PvXwiki
Register
Advertisement

criteria

Three criteria suggested in this policy is just an example. It is possible to make 3-4-5 or 10. It is very adjustable.

I really don't want to see us promote a build just because it is innovative. There are plenty of good builds out there that are pretty cut-and-paste; that doesn't make them bad. For example, a good war build will have an IAS, speed buff, ~3 attacks and a rez. There's nothing innovative about this, but a build featuring those skills would be effective. There's no need for that build's score to suffer.
The potential of build is also a poor quality to judge a build by. Some people can design a good build, designated to fulfill a certain purpose (let's say, e-denial). It may perform excellently at e-denial, but may not be effective at something else (say, spiking). Under these criteria, the build would rate much lower than it deserves. Just because it only does what it was designed to do doesn't mean a build's score should suffer either.
The categorization system of this policy looks great, however. It would certainly be beneficial to any policy we end up implementing.
As far as the voting side, it looks good too, though I think it would be inappropriate to call this an easy voting process. It is complicated, especially when compared to some of the other policies.
Most of this polciy looks really good, but it has a glaring Achilles heel, if you will: criteria. Having a static set of criteria to evaluate a build upon will be harmful to the wiki. It will misrepresent builds that are really good, as well as those that aren't (like the 'best use of a bad elite' builds. They may be good to have, for players who don't have any useful elites yet, but they should never be ranked above those common, uninnovative builds that really work). Evaluation is better left in the abstract; rigid criteria will not be beneficial to maintaining a useful Builds section. - Kowal Krowman (talkcontribs) 00:52, 15 May 2007 (CEST)
I think this policy has a lot of potential, but needs some cleanup.
To me, it depends on how the vetting result would be displayed. If the three criteris (Idea, potential, and strength) are averaged together for a signle score, then I agree that it would artifically help the score of poor builds, while dropping the score of effective but non-innovative builds.
On the other hand, if the scores are not averaged, and each build is displayed based on each score individually, then this seems a very effective method. Scripts can easilly reduce the complexity, so no issues from that. If you're looking for a powerfule build, search for all builds with a high "strength" score, ignoring the other two scores. If you want innovative, search for high "Idea" scores while ignoring the other two. This gives it a lot of flexibility. I would also suggest writing the scripts to be easilly expandable - if a fourth or fifth criteria are wanted in the future, write the scripts from the beginning to support that possiblity.
The main thing that I would suggest changing is to make the vetted results based on 5 categories of 20% (0-20, 20-40, 40-60, 60-80, and 80-100), instead of the currently proposed 25% segments.The preceding unsigned comment was added by 161.88.255.140 (talk • contribs) .
People come to the wiki looking for help with builds, i.e. they are looking for good builds to either mimic or be inspired by. They aren't looking for builds that other people were very creative in making. The strength category is the only one that will actually be useful to someone who needs help with their builds. The other categories are geared more towards pleasing build authors than those who want to use the wiki as a resource. - Kowal Krowman (talkcontribs) 02:44, 15 May 2007 (CEST)
However, if all we saw were the same old builds then the game qould start to get rather dull. Especially since EotN has only 150 new skills, and 50 or 100 (I forget which) are PvE only, someone has to be creative if the metagame is going to change at all and if old players are to stay interested. Another possiblity would be to organize the builds into sub-categories, having a category for most creative, best overall, best potential, best overall strength, best damage(spike and pressure) best defence, best healing, best E-denial and whatever other categories you decided to include.--Foozdood 21:23, 17 May 2007 (CEST)


Why not just bring over the Overall Category sugested by my proposal an place it into this one. May give a 4th definitive build 'rating' that is still abstract enough for some, and detailed for others. May cynch this into place for those who want simplicity over detail. Shireen 18:10, 18 May 2007 (CEST)

Could work just as well to rename it to Design Performance, rather than 'overall'. Shireen 00:44, 19 May 2007 (CEST)

Suggestion

While having a discussion with BrianG, we figured out a way to kind of get around the Achilles Heel that Krowman pointed out. What if we base the actual percent rating (that determines the placement of the build in the hierarchy) based soly on the Strength category (since that category pretty much covers the things that make a build good). Then, we could give people the option of also assigning Innovation, Idea, etc. scores to a build. Those scores would not effect the actual rank, but, would instead be usable as search criteria. That way, we get all the benefits of the policy, while still maintaining a high standard for our builds. Thoughts? DE Sig Test 2 *Defiant Elements* +talk 05:04, 15 May 2007 (CEST)

Good compromise. - Kowal Krowman (talkcontribs) 05:05, 15 May 2007 (CEST)
I agree. However you would need to make the rating more than 0 to 5 to get a real ranking based on just one category. At least to 20.

--Dark Grendl 18:28, 16 May 2007 (CEST)

@DE and Krowman; I thought that was what I said above, but maybe I wasn't clear enough.
@Dark Grendl; Not really. Letting each individual user vote 0 to 5 simplifies voting, and averages of multiple votes goes into decimal places, so there's really an infinite range of average rankings with 4.5, 3.25, 3.14159, etc.

Yes, or something like efficiency + utility. ~Corpselooter 17:14 +1Gmt. 18 May.

Dispelling Some Myths

Ben says:
While having a discussion with BrianG, we figured out a way to kind of get around the Achilles Heel that Krowman pointed out. What if we base the actual percent rating (that determines the placement of the build in the hierarchy) based soly on the Strength category (since that category pretty much covers the things that make a build good). Then, we could give people the option of also assigning
Ben says:
Innovation, Idea, etc. scores to a build. Those scores would not effect the actual rank, but, would instead be usable as search criteria. That way, we get all the benefits of the policy, while still maintaining a high standard for our builds. Thoughts?
Ben says:
So that way, Strength is the only thing that really matters for determining score, but, people lookinfg for innovative builds can still search for them.
gcardinal says:
Yeah sure sounds a like a great idea. How about faking its effect? I am not sure if people will rate build "good" if they know it has nothing to say , how about we make Strengt count for 70% of tottal score, and other 2 for 15% and 15%. that way they cound but not to much to damage anything
Ben says:
Meh... I don't know.
Ben says:
The problem is that the VERY BEST warrior builds would still recieve lower ratings because most of :them aren't creative.
gcardinal says:
hmm yeah
gcardinal says:
I just afraid we will not have any builds in Good, only Great
Ben says:
I still think it would work fine like that.
Ben says:
And voting on Creativity and that stuff would just be optional.
gcardinal says:
sec let me check
gcardinal says:
Okey how about this, does a VERY BEST warrior build has 0 potential ?
Build W/Mo
Strengt : 100% (counts 70%)
Potential : 100% (count 15%)
Innovation : 0% (counts 15%)
I Think this is the best the only problem is caltulating and evrything
And still we wil get 85% tottal score and it will make into Great section
gcardinal says:
Its important that both Great and Good will have builds, if there will be only 10 builds in good, it would be good
Ben says:
What is potential defined as?
gcardinal says:
If this build has potential to do more than it has been designed for.
gcardinal says:
LIke you can farm in 1000 areas with build = potential
Ben says:
No... builds do what they are designed to do... that is just a silly category...
gcardinal says:
No not really
Ben says:
A GvG warrior is good at one thing... being a GvG Warrior.
gcardinal says:
okey let me illustrate that
gcardinal says:
http://www.pvxbuilds.com/index.php?title=Build:W/A_Shove_Hammer_Spiker
Ben says:
Alright...
gcardinal says:
That is a PvP build, but it can be used for AB, CM, GvG, RA, TA -> 100% potential
gcardinal says:
Potential in form of Universal
Ben says:
Universality doesn't mean good though.
Ben says:
Again, a GvG warrior may be the best GvG warrior but not be good at anything else.
Ben says:
Why should that lower its score?
gcardinal says:
It will not lower its score, it will make it Good
And if its a build that is so great for GvG so specificly designed, it may just as well get Innovation score, becouse its a new way of thinking or what ever.
gcardinal says:
http://www.pvxbuilds.com/index.php?title=Build:Mo/E_SoA_Sliver
Build that is
Strengt - 75% becouse its has many major problems like hard to use
Innovation - 100% new way of farming
Potential - 100% can farm 40 different boss's
that will make it a great build
Ben says:
Alright... whatever... just make Strength count for the vast majority at least.
gcardinal says:
And another thing, if we see anything going wrong, it will take just 1 second to change balance in terms of how rating is counted, and it will automaticly recalculate evrething
Ben says:
There is another problem though.
Ben says:
Just in general with your policy.
gcardinal says:
yeah?
Ben says:
Alright... I know you may not agree, but the fact that this gives no way for an Author or an Admin to refute a vote is ridiculous.
Ben says:
Some votes are just stupid.
gcardinal says:
I dont remmber saying that
gcardinal says:
of couse we can refute a vote, if we cant do that what the points of having admins ?
Ben says:
Well also... is there any way that we can force people to present a reason for how they rate a build?
gcardinal says:
we may also need to create a group of people with that kind of permission. not sure about giving that to author
Ben says:
And then can you have Admins remove votes that just make no sense based on the reasoning.
gcardinal says:
Yes there will be rating with box's and shit, but it will be also box where they MUST type some reason for they vote
gcardinal says:
and people can edit they vote and reason any time
gcardinal says:
yeah exactly
gcardinal says:
that was my idea, it was maybe stupid not to mentining it in a policy
Ben says:
So, as an Admin, I can see someone's reasoning and then render the vote null and void if the reason just makes no sense?
gcardinal says:
in short, when you go to build page, next to [Discussion] will be [Vote] when you go there, you will be askt to rate a build using 3-4 box's rating each parametr from 0 to 5 bla bla. Under that you will see current builds score, votes from other people:
User name
Score given to build
Reason given
Each user will see that but admins and (maybe) a bigger group of trusted users will have right to
gcardinal says:
delete a vote giving they reason for deletion of the vote
gcardinal says:
however discussion will not find place on that page, and will be in [Discussion], but in top of the discussion will be current builds rating just as on diagram on the page
Can some of Admins please make a summary of this discussion and edit policy according ? GCardinal 00:37, 19 May 2007 (CEST)

Winner

...This one won... Can GCardinal tell us how many votes the others got? NVM--User:Frvwfr2 frvwfr2 (talk)(contributions) 20:48, 18 May 2007 (CEST)

ick--Coloneh 23:01, 18 May 2007 (CEST)

My main thing is that this is the one all the socks voted for... But whatever. We can flesh this one out. -- Armond WarbladeArmond sig image{{sysop}} 23:15, 18 May 2007 (CEST)

Page is buried again

- Need to put a link on the front page, a LOT of people are going to want to be able to have a quick refrence to it without having to guess and dig where the link is. Voting results are not linked to their policies. Shireen 00:38, 19 May 2007 (CEST)

Ummm... click on the Official Vetting Policy link under the "Writing Your Own Build" Section. DE Sig Test 2 *Defiant Elements* +talk 00:56, 19 May 2007 (CEST)

Transformation

Here is my suggestion on what to do with builds we have no and how to move them under new policy:

  • All current favored builds goes to Working\Good.
  • All current un-favored builds goes to Archive\Trash
  • All current stubs goes to Archive\Store
  • All new builds as by policy will start as Drafts.

It can be a small mess to start with but as soon as that extension is ready it will be okey.GCardinal 00:35, 19 May 2007 (CEST)

Do we need to retemplate anything? Eronth 00:36, 19 May 2007 (CEST)
Pretty much everething. But I have to write an extensions first. Then we will see how to deal with what we have. I will figure out something. Until then just hold GCardinal 00:37, 19 May 2007 (CEST)
Give us a holler and we'll (Me for certain) help out. Gotta break the funk man, quickety split Shireen 00:42, 19 May 2007 (CEST)
Putting all currently favored builds into the working or whatever section is going to solve problems from the previous policy how? We know that old builds in favored needed to be cleaned out, and still do. --Sefre File:Sefresig.pngT*C 00:59, 19 May 2007 (CEST)
Agree with Sefre. Move them ALL to untested or w/e it's called. --User:Frvwfr2 frvwfr2 (talk)(contributions) 02:44, 19 May 2007 (CEST)
Voting will be done on all builds as time goes. And as voting goes builds will be automatically moved around. Moving all builds to Trash will not solve a problem. Taking it step by step will first keep site up and running and will keep it usefull for Readers as well as Writers will have a playground to start with GCardinal 12:32, 19 May 2007 (CEST)
We didn't say move them to trash, we are saying move them to Drafts I think. --User:Frvwfr2 frvwfr2 (talk)(contributions) 15:04, 19 May 2007 (CEST)
Move all builds on the site to drafts. I dont fully agree with this policy, but i think we all know the old one was crap.--Coloneh 21:36, 19 May 2007 (CEST)
FINALLY. Will work on it. -- Armond WarbladeArmond sig image{{sysop}} 06:34, 20 May 2007 (CEST)

Some thoughts...

...that occurred to me when re-reading the policy article:

  • Instead of "percentage of current/max ratings", wouldn't it be better to call it "average rating"? Technically it's the same, but easier to explain/understand.
  • The categories (Great/Good/Store/Trash) are meant to be implemented as wiki categories along with Drafts and Untested, right? (The article doesn't say anything yet.) At the moment there's also an inconsistency between the diagram and the description concerning the category names.
  • Assignment of a category is based on an overall rating (I assume). How is this calculated? I favor the 70/15/15 weighted average proposed above.
  • Should the "Strength" criteria be split into something like "Power" and "Usability"? Not so experienced players might look for a build that they can handle, and accept that it has limited power, while others don't care how difficult a build is to use (or to get the equipment) but want maximum power for a difficult task.
  • What about additional, optional criteria, that don't go into the overall rating but help in describing the build? Some of the ideas of pvxwiki:True Build Ratings could be taken over. Might look complicated, but shouldn't scare anybody as long as it's optional.
  • Search engine: I think this is a very powerful and important tool that can render many discussions obsolete. It should include filtering and sorting with respect to all the "criteria". A close to unlimited amount of additional gimmicks is thinkable and can be implemented later. GCardinal: I can send you the draft I wrote (don't want to publish it here, it might cause some load if too many people use it in the current state).
  • Finally, at some point a trusted native speaker should polish the article text (no, I'm not :-)

I guess that most of these points are being worked on anyway, now that we can concentrate on one policy. Btw: congratulations for winning the vote. --Hhhippo 12:04, 19 May 2007 (CEST)

It says the script will be made within a week. im curious as to who is writing it?--Coloneh 08:26, 20 May 2007 (CEST)

Make a new topic under which ppl can suggest criteria, general as well as optional? --Luuck 11:18, 20 May 2007 (CEST)

MAJOR PROBLEM


How the hell do we stop trolls from 0 rating everything? Coz we gonna get idiots who do that. This policy is pathetic in the way that is has such an obvious flaw. And what does this mean? We have people who don't TEST BEFORE VOTING. Back to GuildWiki again, the past is here to stay. '~\^/~' Napalm Flame Napalm Flame Sig Image (talk)(contributions) 11:26, 20 May 2007 (CEST)

People who rate 0 on all criteria are easily identified as trolls and can be warned/banned. That's already a major advantage of Real Vetting over just Favored/Unfavored.
But you're right, there should be a note in the policy saying that you need to give a reason for your rating. If a rating seems unjustified, the rater should be asked to clarify it. If that doesn't help, admins should have the power to strike the vote.
In general: Let's try and see how it works. There will be more to the new system than just vetting (like you can check which builds your favorite xx-expert is favoring). So the actual rating might become less important than in the old system, reducing also the motivation to push the rating up or down.--Hhhippo 11:57, 20 May 2007 (CEST)
That's why I didn't vote for this policy, and why I decided to go for the scale of 1-10 one, I think it was the one about things like overall, practicality, specialist roles etc. But yeah, I think reasons MUST be given. But that's not enough to satisfy me.
I believe that we must all provide PROOF that we have tested the build, either by screenshots, a report on how the build functions and where you have tested it, or especially for the people that are willing to provide the software (by request sending the full version by e-mail or MSN), FRAPSing your manic mayhem, which would be by FAR better proof as then you can see whether the person's vote is valid or not as you can see their playing style or if they were using the build in a completely stupid way. Obviously if a build is so stupid it violates PvX:WELL then it will just be deleted on the spot. Not so easy to do all of the above, but it is damnnear foolproof. '~\^/~' Napalm Flame Napalm Flame Sig Image (talk)(contributions) 15:14, 20 May 2007 (CEST)
It's also close to making nobody want to vote for any builds. The amount of work it would take just to put your vote on a build is tremendous. Sure you could implement that type of policy, but PvX would lose much of the user base it has now, which still isn't that much. No one would want to vote, and the to verify that everyone did what they were supposed to to vote would also take a tremendous amount of work. So instead of taking a good idea on how to handle the votes, you want a almost perfect plan, but at the cost of many users not voting and ten times the amount of work. I think we should just leave the voting regulations the way they are. Most bad voters will be caught. The only thing that I would say, is that the user had to be logged in. That might already be up there, but I don't know. My 2 cents. Bluemilkman 15:58, 20 May 2007 (CEST)
Then this policy would fail badly. With no regulations and no PROOF, if someone goes and rank 0s about 20 builds and just says it sucks, that would then be considered a perfectly valid vote. THAT is what I feel will fuck this wiki over like GuildWiki was. Heck, this is no better than favoured/unfavoured really. Well, it is in the slightest bit. '~\^/~' Napalm Flame Napalm Flame Sig Image (talk)(contributions) 16:03, 20 May 2007 (CEST)
If they 0 rank a bunch of builds, they'll be checked against WELL. If they don't violate WELL, the troll gets a warning for VANDALISM. Because there's no worse vandalism than tanking build ratings on a BUILD WIKI. Someone add that because I have to go. -- Armond WarbladeArmond sig image{{sysop}} 17:55, 20 May 2007 (CEST)
Did you not read what Hhhippo said about it about three comments up. People who vote 0 for no reason can have their votes stricken. The ones who just go around voting 0 will be identified as such, and could possibly get banned. As long as someone has a good reason for voting 0, then it's fine. The only thing that I could see you being not so happy about is the fact that someone has to tell an Admin that somebody is just voting 0 for no reason. That will leave it up to the normal users to make this policy as good as it can be. Bluemilkman 17:58, 20 May 2007 (CEST)

In reply to the note-box by Defiant Elements: Great. That should end the discussion. Maybe include it in the policy description soon. There are some points that should be clarified anyway, see previous section. Oh, yes: And maybe call it "as per discussion in the community". If only two people decide about details of a policy which was just voted on, that calls for complaints. Even though I don't think anybody will oppose this particular decision. --Hhhippo 19:36, 20 May 2007 (CEST)

Yeh, it probably wasn't phrased in the best possible way, but, I was really referring to the conversation posted in a previous section that already answered all of these questions which obviously no one bothered to look at. But still, I hope the notice box renders this whole discussion moot and we can get on to more important things. DE Sig Test 2 *Defiant Elements* +talk 19:39, 20 May 2007 (CEST)

Something that I would like to comment on is the 'testing required' remark above. A testing requirement is a practical impossibility. It is unnecessary to test a build in order to determine it's quality. Giving reasons behind your vote should suffice. If a testing clause were included, all that we could use as evidence would be screenies. Uploading (at least!) 5 screenshots per build as evidence would be a nightmare. As well, what do you take screenshots of? How much a build heals for? How much damage it does? There are qualities in all builds that cannot be measured simply by numbers. Having to test build will create a huge backlog of builds in Untested. Leaving a reason behind your vote will tell people what is wrong with the build, how it can improve, and should leave build authors feeling a little more satisfied that someone did not simply vote against their build out of spite or whatnot. - Kowal Krowman {{sysop}} 19:55, 20 May 2007 (CEST)

YES! NAPALM HAS SOME INFlUENCE! NAPALM HAS DONE GOOD DEED! WOW! Okay I'll stop capsing and get down to business. Nice one defiant, thanks, much appreciated. But do you plan on giving build creators the right to strike out certain invalid votes like trollers and votes that do not make sense/cannot be applied/are downright stupid/are just incorrect, or do you have to notify an admin? Coz once this policy goes into full effect there physically won't be enough of you at some point to keep monitoring the votes for the above problems. And this doesn't just go for 0 votes, can go for 5s or any number. I can see there will be some very slight teething problems in time to come, but nothing too serious I hope ;) (And by the way, I'm only on about when this wiki reaches it's point of boom... Sounds like the English housing market... typical) '~\^/~' Napalm Flame Napalm Flame Sig Image (talk)(contributions) 20:57, 20 May 2007 (CEST)

Right now, it looks like only Admins will be able to strike votes since voting will occur within an extension and it is unclear whether or not regular users will be able to view the results (which will include the reasons). I would never advocate giving Authors the right to strike votes since I just wouldn't trust a lot of the random Anons who post builds or whatever. Maybe we could get a larger group of trusted users who weren't Admins and give them the power to strike votes. Regardless, we are trying to expand the Admin pool since we do acknowledge that we need since it is just too big a job for 6 people to handle. However, to really answer your question, I would guess (you have to ask Cardinal for the real answer) that all users will be able to see who voted and their reasons for voting, but, that only Admins will be able to strike votes, and, it will be primarily the responsibility of normal users to tag votes for Administrative Review. DE Sig Test 2 *Defiant Elements* +talk 21:25, 20 May 2007 (CEST)
I agree with Defiant Elements, you really don't want the general population to have the right to strike votes. When users did give reasons on GuildWiki and the vote was negative, many anons and even some registered authors automatically said that because it failed for that user, the user obviously did something wrong, and therefore the reason was invalid or seriously flawed. This would even happen when the votes were 5:1 or more towards unfavored - many noobish authors simply refused to acknowledge the community's reasoning. As far as the author was concerned, in their mind the build was the greatest creation in Guild Wars history, and anyone who voted negatively obviously didn't do something right or was flatly incompetent. To keep a handle on the volume of issues, it's a good idea to have names ready to expand the number of users who can strike votes in case the volume of complaints get to high, and users who can strike votes wouldn't really need to be full admins. But there should also be a policy in place to settle disputes (arbitration or whatever) for when two users who can strike votes disagree on if reasoning is valid or not. --- Barek (talkcontribs) - 01:32, 21 May 2007 (CEST)

Protected

Why is this protected still? If it will be the official policy then people need to be able to fine tune points and fix errors on it. Especially points mentioned in the above section.--Sefre File:Sefresig.pngT*C 21:35, 20 May 2007 (CEST)

probably because there is no need for it not to be protected... concsencus has to be reached on the talk page before a change is made anyways. Not a fifty five 21:55, 20 May 2007 (CEST)
True, if someone has a real reason to edit it, we can always unprotect for a short time. But, beyond what 55 said, there is also the fact that if a page is viewed as "unpopular," there is always a greater risk of vandalism, much like with the Builds wipe page on GuildWiki which had a bunch of occurrences of vandalism. Once things settle down and it is properly implemented, we can un-protect the page, until then, we can leave as is. DE Sig Test 2 *Defiant Elements* +talk 22:03, 20 May 2007 (CEST)

Quality vetting policy?

While I think the idea is pretty cool... I don't think this policy solves ANY of the problems guildwiki had... the two major ones being

  • Flood of terrible builds making untested monstrous
  • Poor quality voters who do not read the entire (particularly the usage section) article or think about it

In fact, this probably only aggravates the second problem, as people are probably less likely to comment before rating in this system as the new "vote" has no comment area Not a fifty five 22:16, 20 May 2007 (CEST)

Err, we got rid of the untested problem at Gwiki, it just took time and effort. The big problems at Gwiki were general quality and NPA violations and the ensuing headaches. With this system, they have to leave a comment about why a build is bad, if they choose to vote that way. That kind of mutes your 2nd point; if they don't leave any reasoning (to make it a valid vote), their vote will be stricken. The 1st problem, like I said, only really occured at the beginning of new chapter releases, and we got it under wraps over at GWiki. - Kowal Krowman {{sysop}} 22:22, 20 May 2007 (CEST)
Well for #1 if I recall correctly, guildwiki DID solve the nightfall builds inundation for a brief time in untested but it eventually flooded back to the hundreds. And for #2, I guess thaat's true. We'd prolly have to describe what an obviously flawed reasoning would be in more detail, though. Not a fifty five 22:30, 20 May 2007 (CEST)
I think Krowman gave you the reason why Poor Quality Voters is not an issue. I would add that I would guess that it will be left largely up to the Admin who strikes the vote as to what constitutes a poor build. As to floods of bad builds... well, we aren't going to limit people from submitting builds, but, PvX:WELL does give us broad latitude to delete bad builds which means that less time has to be spent actually vetting them. And, with a little effort, we can keep them from being flooded. DE Sig Test 2 *Defiant Elements* +talk 22:37, 20 May 2007 (CEST)
Advertisement