ATTENTION: If you have an account that was recently deleted and are located in the European Economic Area (EEA), please see this announcement.

PvXwiki talk:Real Vetting

From PvXwiki
Jump to: navigation, search

TBH

Universality doesn't seem to be clear enough, because it's misunderstood by a large portion of the voters here. Most people seem to think that universality should be marked down if the build is for a specific profession (7 Hero Paraway) or because it can't function well outside of where it was designed (Derv Plains Farmer). From what I can tell, it's not explicitly stated (and almost appears to contradict) that universality is meant to be judged only within a build's specific scope. --Jai. - 17:19, May 22 2011 (UTC)

Raging seems like a really good way to be the top on the Active User's List ;D. AsuraSignature.jpgAnvil Godzzz... 17:57, 22 May 2011 (UTC)
"Most people seem to think that universality should be marked down if the build is for a specific profession (7 Hero Paraway) or because it can't function well outside of where it was designed (Derv Plains Farmer)"
First part you're correct. People shouldn't mark down because it's for a specific profession (that would fall under effectiveness as it's how effective the given profession is at it's job). The second part they're correct on. Look on the policy page and it even says as much. Obviously you don't put too much weighting on that aspect, but it does weigh in. ~ PheNaxKian talk 20:59, 22 May 2011 (UTC)
That's basically like saying that if an SC build meant for FoW doesn't work in UW, it should be voted down. If a farming build is meant for a specific area, and it doesn't work anywhere else, then its universality should not be affected. Of course, in general PvE it should, but farming and SC's are a different story. What I'm saying is that I think the policy needs to be more clear as to what exactly should affect universality, and honestly on effectiveness and innovation as well (though effectiveness is more self-explanatory and innovation doesn't even matter). --Jai. - 21:20, May 22 2011 (UTC)
then you start unnecessarily complicating things. It means you have to have one definition for one area, another for a second place, a third for some where else etc. That's why the descriptions are what they are. They may not be perfect but it's a balance of being able to understand what they mean and the impact on the vote. Besides it means one, maybe two at a push, points difference in universality. ~ PheNaxKian talk 22:52, 22 May 2011 (UTC)
People are voting zeroes in uni due to not understanding it, which is somewhat a problem. Honestly it just looks to me like you'd need to remove the location clauses from the universality description to make it be more sensible.
As an example of my stance on this, consider Build:E/Me Imperial Sanctum Runner. It finishes the mission faster than most full teams would and only fails if Mhenlo decides to be incredibly stupid (apparently has a ballpark estimate of 1/20 chance). Uni would be docked for the small chance that Mhenlo can irrecoverably screw you over and that you have limited control over his position, but not for the build being unable to work in a different mission or area, since the latter is irrelevant to the scope the build. -- Toraen talk 23:37, 22 May 2011 (UTC)
^ Basically what I was trying to say, thanks. --Jai. - 23:49, May 22 2011 (UTC)
Idk how Toraean does it. Whenever he comes in hes like "(insert text here) and then everyone's like "Oh that makes sense now!". AsuraSignature.jpgAnvil Godzzz... 00:33, 23 May 2011 (UTC)
I think the "Apply common sense" section covers this. The purpose of "Using in a different location than intended" is just descriptive of how flexible a build is, it tends to apply more to pvp builds than pve anyways. Giving a 0 because the build doesn't work well in other areas is stupidity on the part of the user not misinterpretation of the policy--Relyk 01:25, 23 May 2011 (UTC)
I brought this up because there's been a lot of people who have misunderstood (or maybe just not read) the policy recently. And then, for instance, when Shadow gets told why his votes are wrong, he won't believe anybody until a mod steps in, because the policy isn't clear on how universality is treated. --Jai. - 01:42, May 23 2011 (UTC)
It's honestly common sense, it's not like we have a strange way of rating builds.--Relyk 01:54, 23 May 2011 (UTC)
I would agree, but sadly there are a lot of people who apparently don't have common sense... --Jai. - 04:44, May 23 2011 (UTC)

(Reset indent) The problem I have with removing the location clause is then how do you say a build that's made specifically for a given area, is better/worse than one that's a bit more generic, works just as good and covers more areas? You'd have no real way of distinguishing the 2 apart. They're both just as effective as each other (we'll say. not an impossible assumption), so what reason would i have to take the specific one over the generic one if we didn't take into account the build adaptability to a slight change in location? ~ PheNaxKian talk 03:06, 23 May 2011 (UTC)

If there's a general build that's more effective than or equally as effective as the more specific build, then that's a reason to rate it down. But if the generic build is not as effective in that specific area, then IMO the specialized build shouldn't be penalized.
But yeah, all this would get pretty complicated, I guess... --Jai. - 04:44, May 23 2011 (UTC)

Small suggestion

If someone requests elaboration for a vote with insufficient reasoning, they would be obliged to give it. Threaten with bans, gg. This, of course, when it's obvious the other guy got the message. The point of this isn't to see a single person get banned, but to delegate much of the conflict resolve to the users, rather than the AN. Fucking stupid to see people trollvote eachother in an ele build, then spam my watchlist with the Noticeboard. Just threaten to ban shit if they can't resolve shit themselves! Invalid reasoning eventually comes down to admins, but different thing yo. --DANDY ^_^ -- 19:59, 23 January 2012 (UTC)

Skakid needs to avoid abusing the noticeboard. Everyone needs to stop attacking HerpDerp, it's getting old. Creating a dramafest out of a simple FA build is ridiculous. Threatening bans aren't the solution, only less people acting dumb.--Relyk 21:25, 23 January 2012 (UTC)
Welcome to PvX Relyk, you must be new here. RąʂKɭɘş14/f/japan 21:38, 23 January 2012 (UTC)
The point isn't to threaten bans, but to enforce people to expand their reasoning when asked, through policy. The ban part is just the only real tool we have to punish people for breaking policy. --DANDY ^_^ -- 11:06, 24 January 2012 (UTC)
I thought we did this anyway. The "punishment" for a vote with insufficient reasoning is vote removal (which it something we should push more but admins cba) and the punishment for copious amounts of votes which are obviously stupid is a ban - AthrunFeya Athrun snow sig.gif 11:34, 24 January 2012 (UTC)
^"But admins cba". Say, Rask here made an amazing PvE warrior build using Warrior's Endurance to spam Dervish Melee (not scythe) attacks with a Hammer spec, and I give it a 2-2 vote saying "no real reason to run this instead of a regular hammer build". Rask sees that whereas my vote brings up a point, it doesn't really argue enough, so he asks me to elaborate on the reasoning. I say "fuck you, I won't", and instead of taking it to the AN, I get a ban if I blatantly ignore the request to elaborate my reasoning. Voila, increased user conflict resolve! --DANDY ^_^ -- 12:16, 24 January 2012 (UTC)
There is what I would term voting bias, in that most people don't feel inclined to elaborate on their vote if they don't feel it's necessary. They tend to write more more for a positive rating and less for a negative rating. This isn't a problem as long as consensus is that the build is actually good or bad. The policy already implies this, but people should elaborate on or revise their votes if they see that there isn't clear consensus. It's not fire-and-forget or static like you're reviewing a product you just tried. The AN in a nutshell: "Don't use this page to bitch about votes you don't like. Admins aren't here to solve all your problems for you."--Relyk 15:03, 24 January 2012 (UTC)

Hmm

"A build that works but is clearly inferior to another build should get a lower rating relative to that build. However, this does not serve as sufficient reasoning for a vote and should reflect the build's effectiveness individually as well." Does that mean its forbidden for me to compare (pvp builds where I can reroll in less than a minute) the Glyph of Energy elementalist to a Psychic Instability mesmer, which also has KDs, and aoe damage? Chieftain Alex Sig.jpg Chieftain Alex 20:24, 24 January 2012 (UTC)

If it was forbidden, we would have our categories full of Mo/R's with bows --DANDY ^_^ -- 20:32, 24 January 2012 (UTC)
I'm not sure if you're referring to my change in wording or the sentence itself. You should only compare it to other builds to support your argument, not use it as a substitute for any argument at all. Explain why it's inferior to another build; elaborate on what parts of the build make it inferior in comparison; don't leave the argument as a statement that you believe it's inferior to another build and therefore it deserves a lower rating.--Relyk 22:42, 24 January 2012 (UTC)
I believe his point was can he say a Mesmer is worse than an Elementalist because it does less AoE damage (that's the generalised point he was getting at).
Short answer: It depends.
Long answer: If (in PvP) they're meant for similar roles (i.e. both are AoE Offense, or Healers, or Tanks etc.) then yes, generally you can compare and vote accordingly (though generally the "inferior" one should be WELL'ed as opposed to trashed). In PvE we generally say you can, but you can't knock off/give too much on your vote due to this, because of the time taken to re-roll. ~ PheNaxKian talk 23:50, 24 January 2012 (UTC)

Vetting trash builds

Is it OK to skip the voting and trash certain builds? Or must they all go through a trial phase? 62.220.135.129 06:26, 26 March 2012 (UTC)

Yeah, that's what WELL'ing is for. --Jai. - 06:29, March 26 2012 (UTC)

Not that it would cure anything in particular..

But I'm wondering why not use a 1-10 scale or 0-5 in steps of 0.5 for effectiveness/uni? Cɥıǝɟʇɐıu Alǝx 10:46, 16 July 2012 (UTC)

This suggestion would have had some weight behind it if it were 2007. Frostels 10:49, 16 July 2012 (UTC)
We don't have server access, so we can't change jack. A new misery 12:31, 16 July 2012 (UTC)
Wasn't it on a 1-10 scale back on guildwiki?--TahiriVeila 13:08, 16 July 2012 (UTC)
As far as I'm aware it's always been 0-5. As for the 1-10 idea look somewhere in the archives, this has been suggested before. ~ PheNaxKian talk 16:32, 16 July 2012 (UTC)
Closest would probably be PvXwiki talk:Real Vetting/Archive 4#A suggestion--Relyk 20:17, 16 July 2012 (UTC)
Yeah that was the start of it. Iirc that conversation (as a whole, not necessarily the out of 10 part) went over to the next archive as well, so it might be worth checking archive 5 as well. ~ PheNaxKian talk 21:23, 16 July 2012 (UTC)

k going back to archive 5 then, Universality → Flexibility sounded good. Cɥıǝɟʇɐıu Alǝx 10:29, 17 July 2012 (UTC)

...It'd be nice to have actual reasons for these changes. Currently I feel like it's "Changes for the sake of change", which I have no intention on acting on. ~ PheNaxKian talk 16:32, 17 July 2012 (UTC)
Osnap? We can actually do that? I imagine we couldn't do something nice like double all current votes as well to scale them to the 10 point scale? The reason to do it would be to improve granularity as at the moment 5 is great, 4 is good and 3 is trash. A new misery 19:07, 17 July 2012 (UTC)
hrmm, we probably could. I can't say for definite because I've no real idea of what the database architecture for the PvX extension is like, so I can't exactly set it up for Curse to just hit "go", and it might be there are restrictions in place at a database level to prevent values that aren't whole numbers ranging from 1 to 5.
Plus I'd rather solve some of the more fundamental issues with the system first (I'll have chance to give it a good looking at in a few weeks time) before any changes to the system, unless they're massively urgent. ~ PheNaxKian talk 20:20, 17 July 2012 (UTC)
Oh, ok. Probably not worth it then. A new misery 20:24, 17 July 2012 (UTC)

Special:Recent Votes

I just rated for the first time and found out that Recent Changes does not list Votes. How can the community check "voters" activity and also how do you double-check voters do not change values while everybody else is distracted with something else (malicious behavior)? User Yoshida Keiji Signature.jpgYoshida Keiji talk 10:09, 1 October 2012 (UTC)

It's there now. How long after voting did you check? Probably latency/refresh issue. It says "edited by x on x", btw. Doubles 10:40, 1 October 2012 (UTC)
Votes do not appear in Recent changes, but under the Recent ratings link. (You must be logged in to examine the actual vote, but IPs may examine the summary on the Recent ratings page.) Cɥıǝɟʇɐıu Alǝx 11:55, 1 October 2012 (UTC)

Meta in non-GvG/HA areas

We're overdue for an actual definition of what the tag means for those areas and how we actually decide on them, so make suggestions here. "A build that is very commonly run and expected to be known by players of that area" seems really simple though I'm not sure if it's the best definition. Toraen (talk) 07:51, 7 May 2018 (UTC)

The only issue I see with the definition is the possibility for builds to be meta, and then we go ahead and trash it because it happens to be suboptimal. I'd say for builds to get the meta tag, they'd not only have to be commonly run and fairly widely known, but also have a certain standard of effectiveness and optimisation. Otherwise, we'll be listing what could be called 'meta-trash'. NapalmFlame (talk) 08:48, 7 May 2018 (UTC)
If it's meta-trash, it would still actually get deleted (after 2 weeks grace period) with current policy, although this is not explicitly stated. The meta tag doesn't actually accept Trash as a rating, so the build would be simply tagged with {{Trash-Build}}, overriding a Meta tag if it was previously on the build somehow. Having a certain standard of effectiveness and optimization is what the rating system filters for. Toraen (talk) 09:16, 7 May 2018 (UTC)
Even so, it may still be worth stating with regards to defining the meta tag. Other than that, sounds pretty good to me. NapalmFlame (talk) 09:31, 7 May 2018 (UTC)
Toraen's suggestions seems fine to me. Meta is a category of popularity, not quality. I doubt that meta builds would be trashed as even of the worst meta builds there are variants which would qualify as good, and we'd use them for the build pages instead of the even way way worse crap builds which are variants of the meta builds. (Thinking of those UA builds I've seen in the last months.) I'm just not certain whether there are good reasons to have the extra popularity tag. I'm not that convinced right now. --Krschkr (talk) 20:32, 13 May 2018 (UTC)
Without the "bypasses vetting" quality, the meta tag for areas other than GvG/HA is purely informational, like the innovation score in ratings. Unless maintaining that information ends up being a huge hassle, I'd prefer keeping it. Toraen (talk) 02:08, 14 May 2018 (UTC)
I added some RA only builds to meta because I had forgotten it's only for GvG/HA. I remedied that, but what about these builds? Build:D/any PvP Avatar of Grenth, Build:E/any PvP Lightning Surge, Build:E/any PvP Stoning, Build:Me/any RA Energy Surge, Build:Mo/W RA Healing Burst Monk, Build:Mo/W RA Word of Healing? They aren't for HA/GvG either. Do we keep them meta tagged anyway for the useless extra information as they've already passed vetting (except for lightning surge which would only qualify for a provisional rating)? --Krschkr (talk) 17:48, 12 November 2018 (UTC)
As per current policy, if they're really popular builds in their tagged areas, they can retain their meta tag if they have 5 votes. Otherwise bump them back to provisional. I do not personally visit RA nowadays so I can't speak much to their popularity. -Toraen (talk) 19:42, 12 November 2018 (UTC)

(Reset indent) A year later my opinion on the meta tag in PvE has changed: It is a troublesome element that should be abandoned for most builds.

  • Our take on what the meta is (popularity, not quality) does not match how the majority of players understands this term. Time and again I see people using it to describe the best builds, which is the build quality.
    • This means: The current usage of the meta tag in PvX is deceptive for a large share of its casual (by this I mean non-contributing) users.
    • We determine build quality through vetting, resulting in a great or good rating, based on tests by community members. Tagging a build as meta invokes the incorrect impression of outstanding quality, which devalues the meaning of vetting.
  • The meta tag de facto reduces build diversity.
    • Among the most frequented pages in PvX are the meta category pages.
    • These cover only a fraction of the stored builds. Some of them even are most questionable, such as Archive:Team - 3 Hero Spiritway, which was in meta until few months ago and is now archived because it was horribly outdated and out of use for years or Build:Mo/Me Unyielding Aura Hero which is far away from a great rating.
    • PvX is still the place where most people go to find builds. What's presented in here determines what people try ingame, especially less experienced players which are afraid of already making their own builds. If they are presented with a small array of apparently outstanding builds they're much less likely to test other builds, which they might find more enjoyable or builds which prove more effective.
    • Thus, due to the meta tag and meta category, PvX promotes a couple of builds of which some are, by our means of determining build quality, far from optimal. But it does not only promote non-optimal builds, it also disencourages people from playing good or even better builds just because they aren't already popular. In my opinion we should not only document builds, but offer the service to show people more builds that are good and possibly enjoyable. What we recommend should be solely determined by vetting results.
  • If we keep meta as a means to indicate popularity, we're running into trouble. As the spiritway example showed the pages featured in the meta category are sometimes questionable. For years, no one seemed to administer this category – and nowadays the ingame activity is so low that we can't even easily (or at all) determine what's most popular. Most things we put into the meta category or which we remove from meta is, by a large share, based on intuitive decisions – to avoid saying arbitrary decisions.

I vote for removing the meta tag in PvE, with the possible exception of speed clears, where it's meant to bypass vetting. But we're in need of active members of the speed clear community to keep an eye on what's meta and what not, to keep our database up-to-date. --Krschkr (talk) 17:02, 2 June 2019 (UTC)

Barring anyone coming out of the woodwork to administer general PvE meta, it may indeed be best to phase out the meta tag from non-competitive areas. Its perception as a higher quality than great isn't helped by using a different template color and the way the categories are laid out on the main page, so I'd like to change those issues too. To start I'd merge the separate Good/Great/Meta tags into a Vetted tag that has a rating parameter and a separate meta parameter. Meta would no longer give a solid purple tag, but maybe a flair of some sort on the standard green tag. Hopefully I can also do something with the main page layout.
Getting SC to not be outdated will be difficult because the only real way to fix it is more editors who know SCs, regardless of whether it uses meta tags or vetting. -Toraen (talk) 00:57, 3 June 2019 (UTC)

GvG Meta

GvG happens only once a month competitively – of which about 5 matches show up on observe. This low activity makes handling the meta tag troublesome.

  • We can take the approach people seemed to mostly agree on for PvE: "A build that is very commonly run and expected to be known by players of that [PvP mode]." This would warrant a meta tag for each and every GvG team and single GvG player build we currently have in the build namespace and some more gimmick builds like this. People have high expectations (and the frequency is the next point).
  • Going by popularity doesn't seem viable. If 5 matches end up on observe that's up to 10 different team builds; even if a build is seen just once in a mAT it would already constitute 10% of the builds of which we know. I consider that a high frequency for PvP. Applying more rigorous standards would lead to a single unfluxed meta team build, nec bala.
  • Someone who's not me could ask the QQ discord which teams they still consider good for competitive purposes (meta), which are falling out of favour (archiving candidates) and which are considered acceptable gimmicks (build pages without meta tag). This would shift the meta tag away from the actual statistic meta to a meta of attribution, a pool of builds people consider playing when there is a match.
  • Don't decide on anything and I'll keep distributing tags following partly observation (see here) and partly my intuition, which is sort of the pool of builds to chose from solution without asking top GvG players.

Opinions? --Krschkr (talk) 21:49, 19 September 2018 (UTC)

2 Changes to Provisional Vetting

I suggest two changes.
1) If a build has received a 5/5 rating 4 times (or 3x 5/5 and a 5/4) it should receive a great tag without waiting for a fifth rating. After 4 perfect ratings it's highly unlikely that the fifth will be much worse and drag it below the 4.75. Even a 4/3 rating couldn't achieve that against 3x 5/5, and just barely against 3x 5/5 + 1x 5/4.
2) If a build has a rating below 3.0 after 3 votes it can receive a trash rating instead of going back to testing. It doesn't seem to get accepted by the voters and shouldn't linger in the build space for a few more years until it has 5 votes.
Opinions? --Krschkr (talk) 16:14, 28 October 2018 (UTC)

Tbh 5 votes should be lowered to 4 to begin with. Or maybe even 3. Hmm.. what if the regular threshold was lowered to 3, but if a build became controversion on the dicsussion page then it'd be slapped with a provisional until the build either received 5 votes or an agreement was reached? --DefinitelyNotHanz (talk) 19:29, 28 October 2018 (UTC)
@Krschkr: I was actually thinking of reverting the thresholds to 4.5 for great and 3.5 for good. The current thresholds were intended for the active BM era and are very narrow. Fair on builds below good after a month though, they probably should just be sent to trash.
@Hanz: Well, we tried the 3 vote threshold before this and it's not a great solution. The point of provisional is that we can say, "this is probably the correct category, but we don't have enough people weighing in to be certain" and the users will be aware of that footnote to our recommendation. The problem with only putting builds that get controversy on their talk pages in provisional... is that talk pages aren't active enough. -Toraen (talk) 08:17, 29 October 2018 (UTC)
Also, I am thinking of including Speed Clears in the sections where a Meta build can bypass vetting. I don't see any other way to keep that section relevant. -Toraen (talk) 08:19, 29 October 2018 (UTC)
Okay, makes sense ^^ Regarding lowering the rating range to 4.5, how about 4.6 instead? It's still more generous than before, but a good and a great rating alone probably shouldn't put a build into the highest possible category, even if it's provisional. --DefinitelyNotHanz (talk) 09:08, 29 October 2018 (UTC)
Good catch. I'll make sure that we use 4.6 instead if we change it. Changing these thresholds will require Curse to make a small change to the ratings pages so they display properly, so I can't just do it myself now. Also, while I'm asking them to change it, I'm thinking we lower the wait on new accounts being able to vote. 4 days is just too long. -Toraen (talk) 09:19, 29 October 2018 (UTC)
I've started to like the 4.75 and 3.75 marks. The only issue I know with the lower end is that it made the only warrior hero build drop to a trash rating for a few days. Apart from that I reckon that a rating below 3.75 tells quite clearly that a build shouldn't be part of the build namespace, and instead get archived, trashed or moved to the author's user subpages (depending on how serious the entry was and whether it went straight to a trash rating or started as good). The high end on the other hand is well-suited to ensure that builds only get tagged as great if most people agree that it deserves it. If that doesn't happen there doesn't seem to be consensus that the build is among the best builds currently known. I don't think there is an issue with great builds struggling to reach the >4.75. If we just look at player, hero and team builds for PvE we already have 100 'great' entries next to 120 'good' ones. But maybe I'm looking at this from the wrong angle. Why would you prefer slightly more forgiving thresholds? By the way, I wouldn't mind if the requirements for voting on builds would be lowered decisively. I don't expect a wave of fake ratings to happen just because of that. --Krschkr (talk) 03:29, 5 November 2018 (UTC)
If you're fine with the 4.75 and 3.75 thresholds, then I have no real problem leaving them there. I was just concerned that the high thresholds discourage people from voting honestly on builds, which I've seen in many older votes where a bunch of negative points go into the reasoning but the numbers chosen for eff and uni are higher than I would expect from that - though it may just be those users just didn't elaborate enough on the good points they saw with the build, assuming them to be obvious enough. I can't exactly ask any of them about it now.
As for the requirements to vote, do you have any preference for how long users should wait (if at all) or changes to the edit count requirement? -Toraen (talk) 06:32, 5 November 2018 (UTC)
I don't have a preference. As I said, I don't expect that lowering the requirement a lot would cause fake ratings en masse. Two edits without waiting time might already suffice. --Krschkr (talk) 17:27, 5 November 2018 (UTC)

(Reset indent) 3) Another suggestion. I think we should drop the 1 month period before builds can get a provisional rating. Less than 10% of builds get 4 or 5 ratings within that time (if any) and waiting just causes confusion and may cause builds to be overlooked. --Krschkr (talk) 01:43, 15 July 2019 (UTC)