Nov 18

Theorycrafting: Cautions & Fallacies

For a long time now, I’ve been in to theorycrafting, and I always tend to do my own math to figure things out even when someone else has done so before me, if only to confirm they are accurate (or to practice my own ability, to either try to reach the same result or understand how we differed – I’m not a “natural” mathematician). You might be surprised how often I find an idea or assumption made that I can appreciate the logic that led someone to that conclusion, but find it not true to reality at all. There are times though that it’s not all about maths. Sometimes, the maths takes us off course, misdirecting us while we focus on it instead of the reality.

The question may become more philosophical in nature – is there a way to determine when maths won’t cut it?

First of all, I need to point out that I’m not attacking anyone or claiming anyone is necessarily wrong. There is a tremendous amount of value in people who look at extremes, and it’s OK to look at the “perfect” numbers to come up with. Generally, the people who do this don’t tell you that you’re a fool for thinking that the reality is not quite so intense as they are describing, in one way or another. I’m going to give a couple of examples of fallacious thinking – in both cases, beyond thinking those people thinking this way are wrong, I’ve held no ill will towards them, and generally still had a lot of respect for them.


The More Haste, Less Mana argument

When I first got in to really reading and understanding how to play a druid – somewhere around Ulduar/ToGC – I was an avid reader of all the usual sources for theorycrafting. One thing I couldn’t get my head around was what so many people were saying – having higher haste means more mana consumption.

empty-vialNow, I can understand how someone gets there logically, but it’s just totally fallacious. If you ever read my post about understanding healing meters, you’ll know that the amount of healing you can possibly do at any point in time is dependent on the amount of damage that has been sustained previously. All healers in your group are then vying for a “share” of healing that damage.

At the end of the day, your mana will be spent either healing damage sustained by the raid, or over-healing. If you have low over-healing, the additional healing you’re doing is either healing an amount that otherwise would have been healed by someone else (resulting maybe in over-healing on their part that might not have been otherwise) or allowing you to keep up with damage that you otherwise couldn’t have (i.e. people would have died).

Now, that’s not to say there’s no benefit to that – by being the greedy healer who grabs all the damage before your co-healers, you take the pressure off them and might find you can afford to have a healer switch to DPS. However, the point is, you aren’t spending more mana for the same amount of healing, unless you’re wasting it by over-healing.

Of course, this would be a legitimate drawback if fights went on long enough that going out-of-mana were a certainty, and you had to maintain maximum healing throughout the course of the fight. Not many fights in WoW are like that though.


The Run Speed = Throughput Debate

The first time I remember this coming up, I’m pretty sure my guild was progression on Bastion of Twilight, back in Cataclysm. Back then, there were enchants that gave you a a pure increase to a stat – say +30 mastery (I honestly don’t remember the numbers and it was pre-item squish). There were also enchants that sacrificed some amount of that stat for an increase to run speed – say +26 mastery & 10% run speed. A fellow theorycrafting guildy explained that the run speed equated to a throughput increase, because every time he had to move out of something, he started casting a little earlier, and those fractions of a second sooner he began casting added up to additional casts over the course of a fight. I told him the run speed should only be considered in PvP.

Image: Wowhead

Image: Wowhead

Ignoring that this applies least to resto druids than anyone thanks to our mobility while healing, it is wrong for any class. The reason for this is, there is no fight where the movement mechanics line up cleanly with the end of a cast every time – ie that you have exactly enough time for five three-second casts and as soon as the final cast ends, you are dodging the void zone perfectly. At some point, there is going to be a need to interrupt a cast to move. On some inter-mechanic gaps, you might end up with an extra cast over someone who has the higher stat. On many, you will probably only do the same number, only that yours would be a bit weaker. Overall, you could seek out every situation in the game and there may be one in a particular fight where you can reliably say that the movement speed reliable pays off with an extra cast, however the real world difference is totally negligible.

Of course, he would have been right, if only you could cast the same spells in the same order  to heal unlimited damage and move the same distance at the same interval, so long as that interval matched the total cast time. Boss fights aren’t like that though.


Extreme Cases vs Reasoned Adjustments

I’ve been looking a lot at Haste vs Mastery stat priorities lately, really trying to make sure I have the best understanding possible.

Take Mastery as an extreme case. People (including me) always theorycraft Mastery based on 100% Harmony uptime. Your Harmony up-time might not be that high, and that drastically changes the value of the stat for you in that particular situation.

Let’s compare this example with the effects of Haste on Rejuvenation. Is it fair to apply the same reasoning – that we must look at the extreme case, that Haste increases the full duration of the HoT and grants you an extra tick? Not as I see it. The reason being, Rejuvenation is a 15 second HoT. While I appreciate the intricate skill that comprises predictive healing, you can’t predict 15 seconds in to the future. Rejuvenation will, typically, be subject to more over-healing than any other spell when one is being discerning with their healing choices. Even when you have great chemistry within your healing team (there’s another old post of mine on the subject), your co-healers will generally react to a health deficit by healing, and more often than not, you’ll probably be prioritizing the same targets in a raid healing situation. There’s also smart heals. Even if you pick your Rejuvenations perfectly so that they won’t overheal, and your raid group has perfect chemistry such that the other healers don’t case when Rejuv is on a target, smart heals will do their own thing.

In the extreme case, Rejuvenation may benefit from Haste by an extreme amount. Making a reasoned adjustment though, you might realise that you’re rarely likely to get the full benefit of that increase, and as a result you don’t consider Haste significantly better than Mastery.


Thoughts & Discussion

Where I think we get in to the realms of philosophy is determining where one draws that line in the sand. Why is OK to regard some things in the extreme cases, where perfect utilization is required, and others are not OK? In the case of Mastery vs Haste, I believe that the answer lies in terms of control. Mastery up-time is 100% the responsibility of the individual playing. If you aren’t maintaining full uptime, that’s because you didn’t re-activate it with a direct heal before it expired. It is within your power and your power alone to change that. Rejuvenation, on the other hand, is open to influence from external variables, including how the other healers regard your HoTs on their target of choice, and how many smart heals are thrown out.

Don't you dare forget to refresh Harmony!

Don’t you dare forget to refresh Harmony!

Another way to look at it is how much effort is required vs optional. If you’ve raided, you’ll probably be familiar with the idea that you need to do “everything you can” to prepare your character for a raid. Sure, you flask, you eat food, you have pots, and you learn the fine details of how to play. But with some exceptions, you don’t delve down in to your character and gear setup for every individual fight looking for some small edge. You probably don’t rebind hotkeys between bosses, even if a boss has a mechanic that might benefit from your hand being slightly better positioned. You probably don’t investigate the fight looking for some reason that your typical 4th priority stat is in this situation better than your 3rd, and swap a bunch of gear out accordingly. It’s not necessary to play this way – the game is not tuned to this degree. As players, we try to make sure we do it right the first time, and get the best gear for the majority of situations, and usually reserve changes for when we actually get an upgrade.

Sometimes, when we theorycraft, we can go to the degree of certainty – that X is better than Y by Z amount. It’s actually a good thing that in situations like haste, someone goes ahead and does exactly that, and maths it out so that we understand every boundary precisely. However, it’s important to be able to evaluate those extremes in the context of the real world (…of Warcraft).

Remember, people (especially me!) make mistakes all the time. I’m constantly revising things that I say because I find out it’s not quite right. But to my credit, I’m always checking, appreciating how easy it is to make even a simple error.

If you’re wondering which is the better choice right now, I give a slight edge to Haste. The difference is marginal and it varies by situation and by play style. I wouldn’t say you’re wrong to prioritize either over the other, and whichever one loses out is still your next best stat (by a reasonable margin). They are close enough so that a wide range of variables – what are the mechanics of a given fight, what are your healing assignments, what is your raid makeup, etc – all have a much greater impact on one’s personal healing than the difference between the two – when you’ve got 100% Harmony up-time, that is.

Permanent link to this article: http://sometimesatree.com/theorycrafting-cautions-fallacies/


  1. Zoopercat

    Hey, love this post! I was doing some research for an article I’m writing about how Theorycraft affects our choices in game. This article covers a few of the things I would have touched on, and I found myself cheering along as I read each paragraph. I’m going to grab some quotes and link to this article when I post mine. Great job, I’m a new fan 🙂
    -Zoopercat / @AskMrRobot

    1. Breadfan

      Thank you kindly 🙂 I look forward to reading yours!

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>