Friday, February 24, 2012

Personal Traffic Light DPS Meter

I'd like to highlight a comment from Stubborn on the previous post:
How about borrowing something from a totally different genre of gaming, like my wife's Wii Zumba? When she's performing within 90% of max, her character is highlighted in green. When she's between 60% and 90%, shades of yellow, and below 60%, shades of red. There's absolutely no reason a similar, personal system couldn't be implemented to let each individual player know, based on the minimum dps requirements for the fight, how they're individually performing. That provides the feedback you need without the immediate consequence of creating a hostile environment.
This is a pretty cool idea. A simple green/yellow/red meter that was boss-dependent and displayed your personal DPS for that boss.  If all the DPS stay in yellow or green for the entire fight, the boss will die before the enrage.

Of course, the one thing that might cause issues here is DPS during the execute phase. Most DPS classes see their boss damage jump upwards during the last 20% or so. Accounting for that might be a little tricky. But not insurmountable.

You could make this pretty complicated if you wanted to. If you know the fight, you could draw a damage curve that accounts for phase changes, interruptions, and execute phase. Then just draw the players total damage as time goes on and compare to the curve to determine if the player is ahead of schedule or falling behind.

Still, it's a pretty cool idea, and be enough feedback without needing exact numbers, or revealing performance to everyone. Or maybe the red/yellow/green light should be revealed to everyone. After all, it's not players imposing arbitrary and unrealistic standards on each other. It's the game revealing the basic performance.

17 comments:

  1. Is that supposed to be X% of max character DPS or a sort of minimum DPS needed to down the boss? And how annoying would it be to see red all the time if you are on add duty or kiting or whatever?

    As I mentioned before, it's simply counter-intuitive design to have big flashy numbers pop up when you hit buttons, and then pretend like those numbers shouldn't matter... and then make them matter via enrage timers.

    I forget who said it, but why don't we apply the same philosophy to HP and mana? Green = 90% HP, Yellow = 60-90% HP, Red = <60% HP. No accurate numbers! We surely wouldn't want players to have feedback regarding how much damage they're mitigating, now would we?

    ReplyDelete
  2. Minimum DPS to down the boss. The idea is that if all lights are green, you will not see enrage.

    Obviously it wouldn't work perfectly for every boss. But it would work for a great deal of them.

    Eh, honestly, I wouldn't notice if there were no numbers on health bars. I heal by eyeballing how much health the bar has lost, and how much a heal will refill it. It's already a pretty numberless operation.

    ReplyDelete
  3. Well, as Azuriel wrote, it would fail miserably if the fight is complex and has add kiting/interrupt duty, etc.

    BTW the equivalent is not healing without the number on the health bar, it's healing without SEEING the bars, only a green/yellow/red indicator if you HPS is high enough.

    ...and I can already imagine the screaming on TS after a wipe: ok, so WHO HAD A YELLOW BAR?? YOU TELL ME OR I KICK EVERYONE OUT!! :)

    ReplyDelete
  4. It would be really depressing if you are playing a class/spec that is undertuned to not really be doing much wrong and still see the meter on red because it's tuned to a higher performing class.

    ReplyDelete
  5. @Helistar: lol. "Everyone hit printscreen RIGHT NOW, and post it in the raid's Google Docs. If it takes longer than 30 seconds, I'm going to assume it was Photoshopped."

    ReplyDelete
  6. The problem with the system is that it's not intuitive but is designed to fake it; I consider this worse than a system that is obviously non-intuitive because that one at least makes the users to either read a guide or give up rather than use it in a wrong way.

    It's possible to do a test, ask some people to read the description within a time limit (say 30 s or 1 minute), ask them "X is green, is she able to carry her own weight in the encounter? yes/no/maybe" and see how many people answer yes (which is a wrong answer).

    Also, Recount and Skada get a lot of criticism for not taking all details into consideration even when they do - the traffic light does not.

    ReplyDelete
  7. Better yet, hook up a heart rate monitor to the game, then people who get it to aerobic levels get a 20% boost to dps/healing. Soon all the best guilds will only recruit players who can raid and treadmill at the same time :)

    ReplyDelete
  8. Already exists in the form of WoL. You can compare yourself to other damage dealers there (from same class/spec). If you are over 100% you might be ranked. One should not forget the performance outcome is, in WoW 10m raids, heavily influenced by not only fight mechanics but also your group composition. For example, if you don't have all raid buffs, you will get less performance % compared to the 100% performance because those people have full raid buffs. You should also not forget gear; this is not taken into account in the percentage. In 25m raid (de)buffs are much less of a concern.

    One of my guild members had to be soaker on Morchok HC. The reason for this was that we had nobody else who could soak, and he went frost. He was one of the very few frost mages, and his DPS was high, so he got ranked. Of course we had people ranked before, but it was just a funny mention after the kill. In contrast, if you play a more popular class/spec combination such as affliction warlock on Madness HC then you will have a more difficult time getting ranked.

    Blizzard stated before they considered adding a damage meter in WoW but that it has its downside. One downside was that it'd count all the damage which isn't always fair because not every damage is equal (among the recent ones blobs on Yor Sahj, Tendons on Madness, Rhyolith legs). Thing is, if they'd officially add it people would take it more serious. So that is why, by default, WoW does not have a damage meter. However pretty much one of the first addons I installed back in the days was a meter. After having asked the noob question what that exactly is.

    ReplyDelete
  9. There is an unstated assumption in the original post. It assumes that a full out combat log and traditional DPS meters are bad for the game, as they prove to be a destructive influence on group play.

    Now, a lot of people will disagree with this. I probably do as well. However a lot of people do believe this, and they may be right.

    But just suppose it is true. What can you do to provide DPS players with feedback, without going to the full extreme of normal damage meters?

    ReplyDelete
  10. Feedback is, hands down, the most important element of improving performance at any task. Without it there is no way to measure and evaluate output. The binary (boss dies/you die) and non-individual feedback available within the game isn't anywhere near good enough, and leads to frustration when the fights are too hard and boredom when they're too easy. It also makes tuning nearly impossible since player damage will cover such a broad spectrum.

    In the majority of cases, protecting people from information exposes them to greater harms than providing that information, and dps meters show the same result; sure some people may be treated poorly by some other people because of that information, but the alternative is either frustrating or dull gameplay, no opportunity to improve, and finally - loss of subs.

    Hopefully Bioware will catch on and provide the feedback players need to challenge themselves and improve. The game, the players, and the devs will all benefit if they do.

    ReplyDelete
  11. This idea is about as stupid as the idea that one can judge the quality of a DD by it's dps. It's like back in WotLK where people asked for "your dps" before inviting you to a heroic group.

    If the color is a percentage of the dps required to down a boss it doesn't account for
    - different possible dps of different classes and specs in this very fight.
    - different required dps because of different raid layout (more healer, tank that do more dps while tanking or off-tanking)

    Besides that it, again, encourages all the tricks to "cheat" Recount like
    - not target switching
    - dotting unimportant targets
    - using AoE spells where not useful
    - not moving our of fire
    - not using non-dps abilities like decurse or spellsteal

    And it devaluates buff-bitches. Who wants to play a buff bitch if the game constantly remindes you that you suck?

    A single dps number does not indicate if a DD is doing his job or not. It doesn't matter if this number is shown in Recount or through a color code.

    ReplyDelete
  12. Note that I'm going to state several times in my comment what I think opinions of others are. I might get it wrong and I apologize if I insult anyone by doing so.

    @Rohan
    There is an unstated assumption in the original post.
    Yes. Yes, there is. However I think this is an opinion held by nobody (not even Stubborn - read his post in its entirety) and that's the exact group of people who would be satisfied by the performance indicator you describe.

    Considering only people who have an opinion, I noticed three groups:

    1) People who want as much information as possible and then some and if anything, they would like the damage meters look more like in-game WoL, just more complex. There's less of them than the other two groups so even if they're unhappy with the change, if the others would be more happy, I guess it's a worthy trade.

    2) People who think anything that would allow judging players is bad as it might drive a wedge between friends when some of them are lacking of performance while others aren't. I am not sure whether the proposed change would make them happy because it still is a performance meter.

    3) People who think the performance meters are good but the UI which controls how the information is presented is lacking. The problem with current meters is they're designed mostly by the people from the first group for themselves. I don't think there's place here to get into the reason for this, I believe it's a problem that a lot of the open source software suffers from and more experienced people wrote papers on that.

    While I agree the information would be presented in a better way, the information presented is lacking - one of the reasons people want to change the meters is because they perceive meters as not providing a complex analysis of performance. The current ones do so I would say it's the UI that makes the information hard to discover. However, your suggestion has the meters not to present the information at all.

    So the question is who would be content with replacement of complex performance meters with one that is not (as opposed to either keeping it or removing it altogether)?

    ReplyDelete
  13. Already exists in the form of WoL.

    That's an external resource. Probably a good 70% of WoW plays with the default UI, and they spend their offline time on facebook rather than thinking about WoW. These people have no idea how well they're doing or how well they're supposed to be doing. It's high time Blizzard put a way to measure your damage done and dps in the UI.

    Not only that, but they need to add guidelines on how to play a spec. They obviously have some design in mind, so they should tell the player in a method akin to the dungeon journal thing. A newbie fire mage probably has no real idea what Combustion does or how to use it.

    ReplyDelete
  14. Not only that, but they need to add guidelines on how to play a spec. They obviously have some design in mind, so they should tell the player in a method akin to the dungeon journal thing.

    While people are complaining about MoP's talent and spec system, this is actually what they're doing with it. They took all of the talents and abilities they expected you as the player of that class to use, and just give it to you as you level, and only those abilities you should be using. That all by itself is a massive improvement in instruction as to how you should play your class.

    ReplyDelete
  15. Very interesting discussion. Reminds me of the raging debate in public education: is it appropriate to reward or even fire teachers based on their students' test scores?

    Certainly feedback is good, but quantitative feedback like test scores or percent of theoretical max DPS to evaluate a complex activity can also be less than useful.

    And worse, it rewards ignoring the myriad intangibles, and promotes "teaching to the test" as others have commentd.

    Maybe that function would be useful on a target dummy, but then a PuG would demand a screenie that you can "go green".

    ReplyDelete
  16. To all,
    So I missed this whole thing because I don't read blogs over the weekend, and I guess this showed up after my "offline" time began, so apologies for a very late comment.

    I agree with the commentator who said feedback was the most important way to improve your performance in a game. However, I am in no way interested in a game that requires people to improve their performance. That's not to say people shouldn't be able to, but that the requirement to do so bothers me.

    In Imak's breakdown of responses, I fall squarely into #3. I vehemently dislike the elite culture that's risen up around damage meters. I also agree that feedback's important and that a basic "stoplight" might not provide enough. There's no reason, though, that a stoplight type system can't be made to incorporate many other things. Classes that can interrupt can get a color change for not doing so (and when they do, at their point in a rotation, their color bounces back). Tanks could get damage mitigation colors based on when they should have used cooldowns (absorbing Ultraxion's Shadow Bomb thing much?). Dps working on kiting adds already look like crap on Recount; why would it bother them that they were orange for the whole fight? They know they're doing the right thing. The color codes could take the off-jobs into account and assume a particular number of dps and heals, then simply extrapolate along those lines.

    As complex as recount is (which does measure things like interrupts), there's no reason it couldn't be implemented in a visual form.

    As for being depressing to see red all the time, how is being at the bottom of the meter every boss fight any different? At least this way it would be personal and not something others might shout at you about.

    I do think that providing just numbers is dangerous. NYC just released teacher test scores to the public, and I'm just waiting to see the fallout. The Dept. of Ed. repeatedly warned people not to judge the teachers based on the numbers because some are in very tough schools, some have weird sample sizes, and some have a variance of 35% (I can't italicize, but that should be emphasized)! The top teacher's entire sample is 1 class for 1 year of 32 students. That's it. His numbers are up against teachers who teach seven classes of 35 kids for multiple years. There's no comparison.

    Human beings judge. It's how we survived so long. We make quick judgments. Some of us, afterward, take a break and reflect to see if our judgments were right. Others don't. Recount doesn't really inspire reflection on whether that PuG you just shouted down was new, just spilled soda in his lap but kept playing so as not to cause a problem (but had a massive dps drop anyway), or anything else. It reduces human beings to a single variable: Dps output. I don't like it.

    I believe there are measures of feedback that can work. I think that the option to keep your information personal should be available. High-end raid guilds could simply require their applicants to turn the option off.

    Sincerely,
    Stubborn

    ReplyDelete
  17. @Stubborn: It reduces human beings to a single variable:
    Interesting. When writing about group 3, I meant people who do not like Recount because it fails to reduce human beings into a single variable (or a few variables) properly. Because it does not and because so many people go with substandard performance indicators which are much simpler (a couple of numbers and true/false values) than Recount's output.

    ReplyDelete