-hander from 20-29. But I don’t see it post 30 if you’re actually relying on meleeing.

    Windfury is huge, and it only costs mana once every 30 minutes. Between its 3-second cooldown and its lack of normalization (your AP gets converted to damage at the speed of the weapon you are actually wielding, as opposed to a convenient stand-in number) it encourages big, slow weapons, especially since the extra swings also get a non-normalized AP boost. Ideally you’d pick one up with a speed of 3.9 or higher, since that would let you have a Windfury chance every swing even with Flurry up.

    Of course, there’s only a couple weapons like that in the under level 40 range, one of which is about a 1/6 chance from a rare mob on a 20+ hour respawn timer and the other requires level 39.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: