Thursday, December 17, 2015

Rotations as a Proof of Competence

Continuing on from the previous discussion of Proof of Competence, asking for spec is not the best question. It's a decent question. It's easy to answer and verify. And it does tend to give the information you're looking for. But there are better questions.

In particular, asking for rotation is a much better question. Someone who can rattle off the correct rotation is very likely to be a decent player.  (Rotation here also means ability priority list, not just strict rotation.) The problem is that asking for the rotation is a harder question to answer and verify. You can't just inspect the player and see their rotation.

But what if you could?

In WoW, there are a number of mods which will display the next ability you should use. Some also show the next few abilities of the optimum sequence. You still have to hit the right buttons, but these mods basically show you the best theorycrafted rotation as it happens.

What if this type of UI element, this "Rotation Helper", was part of the base game?

The idea here is that the game doesn't fill out the rotation on its own. Instead, you would have a screen where you could drag abilities into a priority list. Essentially construct your own rotation. The rotation you constructed would then be displayed, and you could hit your buttons to match.

Another player could then inspect you, and just like they see your gear today, they could see the rotation loaded in your Rotation Helper. And perhaps you could send and receive rotations from other players or third parties. If Sally is another Retribution paladin in your raid and is doing better than you, perhaps you can ask her for her rotation and compare it to yours, or replace yours with hers.

The hard part, though, is to construct a Rotation Helper which is simple enough to use, but also powerful enough to construct a decent priority list. A lot of abilities require things like "use when this buff reaches 4 stacks" or "use when this debuff is about to drop off". This kind of stuff is easy to do in code, in a programming language, but is harder to create in a GUI. A Rotation Helper is a non-trivial design problem.

This Rotation Helper wouldn't really help with reactive abilities like healing, or tank cooldowns, or interrupts. But it would significantly help with DPS.

Of course, there are concerns that this type of UI element is "playing the game for you". But such helper mods already exist, and a decent amount of high end players use them already. These type of mods help those players perform better, and would probably significantly help weaker players who don't know about them.


  1. The other problem is you may not have a good handle on the class in question. If I had to ask our Warlocks their rotation I'd have no idea what to do with that information, which makes it a poor general metric, despite being a good metric for someone who IS familiar with the class. Either that or you're just comparing it with what's listed on Icy Veins or some such.

    On the bright side, Warcraft Logs actually has a timeline view you can drill down to for a character to figure out what abilities they're pressing when, so if you really wanted this information, you can get it, albeit via external tools.

    Rotation helpers do exist for the more complex classes--I fully admit I've taken EnhaPrio and fixed it up myself locally despite the author having dropped off the face of the planet this expac--but it'll be interesting in Legion if they're as required given that many classes are getting simplified significantly. But I'm capable of modifying the code because I am intimately familiar with my class and nudging bits and pieces is a good exercise. An interesting idea to make such a thing baseline.

    1. Well, I think it's the same issue as asking for the spec. You have to know what the cookie-cutter spec is in order to evaluate. Is an ability priority list that much more information?

      As for logs, that's fine after you've run with the person. But we're looking for an element to make a decision before you run with her.

  2. Honestly this is looking more and more for a solution looking for a problem. In particular, I find this "solution" to be a bad idea in general, as it turns playing your character into playing Guitar Hero. Sorry, I don't play MMOs for that.

    I'm getting the feeling that you got caught in the Blizzard reasoning of getting anyone to "compete" regardless of preparation, reflexes and engagement. I find that the "difficulty level" approach is much better: you have stuff like LFR where the importance of these matters is minimal, you have Mythic, where it's critical. In LFR the proof of competence is more or less irrelevant, in Mythic any guild will test you for weeks before making a decision, so you don't really need a shortcut.

    What you propose is just "gearscore for rotation". Gearscore was a cheap shortcut to evaluate people which turned out to be an almost complete failure, I don't see why this should not end the same.

    1. That's entirely possible. However, the question is not "Will this work 100%?", but "Is this better than asking for spec, or gear score, or achievements?"

      If something like this is not implemented, that's perfectly fine. Players will just insist on cookie-cutter specs instead. Maybe we're okay with that.

      As for the Guitar Hero analogy, maybe that's true. But the same could be said for boss mods and similar. Perhaps one day players will start insisting that you have a Rotation Helper just like they insist you have a BossMod.

  3. Crap: "...looking more and more LIKE a solution..."

  4. Much of what you want is already available in the "Bitten's Spellflash" add-ons for WoW. Here the writer of the add-on fixes the priority list, rather than you being able to set your own; but it does account for different talent choices for instance making different spells available or modifying spell actions.

    Then it's just whack-a-mole (or Guitar Hero, as Helistar puts it). Hit the lit-up button. Even if you know nothing about your DPS class, just following the lights gets you a halfway decent rotation. I'm not saying it's a perfect rotation; and knowing what the spells actually do, and when you might need to hold off using an ability until a particular moment in a fight - that will always make you better. But Bitten's Spellflash can get a noob a long way.

    My quick WoW "indication of competence" is this:

    * Have they appropriate ilvl gear gemmed and enchanted - I can quickly look this up in the armory, which also tells me if they forgot a talent or a glyph.
    * Have they got (or will they get) the voicecomms software we are using.
    * Have they got DBM or similar installed. I simply ask the question. I still need to invite them to the group to check, but if they don't check out, they can easily be booted for lying.

    Those steps can be completed in about 30 seconds, and are usually good enough for pugging.

    1. Yeah, the existing mods were the inspiration for the post, but I'd like a system where you at least started by building your own rotation. And as you say, it would improve performance at the low end significantly.

      Personally, I think your checklist is excellent. Someone who cares enough to keep her gear gemmed and enchanted is usually good enough.

  5. I'm all for rotational suggestions, but you need to know what each action does rather than rely on a third party program to tell you what to do next. Following the preset flashing lights doesn't seem a measure of any kind of competence. Often enough player's mods suddenly don't work, and they're stopped cold and can't play.
    So many factors come into play in group situations. For me, having someone who is fully engaged, watching not only their own play, but what everyone else is doing in addition to the mob activity is the person I want to see in a group.

    1. Yeah, that's why I suggested a Rotation Helper which starts empty. Where the game expects the player to build her own rotation, at least at the start.

      As well, if it's part of the base game, it won't suddenly stop working. It would be part of the dev's testing to make sure it was always working, just like the current stock UI.