I received an email today from a regular correspondent commenting on my recent CFN post on vanity and gear choice. He has been having a running argument with some DK friends about what slot was more important to upgrade first – shoulders or trinket.
The trinkets, my correspondent correctly points out, have a very high Resilience/Conquest Point ratio, and the on-use effect is very potent for burst kills. The shoulders have a lower R/CP ratio, and their secondary stat is Expertise – not something that Frost and Unholy DKs often go for. But invariably, the shoulders got chosen over the trinkets because they looked cooler. They’re the flashy upgrade, not necessarily the smart one.
But here’s the thing – how much do secondary stats matter? Is choosing the flashy upgrade a really bad move? Once you get past the primary stats and Resilience, how much are these secondary stats influencing decisions? I’ve maintained in my gearing guides that you should pick gear with your spec’s preferred secondary stat first, but is that always the case?
MODEL WHAT YOU KNOW
I turned to Ask Mr. Robot for help on DKs, but his stat weights confused me – sometimes Expertise was very good, other times it was very bad, and I couldn’t tell you why. There are nuances to each and every spec, and I realize I don’t know enough about DKs to really answer my correspondent’s question.
But I do know warlocks. And I know warlocks in PvP. So I built my model around warlocks, instead, with the hope that people who know more about other classes can adapt it to suit.
My approach is relatively simple:
- Gear can be evaluated by assigning weights to each primary and secondary statistic and totaling their values. This number is a quantitative representation of the relative value of the piece of gear for each spec.
- This gear value can then be compared to the cost of the gear, allowing you to determine how effective your purchase will be.
- By comparing the value/conquest point ratios between gear, you can determine which piece will be the best value.
Obviously, this method evaluates the full value of the gear, not the incremental upgrade value. You could easily modify it to suit a specific upgrade query by using the difference of the Conquest gear and your current gear instead of the full value of the gear. But for a starting model I think it’s okay to evaluate the full gear.
I took the stat weightings from Ask Mr. Robot because they were 1) easy to find and 2) relatively neutral. (Just keepin’ it real, yo.) I normalized them down to 0-1, and then added Resilience at 1.00, to reflect that I think it is as important as Intellect for PvP. I think it’s fair for all classes to do this – Resilience should be as important as your primary stat, not more, not less.
Once I had the stat weights in place, I then filled in all of the gear with values from the Ruthless Gladiator’s Felweave set. I left Hit in there, because it’s important, but I did skip Spell Penetration because there are several ways you can reach the Spell Pen cap, many of which can depend on your profession. Spell Power also presents a little bit of a tricky stat to weigh, but since weapons and trinkets use SP not Int I kept it in there and used the stats from the robot.
Here’s the spreadsheet if you’d like to follow along.
Applying the weights to the values on the gear, I came up with a normalized value for each spec (rows 23-26). This number doesn’t represent anything in game, really – it’s a relative measure of gear that only makes sense in context of the rest of the gear available.
I tossed in the Conquest Point cost for each item (line 21), which then lets us look at each piece in terms of purchase value it has – how much bang for the buck we get for spending Conquest Points. This is the value we really want to use to compare gear, because it’s obvious to see that some items have more oomph to them than others – that’s why they cost more. No, we want to compare based on what we get for our hard-earned CP.
My first pass through, I thought that I could stop here – getting a number for each item should be sufficient, right?
Well, not quite.
You’ll note that the weapons (MH/OH), trinkets, and pieces with Hit are better values than the set pieces. I think this has to do as much with the set bonuses not being included in each item’s budget, as with the set pieces costing substantially more points. These items also tend to have more sockets, which further diminishes the item’s onboard statistics. They should probably be higher than they are.
I initially ignored enchants, gems, and set bonuses because when you upgrade from PvP set to PvP set, they are static, while the rest of the gear changes around them. That’s the correct way to capture the upgrade model – ignore stats that don’t change or are not important (like Stamina) and only evaluate based on differences.
But this model only evaluates the gear, without reference to what we’re upgrading from. Even a character new to PvP is upgrading from something; they’re not coming into PvP stark naked.
The solution, of course, is to apply the same analysis to the gear currently worn by a character, and then to compare the difference between the two. That difference represents the value of the upgrade, which can then be divided by the cost of the upgrade.
So the correct model is:
- Assign a normalized value to each piece of gear based on stat weights for the spec and the amount of primary/secondary stats on the item. This includes both the current gear and the future gear.
- Get the value of the upgrade by taking the difference between the future gear and the current gear.
- Divide that upgrade value by the cost of the upgrade.
- Compare that value between the different gear slots to determine the most efficient upgrade path.
Since there’s no way to be accurate to any given Warlock’s upgrade situation, I chose the Season 10 Honor gear with S9 weapons as my baseline in the spreadsheet (rows 39-62), and then compared the Conquest set to the Honor set.
See, it’s important to think about gear purchases this way – you’re not purchasing a new set of gloves with +300 Intellect for 1650 Conquest Points, you’re purchasing an upgrade of +37 Intellect from your current gloves for 1650 Conquest Points. The key is to evaluate on the cost of the relative upgrade, not just the cost of the new gear.
Here’s how it looks.
As a key:
- Trinket 1 is the Ruthless Insignia of Dominance, which procs randomly.
- Trinket 2 is the Ruthless Badge of Dominance, which is an on-use “pump” trinket.
- The PvP Trinket is the faction Insignia.
- Ring 1 is the Ruthless Band of Accuracy, which gives Hit.
- Ring 2 is the Ruthless Band of Cruelty, which gives Crit.
When we change the model to evaluate based on upgrade value and not just purchase value, the set pieces – hands, legs, chest, helm, shoulders – are more attractive because we aren’t ignoring their sockets, socket bonuses, and set bonuses when comparing them to other items. We’re now comparing them to earlier versions of the same gear (which correctly nullifies the effects of those bonuses) and allows us to clearly prioritize upgrades.
I think it’s interesting to see how the upgrade lists differ between specs, all driven by the stat weights originally assigned. Items with Hit are preferable to those with Haste, and the relative value of Crit can play havoc with the lower areas of the ranking.
Since Intellect and Resilience are weighted the same, items with a high Resilience/CP ratio (like trinkets) place higher than you might expect. Take a look at the Intellect/CP and Resilience/CP of the Conquest items to see what I mean:
Since this is Warlock gear, Spellpower devalues the MH weapon’s Intellect budget, as well as the trinkets. And the Intellect / CP chart has a strong correlation to the number of sockets on a piece of gear.
But look at how much Resilience those trinkets give for their cost! If you’re trying to gear up for PvP and improve your survivability, trinkets are the way to go. The set pieces weigh in favorably too here, and that’s before the set bonus kicks in.
INTERPRETING THE RESULTS
Now, all these numbers are well and good, but PvP is not just about numbers – it’s about performance. Is the random proc of one trinket better than the on-use effect of another? If we’re to just evaluate it by the numbers, absolutely yes. One gives more value for the CP (and more DPS, incidentally) than the other.
But this is PvP. Affliction warlocks, for example, are not known for their burst, and having an on-demand trinket that can be used to burn down an opponent at the right time is critical.
I think it’s interesting to see how dependent all this is on the stat weights you choose at the beginning. If you were to go on a straight primary stat / Resilience evaluation, you wouldn’t see a lot of the subtle differences in the gear pieces. Small changes to your stat weights will have a big impact on the gear’s purchase values, so these lists should not be set in stone. They are absolutely dynamic and need to be adjusted to your specific situation – which you should do.
Generally speaking, the pieces high on these lists either have a lot of the primary stats warlocks want – Intellect, Spellpower, or Resilience – or they have the secondary stats – Spell Hit, Haste. When all other things are equal, choose the gear with the best secondary stat – but it’s interesting to see that all other things are not always equal.
It’s nice to see that my general rule of thumb holds – get the stuff that has your secondary stat first – but you should also look at the rest of the item’s budget and cost when choosing what to upgrade next, and use your head.
All those disclaimers aside: the MH/OH weapons are always excellent buys.
Getting back to the original question: are players who upgrade their shoulders first doing so out of vanity, or out of logic? If the secondary stat on them isn’t the best one for your spec, how important is that?
I think there’s a strong argument to be made that the shoulders are not a good first buy for warlocks – you get a bigger boost of Resilience from the trinkets.
But there could be other reasons to get them which outweigh the numbers. Perhaps it is important for the player to show others that the character is progressing into the next season of gear. Perhaps the new shoulders are a bigger upgrade due to not having a complete set of last season’s gear. Perhaps the player just likes the look of them! That’s all okay.
It’s okay because eventually, with enough time and focus, you can get all the gear, and it won’t matter which order you got it in.
However, when you are sitting there at the vendor, wondering what to get next – don’t be so quick to upgrade only the visible pieces.
New shoulders are great and all that, but there are other pieces that probably have more bang for the buck.
Like, you know, gloves. :)