Wednesday, October 14, 2015

DPS Meters and Player Behavior

Whether DPS meters are good or bad is a hotly debated topic. I saw an interesting Reddit post by WitcherMog that approaches the question from a different angle.

In FFXIV, there are no in-game damage meters. However, you can run an external parser that reads your combat log and gives information for your group. FFXIV runs on a  "don't ask, don't tell" policy when it comes to meters, though. If you reference them in chat, it's considered harassment, and you get punished by SE. So meters are pretty much the province of individuals and high-end static groups.

WitcherMog ran 100 instances with his parser active and compared player behavior with parser results. He found that harassment of other group members for being slow or wiping uniformly came from low-dps players.

The negative behavior indulged in by high-powered players (as measured by gear level) was to slack off during fights. Of course, negative behavior in either case only occurred in a minority of runs.

This is not quite the same negative behavior exhibited in games with public meters. In WoW, any harassment is more likely to come from good dps towards the low dps.

The results do make sense. In a game without meters, a good dps makes the run faster and less likely to wipe. So really, she has nothing to complain about, as all visible outcomes are positive. Meanwhile the low dps player believes he is doing fine, believes that he is really a good player, so obviously any problems must come from the other players.

In a game with meters, on the other hand, the poor player cannot blame others for visibly bad outcomes, because each individual performance is quantified. However, even if the outcome is visibly good, a negative high-dps player can feel aggrieved that she is "carrying" the group, and incite harassment of the others that she feels are not playing up to par.

So it's really a choice of picking your poison. Without meters, runs that have negative outcomes like excessive slowness or wipes incur harassment from the very people who most likely cause the problem in the first place. With meters, a run that is successful might very well still see harassment, just because one player thinks another player is not living up to an arbitrary standard.

There's no obvious best choice here. Would you rather have poor players never realize that they're bad, or have good players hold others to an unnecessary standard?

Of course, we must reiterate that, especially in FFXIV, we are talking about a minority of outcomes. Most runs are successful and go just fine.

22 comments:

  1. It's the Dunning Kruger effect in action.

    The incompetent don't have the skills to recognize that they are incompetent. That's the first key finding of the study. The second finding, though was that when the incompetent were shown their results (shattering the illusion) and then given a little bit of training, they were able to tell that they were terrible.

    When there are no meters, the terrible people don't know that they are terrible. They don't know what good DPS looks like without being ABLE to do good DPS.

    ReplyDelete
    Replies
    1. That's exactly right. You need feedback to be able to evaluate your performance. The lack of meters makes it harder to improve as a DPS player.

      Delete
  2. Ironically, I'd just read that thread on Reddit, closed the tab abd started in on my Feedly, with your post here being 2nd from the top. Quite the coincidence.

    Gotta agree on the DK effect Phelps mentioned -- the incompetent simply don't know they're bad. I've been fortunate in my Duty Finder groups and have really only had a couple of bad ones here and there. One that stands out in my mind was the Vault run where the ninja didn't dodge avoidable damage, forcing the healer to spend more time healing him and less time healing the tank and/or dps'ing, so we'd get to the point that the healer'd be out of mana and we'd wipe. The ninja kept blaming the healer, but the healer and the tank blamed him while I pointed out that he wasn't avoiding damage like he should. so he rage-quit. Instantly replaced with another dps and we 1-shot that boss and the healer had more than half his mana remaining this time and it never felt like there was any problem at all. I expressed my surprise that a dps player could cause so much trouble, and the tank said something like "Yeah, FFXIV actually makes the dps responsible too."

    ReplyDelete
    Replies
    1. Yes, a bad dps often has a more subtle effect on the group. A bad tank or healer, in contrast, is very obvious.

      Delete
  3. I like the way SWTOR does it, which also lets you parse externally but only for yourself.

    I can honestly say that in this case I would choose ignorance over constant judgement any day. I'm less annoyed by wiping because people aren't up to snuff than by succeeding while someone berates people about numbers the entire time.

    In the grand scheme of things it's really only a tiny portion of content that requires you to optimise your output, and you can easily choose to do that with people who are all OK with parsing and having their performance measured.

    ReplyDelete
    Replies
    1. True, but only parsing for yourself requires a lot more coordination when you're actually working as a group. Still, it's a good solution, as you have to opt into the group parsing.

      Delete
  4. "Would you rather have poor players never realize that they're bad, or have good players hold others to an unnecessary standard?"

    That is not really the choice you are given in FFXIV.
    98% of my runs in this MMO are smooth and there's also no blame going around (exactly because there are no meters). Since FFXIV doesn't have the hardest dungeons, it really doesn't matter who was better in the group or not, as long as it's smooth and company is pleasant.

    The choice you are presenting would only exist within those remaining 2% of "bad" runs then, where bad players blame others. Do I prefer that over the constant bickering in WoW from so-called good players? - Hell yes. Every day of the week.

    While it's an interesting experiment from that player on reddit, the big difference is that it's a theoretical approach. Nobody in FFXIV pugs cares about these parses and most people don't use them. But in WoW they do.

    ReplyDelete
    Replies
    1. I think that depends on two things: what content you run, and how good you yourself are.

      If you only run the easy content, then you'll never see the issue. But in more challenging content, the content where it's important that people stay calm and not be jerks, you'll see more futile bickering. In challenging content with parsers, if there's a DPS issue, you can fix the problem. Without parsers, there's just arguing and the group breaks up.

      Second, it does depend how good a player you are. You, Syl, are a good player. Thus you can carry groups to a degree. No-parser groups only have issues when they fail. So you don't see a lot of negative behavior.

      In contrast, imagine if you were a worse player. Your groups would fail more often, and you would see more negative behavior. In contrast, with parsers you would be able to measure your performance and improve even in successful groups.

      I think there's value in that second scenario. No parsers make a better environment for good players, but does it make a worse environment for poor players who want to be good?

      Delete
  5. For me the rule is simple: if it matters, I must be able to get feedback about it. In World of Tanks I see when my bullet hits because I get a message and see the damage inflicted, while I get nothing if it misses.
    In an MMO, if aggro is important then there must be some feedback indicating if I have aggro or I'm far from having aggro or I'm about to get aggro. Even if it's not ultra-detailed and just uses "red/yellow/green", it's enough.
    The same way, if DPS matters, then I need to have a way to get feedback on my DPS, maybe just a number at the ennd of the fight. Same for received damage, a tank needs to have feedback on damage mitigation, a healer needs to know about overhealing, and if avoiding damage matters, you need feedback on how much you avoided.
    Of course this does not need to be broadcast and can be simply "personal" information appearing on my interface.
    Things change when in a group: if the success AS A GROUP depend on DPS, then I need to know the DPS of the group, and if the composition of the group can be changed, I need to know how to change it.
    If you don't want people to care about performance, then design a game where performance doesn't matter, it's as simple as that. Using an "official" built-in damage meter can give the best of both worlds: you get a damage meter only in those places where DPS matter, say a "hard mode", in "casual mode" you get no damage meter, but the fights are also tuned so that auto-attack (i.e. not being AFK) is enough.

    ReplyDelete
    Replies
    1. I agree with this. FFXIV does the first part with built-in threat meters, which are hugely helpful. But I think we haven't really gotten to the point where this is seen as necessary for damage.

      Delete
  6. Of course there is an obvious good choice: built-in meters and automatic removal of low DPS players. So there is no one to blame, as he is kicked.

    ReplyDelete
    Replies
    1. True enough. The hard part though is defining low DPS in such a way that you don't remove people unnecessarily.

      Delete
  7. "In a game with meters, on the other hand, the poor player cannot blame others for visibly bad outcomes"

    I have noticed a tendancy for low-DPS players (and low-HPS players) to blame wipes on players failing at raid mechanics: such as not targetting the right adds, or not being in the right place at the right time.

    I wonder if this is a defence mechanism, aimed at deflecting the pointy finger of blame from them. The subtext is "My DPS was bad because I was continually retargetting and moving to where I should be. Hers was good because she was standing in the wrong place DPSing a fixed target rather than moving and targetting the correct add."

    And sometimes they are right.


    ReplyDelete
    Replies
    1. Interesting point. Shifting concerns to an unmeasured area. It's a hard thing to say, "I did poorly on that fight." We often make justifications instead.

      Delete
  8. I'm really only familiar with Wow dungeons.
    It seems to me that the players posting DPS meters are rarely someone who was a good part of the team. They're the Draggy Tanks, the Warriors and Death Knights who are dps but are pulling everything.
    On the other hand, for me dungeons are always a sort of proving ground for how well I'm playing a class, and the classes I traditionally have struggled with--shaman and rogue get cringe worthy dps, which leads me to try to figure out how to play them more effectively.

    ReplyDelete
    Replies
    1. I think that, to a degree, this is because of the changes WoW has made to the basic combat skeleton. You don't really see this behavior in older-style trinity games like FFXIV or SWTOR.

      Delete
  9. Based on outcomes, given that there is no sum of money you can pay me to tank for a WoW group, and I barely tolerate instances versus FFXIV where I tanked every instance I ran and enjoyed every one of them, I'd say DPS meters are a bad idea.

    ReplyDelete
    Replies
    1. Hah, that's true. Like Syl pointed out, the proof of the pudding is in the eating. No meter runs are definitely more enjoyable.

      The only question I have is if no-meter games are enjoyable for everyone, including the less skilled players.

      Delete
  10. I wonder how much of the bad players complaining about other players in FFXIV is a variant of "I believe they doth protest too much". Basically, how many of them know they're bad and are just deflecting/redirecting to keep the heat off themselves?

    It's funny, because in my experience most of the people I've encountered complaining about others in WoW are good at DPS, but mediocre players. The number of times where I've been "called out" on a boss or trash in LFR, and then I crush them on meters as either healer/dps the next boss has been pretty numerous.

    I think part of the problem is in WoW, a 25-player raid only needs 4% of the population to be jackasses to be likely to have at least 1 jerk player in a run. I'd bet most people have far more productive/decent 5-player dungeon runs than LFR runs. This is exacerbated by WoW's push towards LFR at the expense of dungeons, versus FFXIV where 4-player dungeons are a large part of the gearing endgame.

    Compare that to 4-player dungeons or 8-player raids in FFXIV, given an equal percentage of jerks you're less likely to find one in your group at random. I distinctly remember the Crystal Tower (24-player) raids having occasional jackasses as well.

    Anyways, that was a tangent. The experiment is pretty interesting, and does exhibit qualities of the Dunning-Kruger effect as mentioned above. Pretty fascinating.

    ReplyDelete
    Replies
    1. Whoops, missed a sentence. Good DPS as in they pull decent numbers, but mediocre players because they stand in everything, miss mechanic cues, etc.

      Delete
    2. I don't know, Talarian. In your formulation, the bad negative players would have to know that they are playing badly, and choosing to continue to play badly. That seems unlikely to me.

      To be honest, sometimes I feel in charity with them. They must have much lower levels of success than we do. If 95% of my runs are smooth and clean, I can shrug off the odd wipe here or there. If your success percentage is 50%, I can see people getting frustrated that "they're always getting bad groups."

      Delete
    3. I dunno, there are plenty of folks who I know play badly in WoW and are seemingly fine with it (they call those of us who play well "elitists" because after trialing them we said they weren't good enough for our raid), but on the other hand, those same people aren't likely to be calling other people out on their bad play, so I suppose that's a good point.

      Delete