Encrypted Text: The fuzzy math of theorycrafting
Every Wednesday, Chase Christian of Encrypted Text invites you to enter the world of shadows, as we explore the secrets and mechanics of the Rogue class. This week, we'll be talking about the methods used to break down gear and talents into their fundamental parts, and compare them intelligently.While it's not recognized as an official term by the Oxford Dictionary, "theorycrafting" is definitely the biggest buzzword in the Rogue community. Originally coined by Starcraft players looking to use their mathematics knowledge to perfect their strategies and unit build orders, it refers to the idea of using math to guide your choices, instead of simply playing from your gut.
From the lowest level of forum troll to the most serious raiders, many Rogues love to punch numbers into spreadsheets and talk about the PPM (proc-per-minute) chance of Mongoose. Theorycrafting is a largely arcane art, and unfortunately an inability to play ball with these math magicians can be grounds for ostracism from the discussion at hand. Asking for talent or skill help on the Public Rogue Forums will likely result in recommendations like "Check the spreadsheet or delete your character."
Obviously you've become attached to your assassin of the shadows, so after the cut we'll talk about how to become a theorycrafter even if you weren't first in your algebra class.
The basic concept behind theorycrafting (will now be referred to as "TC'ing") is to minimize or maximize something. You'll hear it called "min/maxing"; the idea is to raise a desired value and to decrease an undesirable value. An example is Attack Power: it is often considered the most important statistic for a PvP Rogue and can generally be used to gauge the gear level of the Rogue in question. Something you may wish to minimize is your chance to miss: hitting more often is the core behind every Combat raider's gear and talent choices.
The way to properly determine which path to take is to break each path into its fundamental parts, and to then compare the parts. Looking at two items with different stats can become very tricky, so by converting them to the core stats (Attack Power, Critical Strike Rating, etc) you can greatly simplify the process. The hardest part of TC'ing is to properly disassemble the elements while considering things like buffs, debuffs, party members, and the target you're attacking.
A quick example of how to break down stats on items:
Ring of Strength: +15 STR
Ring of Attack Power: +15 AP
For a Rogue, 1 STR is equal to 1 AP and so these rings provide exactly the same benefit, barring outside influences. Using the basic formulae for Attack Power and for Critical Strike will allow you to break down any item in game. The core values for DPS are traditionally thought of as: Attack Power, Critical Strike, Hit, Expertise, Haste, and the newly-revamped Armor Penetration. Let's take a very controversial example of gear choices and look at the theorycrafting that goes on behind the scenes.
Two new sets of leather PvP bracers became available for Honor Points in patch 2.4.2, Guardian's Leather Bracers and Guardian's Dragonhide Bracers. They're available to both Rogues and Druids, with the first being tailored towards Rogues and the latter designed for our kitty cat friends. You'll see a large split between Rogues on which they decide to pick up.
In order to evaluate what pair trumps the other, we'll have to use the basic stat formulae that I mentioned above. I will break down this first example in detail. 1 point of Agility converts to 1 Attack Power and 0.552 Critical Strike Rating. 1 point of Strength converts to 1 Attack Power. We convert these stats, and then add them up together.
Dragonhide: 50 AP, 33 STAM, 17 RES, 38 CRIT, Red Socket w/ 2 RES bonus.
Leather: 60 AP, 37 STAM, 15 RES, 30.5 CRIT, Yellow Socket w/ 2 RES bonus.
The simplest way to start is to only focus on the differences between the sets. So, we'll remove all the stats that are the same on both items and remove the defensive stats (stamina and resilience), because they are similar and most Rogues focus on damage anyway. To make comparison easier, let's consider a situation where a Rogue is playing by himself, with no outside buffs or influences. Here's what the difference comes out to be:
Dragonhide: 7.5 CRIT, Red Socket
Leather: 10 AP, Yellow Socket
If we use Blizzard's iLVL scale, AP is typically valued at 2x the value of Critical Strike Rating, as seen by the gems for AP when comparied with the gems for Critical Strike Rating. This would put the Dragonhide Bracers significantly ahead, as the 7.5 Critical Strike Rating is worth 15 AP. However, this scale is largely inaccurate: it fails to compensate for talents, buffs, and specific playstyles.
For example, many Rogues will have the talent Sinister Calling, which increases their Agility by 15%; and also Deadliness, which increases their Attack Power by 10%. Considering that the Guardian's Leather Bracers have more Agility and pure Attack Power, this will mean they see increased benefit from the talents. Let's reexamine both pieces from the perspective of a Rogue with these talents. I converted Strength and Agility to their raw stats beforehand, removed the unimportant defensive stats, and factored in the two talents. The difference between them comes out to be:
Dragonhide: 7.25 CRIT
Leather: 11.5 AP
You'll see that while the AP and Critical Strike Rating on the Leather Bracers scaled better with the deep Subtlety talents. Now, in order to determine which is more valuable, the AP or the Crit, you'll have to figure out which increases your overall DPS by a greater amount. The key is to remember to evaluate your talents, buffs, and opponents when TC'ing to ensure that your results point to the correct choice. For example, a player that takes less damage from Critical Strikes will make AP more valuable; while a Rogue who has talents that proc on crits like Seal Fate and the brand new Focused Attacks will see the balance ratio tip away from AP.
Focusing on every stat and hoping you remember which talents affect what and so forth can get out of hand very quickly. Luckily, some of the fantastic Rogues over at Elitist Jerks (my favorite community for theorycrafting) have worked up a DPS Spreadsheet. What this allows you to do is easily plug in your gear and talents, and it will do all the math for you. You can then experiment changing talents, regemming, and enchant choices; it will show you if your damage will go up or down. This is perfect for trying to decide whether a piece of a gear is an upgrade for you, or which piece to upgrade first.
It does take some getting used to and some initial startup time, but once you have your character's information entered into the spreadsheet, it is trivial to test new builds and gear without even having to log in. While none of this theorycrafting stuff is worth more than actual in-game testing, it can give you a very good idea of which way to go when deciding between Find Weakness and Focused Attacks or between buying your Season 2 helm or chest first.
Ultimately, if your goal is to do the best DPS you can, you're going to have to theorycraft every piece of gear, every talent choice, and every ability that you use at any given time. WoW can easily become a math game, and unfortunately for us Rogues, there is a "right answer". However, it's also important to remember that it IS a game, and that sometimes you have to worry less about that last 1% of DPS and start trying to have some fun.
So what if that sword that dropped is an upgrade: you like the look of your Heartless! I'll admit that my Warlock used the Argent Avenger with a Strength Enchant on it; and I took that sword and raided from Molten Core to Ahn'Qiraj. I got invited to more raids because I was "that Warlock who uses Rogue weapons" rather than because I did more DPS than anyone on the server. Try to balance being "all you can be" against "have some fun" and I'm sure you'll find a happy medium where your guild is content with your damage and you get to play in a way that you actually enjoy.
Filed under: Rogue, Analysis / Opinion, How-tos, Raiding, Classes, Talents, (Rogue) Encrypted Text






Reader Comments (Page 1 of 2)
echave Oct 22nd 2008 11:10AM
Chase-- while I haven't had time to read this thoroughly, it's a refreshing change of pace to have someone frequently posting to WoW Insider articles of some depth. While I'm a relative noob in-game and don't feel qualified to judge on mechanics or other's interpretation of them, your past few articles have given me plenty of food for thought. Thanks.
edenxiii Oct 22nd 2008 11:40AM
I agree, Chase is definitely one of the better writers on staff. He really knows his stuff and takes the time to back up his claims with the real facts driving them. Keep up the good work my man :D
MoonHemo Oct 22nd 2008 3:06PM
theorycraft ftw
know the right answer and have perfect play
jabba Oct 22nd 2008 11:42AM
I have to agree-as a fellow rogue for some time now its been hard to come across any new information or even in-depth articles that are new/relevant (possibly due to all the previous TC that has occurred).
Ive recently started a shammy but im really getting back the interest in my rogue from reading some of your articles and ofc the new talent changes :D
Keep up the good work and I look forward to reading in the future! :D
jabbawonga EU Karazhan
fuzzbox Oct 22nd 2008 11:54AM
unfortunately it hasn't been updated for the 3.0.2 patch yet, but you forgot to mention a GREAT theorycrafting tool- www.warcrafter.net.
Great for subbing in gems, equipment, enchants, and seeing exactly how they'll change your stats without having to actually make the calculations yourself.
Worcester Oct 22nd 2008 12:15PM
I've always turned to www.shadowpanther.net for my theorycrafting... especially since I stink at math. I've also set up some scale tags using the Pawn addon based on SP's formula. The author is still maintaining the site, as you'll see several updates in just the last few days.
I'll have to look at warcrafter.net, though. It sounds interesting.
Subtlety FTW!
Grubba Oct 22nd 2008 12:31PM
The oft-cited problem with shadowpanther is that it assigns static stat values to rogue gear, when that is not how it actually works. The real value of stats depends on a more complex system of variables, such as talent build, group composition (or lack of it), and the amount of other stats that you have.
As a reference tool for raw data like item lists and their stats, it's a fine tool. For valuation purposes, it is too oversimplified.
Worcester Oct 22nd 2008 1:06PM
I've recognized that it has limitations, but for my limited concept of theorycrafting it works well for my needs. Also, the scale tags I use have formulas for PvE and PvP values for MH, OH, and Dagger builds. Taken as a whole, the mixed values can at the very least point you in the right direction for an upgrade.
While it might not be as precise or in-depth as Elitist Jerks, I find it very helpful for a more casual playstyle. It helps take the guess-work out of gear-selection without needing a math degree.
Doch Oct 22nd 2008 4:19PM
Warcrafter.net is an awesome site/tool, unfortunately it has not been updated yet, but should be in the next couple of days.
Grubba Oct 22nd 2008 12:23PM
Good article as an introduction to theorycrafting to people unfamiliar with the term or the concept.
The only nitpick I have with it is saying that that it has less value than in-game testing. Assuming that the formulae used in the spreadsheet are correct, it is a more valuable discovery tool than live testing, because the numerous random variables that can affect live testing (lucky--or unlucky--crit streaks, player error, distractions, etc.) are accounted for and normalized on the spreadsheet.
If you want to make the argument that playing how you want is more fun than playing by the numbers, that's another argument entirely, but for the types of players who are going to go run DPS tests to maximize their damage, I think your statement of value is inaccurate.
madsushi Oct 22nd 2008 12:31PM
I agree with your points, especially on the topic of random procs and so forth. A true DPS test in-game will require a very large sample of data to be accurate (especially when you want it VERY accurate) and can require you to find 5 guys to buff you all the time. Often, in-game DPS tests are done poorly, resulting in invalid data.
The key reason behind my statement that in-game testing is more important than TC'ing is that I don't honestly believe there is a 100% accurate spreadsheet out there, as Blizzard never reveals every aspect of gameplay and mechanics to us. Testing damage on a live realm (with the proper testing methods, sample size, etc) will ensure that any small change that Blizzard has made will not go unnoticed. The sheets are very very accurate, but running these numbers through Blizzard's formulae is the truest test (if done properly).
Thank you for your comment, and reminding me that it's important to know that math doesn't lie, but poor testing can. :)
Grubba Oct 22nd 2008 5:33PM
Your point about not having a spreadsheet that is 100% accurate to Blizzard's own calculations is a good one. For all that has been done, calculated, theorized and even occasionally blue-confirmed, there are still some areas where we're still working in a black box.
I would only add that the mechanical work done on the EJ rogue spreadsheets is largely the result of exactly the kind of extensive in-game testing that you describe.
Sabotage Oct 22nd 2008 12:23PM
Excellent post Chase. Glad to see that the Rogue community has a skilled writer with a brain.
Jorad Oct 22nd 2008 2:35PM
Great to see a complimentary view of TC.
I know many people who snub their noses when someone begins to talk about the TC stuff and actually evaluating Talents with numbers behind it.
(Of course these same people then choose their talent specs by copying High-End-Guilds who looked up the theory for sure...)
armi0022 Oct 22nd 2008 12:38PM
Good article that steers rogues in the right direction for gear evaluation.
Slighted- 70 UD Rogue
Lollumad Oct 22nd 2008 12:41PM
great post.
i enjoy reading all of them.
very helpful =]
good work!!!
ravenice Oct 22nd 2008 12:46PM
Great post! The only caveat I would add, is that Theorycrafting, like any tool, should be preceded by *WHAT YOU ARE TRYING TO DO*. Your ends necessarily dictate how you min/max. Do you plan to raid 10-mans? 25-mans? Run heroic 5-mans? PvP Battlegrounds? PvP arenas? Which bracket? Which comp? Plan to solo? Dual w/ your girlfriend? Farm? One of the best way to flag yourself as a noob is to ask the question "What's the best spec, gear, weapon for XYZ?", without specifying what you plan to be doing.
To give just an example of the top of my head: although the Ring of Strength and Ring of AP are statistically identical, the assumption is that there are no other influences (which is usually untrue). For example, if your rogue plans to raid on a regular basis, the Ring of Strength is clearly the better choice because the strength stat scales with Mark of the Wild (from druids) and Blessing of Kings (from paladins). Certain talents (like Sinister Calling for Rogues) scale on stats, while others scale on AP or pure percent damage.
superfrank Oct 22nd 2008 12:57PM
Mark of the Wild does not scale anything - it only adds on a fixed amount of the stats unlike blessing of kings which increases stats by (up to) 10%.
So don't take this the wrong way but you've just "flagged yourself as a noob"
ravenice Oct 22nd 2008 1:32PM
Thanks, superfrank. You've caught a factual error that doesn't invalidate my point at all. You win a cookie.
Bunyipee Oct 22nd 2008 1:46PM
I have to say I disagree a bit with your list of situations.
Once you get down to it there are 3 ways to play
1. pvp
2. raiding
3. solo/do a bit of all.
for raiding it really doesnt matter whether you are in 5/10/25 content the basic setup and stats you need are the same. I would say its the same for pvp. for solo it really doesnt matter you probably want to balance all stats, though I would say that solo builds are for lvling alone.