Defining Good Rotations – Part I

This is the first in a series of posts discussing our tanking rotation.  In the first two parts, we’ll be investigating the topic in an abstract sense to try and describe the features that make a given rotation enjoyable and interesting.  In the finale, we’ll look at paladin rotation mechanics in Cataclysm and see how closely they match up to these ideals.

The dreaded Exclamation Point of Damocles.

And at that moment, Neckbeard realized he had become what he most feared... an NPC.

Interactions with the virtual world are important.  They’re what separates the passive experience of a movie from the interactive experience of a video game, and they breathe life into what would otherwise be a fairly static and uninteresting environment.  These interactions can take many forms: “talking” with NPCs, harvesting resource nodes, or making Chuck Norris jokes in trade chat all qualify.  But when we talk about a raid encounter, most of the interaction takes place in the form of a spell rotation we use to beat the stuffing out of a boss mob.

We’re using a loose definition of “rotation” here, in that we’re including everything from predictable and repeatable cast sequences to priority queues and everything in-between.  The key is that your rotation represents the spells you cast and the decisions that go into casting them.  The rotation is the primary ways your character interacts with the denizens of the game world.  If you think about it, most of us have probably spent more time bashing in skulls than we have navigating NPC dialog trees or selling junk items to vendors.

As such, it’s important that the rotation is well-designed.  If the rotation is tedious and boring, game play will feel similarly tedious, and people won’t want to play that class/spec.  On the other hand, an engaging and enjoyable rotation will keep players interested and involved in their character, and they’ll have more fun as a result.

But what goes into designing a good rotation?  It’s difficult to unambiguously define “good” in this context, because you’ll get different answers from different players.  However, I think it’s safe to say that there are at least a few major building blocks that should be considered when evaluating the quality of a rotation’s design.

  • Tempo/Pacing
  • Interactivity
  • Resource Management
  • High Skill Cap

In today’s installment we’ll discuss the first two concepts, which are closely related.

Be happy we're not using the "Freebird" strategy for this boss.

During the burn phase, we're going to segue into "All Along the Watchtower"...

Tempo / Pacing

Having an appropriate tempo is one of the most important design considerations of any rotation.  If the pace of the rotation is too slow it can feel boring and stale, and your brain might start dozing off during the pauses.  Having too many empty global cooldowns (GCDs) is a simple example of how this can happen, but it can occur in other more subtle ways as well.  If warriors had to wait 20 seconds between Shield Slams, for example, their rotation might feel tedious.  By keeping the period of the cycle short, the player’s mind can’t wander off; they’re subconsciously thinking ahead several GCDs, looking towards that next Shield Slam.

The pacing shouldn’t be too fast either, otherwise it starts to dominate the player’s attention at the expense of other jobs.  With all of the things we’re keeping track of at any given time in a raid encounter, nobody can afford to tunnel-vision their rotation.  If the rotation forces the player to make difficult decisions too frequently, there won’t be enough “mental bandwidth” left to handle these other tasks.  The GCD is one mechanism that helps prevent this from occurring, keeping the time between decisions relatively long.  But there are still off-GCD abilities that can increase the pacing (Heroic Strike, for example).

One of the things I vividly remember about learning to play my warrior in Wrath was the extra attention I had to pay to slip Heroic Strikes in between GCDs.  It was effectively like having a shorter GCD, because at the half-step in-between every GCD I was checking my rage and making another decision – should I Heroic Strike or not?  It made tanking on the warrior feel very different than tanking on Theck.  The warrior felt more frenetic for a couple of reasons.  Not only did the rotation have these off-GCD decisions being made in-between the usual GCD-based casts, it also had a shorter cycle time (Shield Slam – RevengeDevastate – Devastate – repeat) and proc mechanics which could reduce that cycle even further (Sword & Board).

By comparison, tanking on Theck was largely GCD-based rote repetition.  The 969 rotation gave us a very strong heartbeat-like pattern to follow, and muscle memory ensured that I could perform it almost without thinking.  So it felt slow and melodious in comparison to the warrior.  I never felt that it was tedious – having something to cast every GCD prevented that – but it was so predictable and low-maintenance that it felt like a lazy Sunday afternoon to the warrior’s busy Monday morning.


I have Crusader Strike bound to Y.

AFK a sec, I'm on auto-tank.

As I said earlier, the rotation is one of the primary ways you interact with the game world.  But if you’re just hammering out a pre-defined sequence of abilities, it’s not so much interactive as it is active.  When you can distill your rotation down to a /castsequence macro, you’re less a player than a robot mindlessly executing a subroutine.  969 was notorious for being simple enough to support this level of automation, and it’s one of the reasons that Blizzard heavily re-designed the prot rotation for Cataclysm.

A “good” rotation should require the player not just to act but to react, adjusting ability usage on-the-fly to accommodate new developments and respond to events.  There are a number of ways this can be woven into a rotation.  Mechanics like Revenge or Overpower, which reward the player for responding to combat table outcomes, are possibly the most intuitive and straightforward method.  And of course there are ways to encourage interactivity based on the player’s resource model, a topic we’ll discuss in more detail next time.  But the most common mechanic by far is a simple probabilistic one, where there’s a small chance that some event causes a disruption in your routine.  We call these “proc” mechanics, a name that has its roots in early online MUDs.

No matter how that interactivity is implemented, it has to be an important part of the rotation.  If the player can completely ignore the interactive element with little or no penalty, it’s not being very effective.  A player shouldn’t just be happy that a proc occurred, for example, she should be excited that it occurred because it allows them to do something fun that they either couldn’t do otherwise or couldn’t do as frequently.  Sword & Board and Divine Purpose are good examples of a simple proc done right. “I get to smash something in the face for massive damage again for free?  Don’t mind if I do!”

Interactivity and tempo are obviously related – forcing the player to react to unpredictable events more often increases the tempo even if the rate of spell casts hasn’t changed.  It’s arguably a better way to boost the tempo than simply decreasing the GCD,  adding off-GCD abilities, or asking the player to juggle a larger number of spells in their primary rotation.  Making reaction-based decisions about what to cast also adds depth to the rotation and rewards the player for being attentive.

There is a limit to how much interactivity you can inject into a rotation though.  If there are too many procs to watch, it starts taking more attention and cuts into the available mental bandwidth for other tasks.  Since WoW has a huge variety of player types in its user base, care has to be taken to make sure that the rotation isn’t so difficult that the average player can’t function.  The ideal situation is to implement the extra interactivity in such a way that it rewards the excellent player somehow without crippling the average player.  This ties in closely to the way the rotation scales with player skill, a point we’ll talk about next time.


Stay tuned for Part II, which will tackle the more subtle issues of resource management and skill-based scaling.  In the meantime, what sort of tempo and level of interaction do you feel is right for a tanking rotation?  What proc mechanics do you like or dislike, and why?

This entry was posted in Theck's Pounding Headaches, Theorycrafting and tagged , , , , . Bookmark the permalink.

One Response to Defining Good Rotations – Part I

  1. Terve says:

    Greats series of posts relating to rotations and game play in general. Sometimes I get too caught up in the moment to sit back and think about the big picture – very refreshing article.

    Thanks for all the wonderful posts… without them I might have to… *shudder* …work!

Leave a Reply