Where Gating via ilvl goes wrong

I was thinking about the upcoming LFR, and ilvl requirements, and listening to some feedback from guildmates who have run LFR, and I made the following statement in the email thread that we spam with theoretically blog-related discussion:

I can meet the minimum performance requirements for LFR with a freshly levelled 85 in quest gear – why is there no method for me to prove that and circumvent the 5-man gearing treadmill that I so hate?

What’s so bad about Item Level?

Ilvl suffers the same problems that Gearscore suffered in Wrath – it is a non deterministic indicator of performance.  Blizzard even went to far as to make an April Fools Joke about it (since removed).  Because you can be carried to gear, you can achieve a very high ilvl without actually being good at the game.  For proof, I direct you to the endless cries about bad pugs in LFD.  Ilvl indicates time investment far more than it indicates any minimum level of performance.

This ties in to the idea that World of Warcraft does not inherently offer any form of feedback mechanism to help someone get better at playing.  Without external tools it is nearly impossible to analyze a DPS rotation and determine the optimal way to play.  If you are, hypothetically, interested in your own performance but not aware of Elitist Jerks, you can spend hours practicing at a training dummy and generate no useful information about your own performance.  All you see is some numbers without context – it’s impossible to evaluate the marginal effects of haste, or crit, or mastery, or rotation changes.

Further, the ilvl system can be gamed.  A lot of us did it to get into 5-man heroics a little bit earlier – we bought cloth rep rewards, or spirit plate, and continued wearing our ilvl277 gear having gotten a high enough “fake” ilvl to get into the launch heroics.  You can do it now, given enough gold – simply buy high ilvl BoEs for all the slots you can, or crafted gear – whether or not you can wear it.  Instant ilvl.  And ilvl doesn’t take into account gems or enchants at all – which is a fairly significant portion of the actual contribution gear makes to a character.

I can circumvent ilvl with gold.  Or I can “earn” it with time investment, which is no indicator of skill or performance.  But I have no in game way to prove that I’m good enough to skip it.  By the time I gear up a character enough to participate in LFR, I won’t just be in a position to contribute, I’ll be in a position to carry.

Gearscore, by whatever name you call it, is a silly gating system.  It leads to some people just jumping through hoops to ignore it, it actively discourages people like me from participating in the systems in gates.  I’m a good player, and an experienced raid leader – I would be an asset to LFD and LFR, if I was encouraged to participate, but as it stands I am discouraged from doing so.

The gear based gating system creates a Blue-endorsed perception that gear is an appropriate proxy for performance, and I would agree that this statement is true… but only for the top 1% of the player base, at most, and you can’t decide who the top 1% is based on their gear level.  In-game performance checks and feedback as gating makes for a much better system, if gating is required.  Ilvl, at best, defines maximum performance potential – and it’s somewhat rare to find people reliably approaching that potential.  It is a very weak proxy for performance – it’s time to let it die.

A Different Sort of Gating System

I propose this: set up solo “scenarios” in 5.0.  If you pass it, then you qualify for LFD.  If you don’t, gear up some more, and try again.  Give us a performance-based metric for qualifying, rather than a grind-based metric.  Anyone can step up and overgear it – with relevant gear only.  It’s solo, so you can’t be carried – you progress at the rate of your skill.

A system like this allows the designers to offer some actual feedback to players.  You’re either succeeding or failing at your task.  If you’re failing, the game can offer you some tips or feedback.  You can even add a few mobs to act like party members – the tech already exists from the Ursoc Quest in Grizzly Hills.

Why would I prefer a system like this?  Obviously it benefits me, I can circumvent some of the grind, because I expect to be able to pass any check relatively early.  It benefits LFD and LFR as a whole, in my opinion, by guaranteeing a performance threshold from the people participating.  By actually creating a system that offers a gate to the content which is relevant to succeeding at the content, it encourages players to consider their own performance – something the game is sorely lacking right now.

I don’t see a way to actually fix the ilvl system as it stands.  I do see the potential for a quest-line based solo-instance type of gating system to offer real and tangible benefits to the game and the community playing the game.  In the end, in any group-based content, it’s the results that matter, and the results are made up of individual performances.  Exceptional individual performance can carry poor individual performance, and that’s fine – but the end result of a system that forces you to prove that you can not stand in some fire and put out some minimum level of output is that everyone’s experience is better.

Final Words

In all fairness, since I started out with a reference to our email thread, I should point out that my co-bloggers don’t strictly agree with me here.  Clearly they’re not as lazy as I am.  I personally see the current system as a disincentive for players like me to participate in LFD or LFR, because I can more easily leverage my guild, skill, and reputation on server to join GDKP runs for the gear, or zone into the instances manually and skip the whole system.  Perhaps I am alone in this, but I feel like I would be an asset to the system, if I was encouraged to be a part of the system, rather than encouraged to skip it completely.  I’m not asking for a free pass – I’m just asking for a system that lets me put my money where my mouth is.

This entry was posted in Mel's Random Musings. Bookmark the permalink.

29 Responses to Where Gating via ilvl goes wrong

  1. 0xfx says:

    I thought the highest quest rewards were ilvl 333, how can you get the 365 needed as a fresh 85 in quest gear?

    • Meloree says:

      You can’t. Hence the request for a performance-based system.

      • Ngita says:

        The highest quest rewards in 4.2 are 365, the highest in 4.3 will be 378. Sorry not doing 90 minutes of questing post 85 before jumping into zx with 308’s in those slots is lazyness, let alone sub 333 weapons that could been upgraded from the cauldron of carnage in Twilight Highlands.

        While i disagree with the basis of why the gating system should be changed your solution does sound better. Be aware that the soio part of the current legendary, according to the forums is both not a challenge as it was far too easy and impossibly hard for their class.

      • Meloree says:

        Postulate: I can gain more of an absolute performance benefit by spending 90 minutes doing research on a class, some UI tweaks for that class/spec, and practice at a target dummy, than I can by harvesting whatever gear I can get in 90 minutes of questing.

        Further, there aren’t anywhere close to a full set of 365s available in 90 minutes. I would be hard pressed to get a fresh level 85 to ilvl 346 (to queue for ZA) without buying a bunch of crafted gear, or without investing a lot of time in launch heroics.

        Either way, complaining about the time investment is the least part of the basis of why the gating system should be changed – there’s a fairly large number of flaws.

      • Ngita says:

        I agree 100% your postulate, but while their is not a full set of 365 indeed only 1-3 items depending on the character, they are indeed higher then 333.

        But if I am inspecting your character, i have already decided something is wrong, I am just trying to ascertain what it is. The 90 minutes spent on training dummies probably never happened.
        I do admit I have not levelled a fresh 85 this expansion ,They all spent time @ 80 in wotlk.

    • Palisis says:

      You cant. Even with the few quests in 4.3, you won’t get anywhere near 378. I have 4 recent 85s (dinged after 4.3) and I still had to do some farming of cata/raz heroics before getting into ds heroics, and then being able to get into lfr.

  2. Lord Kaladar says:

    On the note of your ‘solo’ performance mechanic, who is to say someone wouldn’t simply let another play do it for them? If people are willing to pay for gold, character leveling, etc, why not pay for this as well?

    That being said, I think it would be a very cool mechanic. Could also be a potentially viable way to ‘solo’ content that people won’t group for anymore. And let the more casual (or loner) players see some dungeons without suffering the heathens in LFG.

  3. Adam says:

    Blizzard: “we want all players to be able to experience the content” Good or bad that statement applies. So not going to happen.

    I know exactly what you mean in theory its a grate idea what you have suggested. But still it would be something extra that i have to grind and this is something i do not like, i hate grinding and why should i have to grind something to prove myself when i already know i am good.. Yawn boring.

    If your that good, and you want to raid then join a guild. The Looking for raid tool is for people who have no clue or alts imo. Anything with sense or skill will join a guild and raid.

    The LFR tool clearly is not aimed at people like ourselves. So we should just not use it and stay clear and enjoy normals/heroics. We do have a 2 night a week reset hopefully coming soon so if your intention was to use LFR to pass some time when your locked I guess that will help you a little?

    The normal modes in 4.3 by the way are much more easy than Firelands, Bot, BWD was. So your have no issues doing them on alts or people with lesser gear or a good pug.

    The LFR is actually so easy, I was tanking on my DK and managed to do 3 boss without dipping below 75% hp with 4 DC’s and half the group lagging on the PTR. Its a joke it really is laughable.

    If you ask me one issue with LFG bigger than the one described here would be LAGG! So many people with old machines that just wont cope with 25man mode with projected textures turned on.

    Your going to get people in your group running <10fps and unable to move out of stuff because they cant see it, or because there PC will just freeze. Your also going to get people with damn bad internet and they will DC every 5 mins (this was the biggest issue I had with using the LFR tool, i have used it 5x now for about 2h's each time and it was always the same old thing)

    I would rather take a DPS with no gems, who has no clue on rotation, even a crap talent spec and wrong gearing / reforging than someone who would just lagg as soon as you pull.

  4. Esoth says:

    Great post. I would absolutely love a gating system like this. They could set up a bunch of interesting lore and side-quests to get access to join the legion of heroes (players) who are joining the fight (LFD). Only the best players could get in right away, but that doesn’t matter – just gear up a bit more and/or get a bit better at the game. Then you can get in to this cool system with a sense of accomplishment and the knowledge that your teammates have also passed this test. Could also toss in some achievements for getting to this point. I love it.

    • Meloree says:


      I think that a huge part of the appeal of a system like this, to me at least, is the knowledge that your teammates have also passed the test. That has to be a morale boost to the community that’s participating.

  5. Ohken says:

    Such a gating system sounds really fun.

    A significant downside is that it is more content Blizzard has to create, and it’s content that players won’t even play through but once per alt. It’s not just an extra rule that their programmers can implement.

    I must say, though, that the current system seems quite good in practice, despite the flaws you point out. Partially this is because people who have played a lot have already figured out ways to game the ilvl. Partially, it’s because it doesn’t take THAT much time to increase the ilvl. And partially it’s because, when playing an alt, I actually kinda want to repeat old content rather than jumping immediately into the new. It’s refreshing.

    • Meloree says:

      You only go through Hellfire Peninsula once per alt – if that with LFD levelling. The content would get used. It doesn’t really even require particularly new content. Dress Shattered Halls back up for a tank challenge, or dress up Shadow Labs for a DPS challenge, etc. Especially if it’s training grounds/qualification themed.

  6. Pitsum says:

    I agree with your thinking on the i-level gating system. Player performance and i-level have little correlation to each other. However, I feel there is no need to create a skill test and use it as a gate.

    Why can’t Blizzard know if I am a skilled player or an underperforming player by the time my character reaches level cap?

    Players – not characters – should be given a skill rating. This skill rating could be like the match making ratings for arena players. No need to hide it though. Make it visible so players see it and can decide to work on it – or not.

    As you correctly point out, i-level sets the bar for maximum possible performance. Skilled player are going to be near the maximum possible performance for a given i-level while underperforming players will not. Blizzard can be comparing actual in game performance against what other players of the same class, spec and i-level range are doing. It is a simple matter of checking an individual’s performance against an ever evolving matrix of data. This works for tanks, healers and DPS. I-level 334 Disc priests are compared to 334±2 disc priests and i-level 361 arcane mages to 361±2 arcane mages. Skill level is checked, calculated and adjusted on the fly after the completion of any group encounter.

    Content only becomes available to a player when either your skill level and / or your i-level reaches a point where you carry your own weight. Progression still happens as it currently does but content for underperforming players opens up when they reach higher i-levels while content for skilled players has a much lower i-level requirement. For example, a skilled fresh 85 might get into heroic 5-mans right away while underperforming players would need to grind gear to obtain a much higher i-level.

    Faking out i-level no longer gets a player into content they should not be in as their skill level would too low for the content. Players would be – roughly – carrying their own weight.

  7. Kaarn says:

    I cannot imagine a scenario based test that would actually provide a superior metric for LFD performance than the ilvl based we have now. Most groups that I see fail are not because of lacking DPS/HPS or because the tank can’t generate enough threat. Which are the things you would probably test for in scenarios. Most groups fail because they don’t know about or choose to ignore encounter mechanics (necessary CC/movement/interrupts/dispels) and/or grossly overestimate their abilities in a random group setting.

    I don’t claim that ilvl gating is perfect but it was easy to implement and forces people to put some game time in their characters. No, this will not make them better players but it keeps people playing. I guess whatever gating mechanism you use it will be abused anyway and a lot of bad play isn’t because of lack of potential performance but because people just don’t feel like putting in the effort at the moment (Hey, I’m eating a pizza here/This is just a Zul/Whatever).

    I kind of like your idea of using scenarios as a additional gating system for high skilled players. So they would not have to gain gear up to an arbitrary amount but instead could prove their abilities. However, I think that this niche is too small to put in the necessary development time of designing scenarios when you already have a system in place that achieves more or less the same.

    • Theck says:

      I don’t understand how you can claim you can’t imagine a better metric than ilvl for success in LFD while at the same time pointing out that the two or three major reasons pugs fail is 100% independent of ilvl. So, in your opinion the current ilvl check is the only way you can imagine to test if players will stand in fire or interrupt on cue? That logic doesn’t make an ounce of sense.

      There’s no reason you couldn’t test for interrupts, movement, and not-standing-in-fire in a skill-based test. In fact, I’d argue that those should be mandatory parts of such a system.

      And it’s not even hard to imagine how you’d do that. What’s the most obvious way to test if someone will succeed in a boss encounter? Ask them to succeed in a boss encounter! Give them a tweaked version of Akil’zon, an NPC tank with infinite aggro and an NPC healer that only heals the tank, and ask them to kill the mini-boss. They’d have to switch targets to the eagle to prevent the healer being carried off, they’d have to move to the lightning storm location to avoid dying, and they’d have to interrupt a slow-cast-time chain lightning, all while putting out a respectable amount of DPS to kill the boss within the specified time period.

      If they can do all that, which should be possible for any competent player within a couple tries, then they’re all set for LFD. And if they can’t, then the skill test itself serves as the system the game uses to *teach* them those skills, introducing the instructional component the game currently lacks.

      • Kaarn says:

        I can’t imagine that any test by any metric will actually improve LFD performance. I actually don’t think that many people are as bad as they seem. I see many raid equipped people (even before nerf rounds) perform badly. So it’s not that they don’t know how to perform well, they just don’t want to for whatever reason.
        So you have someone who can successfully interrupt in Nefarian P2 but won’t have a single interrupt in a 5 man. What will a scenario achieve here that ilvl doesn’t? He will pass both gating systems but will still perform badly in a real 5 man because he doesn’t put all his abilities to use.

        Unless a test can somehow enforce that people will put in the same level of effort in a random setting which they use in a guild group I don’t think much would change. That is why I think you can save the effort of developing a scenario. Just stick with ilvl gating because it also just shows a certain level of potential which may or may not be used.

        Maybe I’m just too pessimistic and it’s really that people don’t know how to play properly. In that case an instruction scenario might actually help.

  8. Adam says:

    Skill rating will always be a bad idea, It would make little business sense. Casual players (or bad players who play casual or play allot but are just fail) will be penalised and just get fed-up and stop paying every month.

    Like i said the issue in LFR for most people will be other people’s INTERNET CONNECTION, COMPUTER HARDWARE (LOW FPS/CRASH MID FIGHT), AND “OH SORRY G2G IT’S DINNER TIME!”

    Ilvl/Skill will be the least of your worries trust me, you can seriously just 10man these 25man LFR rights in bad gear if you have a brain. Its so damn easy you could even go as far as using a 1 button macro for everyone in the raid.

    I would even bet money that you could even multi-box 25 people and do one of the new raids by yourself!

  9. Zaey says:

    I am not familiar with the quest you referenced, but it occurs to me that the portion of the legendary staff quest that occurs in the nexus is an excellent example/prototype of a skill gate. It requires a mix of sufficient dps and the ability to use a handful of your classes utility abilities.

  10. Rookey says:

    If warlocks played their character in LFR or LFD the same way the nexus staff quest required us to, we would be performing at about par with a typical retarded pug idiot.

    • Zaey says:

      You’re being purposefully obtuse. If I play any encounter the way a completely different encounter required me to then I will perform poorly. The point is that one demonstrates that they can determine what an encounter calls for and use their class’ bag of tricks to mold their play around those requirements.

      • Rookey says:

        That’s the problem with gates. Its the only threshold that really maters, so a typical underachiever will only put enough effort to get through the gate. When the requirements for the gate are so different than the requirements of what is beyond the gate, poor performance results. The old gating system rewarded those who put minimal effort into acquiring the most gear, and that is the type of player we see in the pug. If a new performance-based gating system is not designed perfectly, there will still be an offset in what type of players get through the gate and what types of players perform well in raids and dungeons, and if not designed well, that offset will be large.

        The SATs for college and the DMV driving tests are a couple real life examples of gating systems that people put specific effort into passing and many severely under-qualified people get to the next stage. Just take a college 101 class or be a pedestrian in china-town, you will know what I mean. At least people can flunk out or get kicked out of college and lose their drivers license.

        A WoW gating system will always only test that a player is able to meet the requirements of the test exactly once. As long as the test is in the form of a gate, LFD and LFR will always be bad, no mater how good the test is. Mel was only half right that the solution is to provide better test content. The scope of the test needs to be much wider than a singular gate event that occurs completely outside of a real raid or dungeon.

      • Zaey says:

        Everything you’ve said is correct. No gating system will ever be perfect, but they don’t need to be, they just need to be a demonstrable improvement over the current system. A drivers license doesn’t make you a good driver, but it does prove you have a basic understanding of how to implement the mechanics of driving within the basic constraints set forward for traffic. It is a far superior “gate” to driving than the ability to afford a car, which is essentially the gating mechanism that WoW currently uses.

  11. Mozen says:

    I’ve always thought each role should have a series of instanced quests that give, say, an achievement that proves you are able to dps well or heal well.

    [Apprentice Soldier] – do X dps in
    [Heroic Soldier] – do 1.3X dps in
    [Twilight Raider] – do 1.5X dps in
    [Heroic Twilight Raider] – do 1.7X dps in
    [Fireland Raider] – do 2X dps in

    [Apprentice Mender] – keep NPCs up for X seconds in
    [Heroic Mender] – for 2X seconds, [Twilight Mender] for 3X seconds etc.

    [Apprentice Tank] – not let any NPC die in
    [Heroic Tank] – not let any NPC take over x% damage in
    [Maintank] – this may test for say tank switches that require watching buffs/debuffs, or using cooldowns at the right time, etc.

    Tank achievements obviously require all previous achievements to activate, and each tier’s new tanking achievements will have the mobs scale up to prove that your “gear+skill” set is really good enough to enter the next tier.

    All three roles will have to move out of the fire etc in all of the quests. Make environmental awareness a basic requirement.

    These achievements will then act as practical indicators for both skill and for gear. Basically like a certification system, and LFD/LFR can then require a specific achievement to queue for the role at each tier. Certain behavior can be flagged for instant failure, like mixing in mail and leather when you’re a plate wearer.

  12. Pingback: Pandaria.nl – World of Warcraft Mists of Pandaria blog » Blog Archive » A proposal for scenarios in Mists of Pandaria

  13. Pingback: Forgotten Children | Blog | A proposal for scenarios in Mists of Pandaria

  14. Pingback: A proposal for scenarios in Mists of Pandaria | Owned Gamers

  15. Pingback: A proposal for scenarios in Mists of Pandaria | Dan Desmond

Leave a Reply