User:Toraen/Ideal Vetting

These changes would require recoding of the vote page and likely the conversion (or in the worst-case scenario, deletion) of existing votes, so it's unlikely that policy changes outlined here would come to pass anytime soon even if supported by the userbase. I just couldn't leave this alone though so here it is.

What current policy does right
The threshold for vetting a build is 3.75 out of 5.00. This means that it's rather easy for low votes to drag down the score. This is largely a good thing because non-meta builds should have widespread approval to get vetted. It should not be easy to slip mediocre builds into Good with the reasoning, "it works". This was the basic line of reasoning that led to the removal of the Other category.

I feel that any new vetting policy for PvX should preserve this feature of vetting. Occasionally, niche builds that should be vetted will fall to it but the alternative is an influx of bad builds.

Problems with the current policy
The number scales and the split between Effectiveness and Universality are not well understood by new users, and there's no clear agreement on what the numbers mean. They would make sense for well-defined absolute scales of the various factors influencing a build's viability, but no such scales exist. A difference of 1 point in effectiveness (0.8 overall) or 1 point in universality (0.2 overall) are not tied to specific amounts of performance. Builds are always vetted relative to what else could be run, and that changes between updates and meta-shifts.

Suggested Changes
In my proposed system, voters can choose Great, Good, Mediocre, Poor, Bad, or Trash for a build, and supply reasoning to support their choice. There is no split between Effectiveness and Universality (which are linked to each other anyway). Only builds that are Good or Great overall are stored (cut-offs are the same as Real Vetting). Builds below Good are deleted after two weeks.

Semi-formal definitions of the ratings are as follows:
 * Great 5 point vote: One of the most effective choices in many of the expected situations it could encounter. Builds in Great tend to be seen commonly as a result of their broad effectiveness.
 * Good 4 point vote: The build has a niche where it is superior but is not as suited to as wide a range of the expected situations as Great builds. They tend to only be seen when their advantage can be put to use.
 * Mediocre 3 point vote: The build is outclassed despite building as best it can for its role(s). Builds in this category tend to seem promising on paper, but fall short in practice.
 * Poor 2 point vote: The build optimizes itself for a niche that has little value and as such isn't useful. Builds in this category tend to demonstrate lack of knowledge about the metagame/area that it's designed for.
 * Bad 1 point vote: The build has severe flaws that prevent it from being viable. Its few advantages, if any, are not significant enough to warrant its use. Builds in this category tend to demonstrate lack of understanding of game mechanics.
 * Trash 0 point vote: The build clearly does not meet the standards of PvXwiki. Builds in this category tend to be troll submissions or not much better than that.

Probably should rename the current "Trash builds" category to something not associated with the troll submission rating category.

Biased Author Reasoning
The point of removing Universality would be to make voting simpler and vote disputes easier to resolve. There have never been standards for taking off certain amounts of points from Effectiveness/Universality for specific issues (which is what voters in the current system do), and attempting to develop such standards would be very time-consuming and inflexible to meta-shifts. In the event that certain build qualities become less/more valued, we'd need to update those standards. Voters would in all likelihood just ignore any policy-set standards even before they became outdated.

In my proposed system, you'd have a list of final scores to choose from. Do you think the build is Great? Good? One of the 'should not be stored on the site' options? Whichever it is, you would directly vote for that category. Vote disputes would be easier on admins this way because each rating has a specific meaning now. They technically do now since the ranges for each category are known, but now we have indicators of just what deserves a 0, 1, 2, or 3. Since trash votes were most of the drama before, this would ease the burden somewhat (who am I kidding though, right?).

I also said that Uni/Eff were linked earlier, and they are very closely so. Most builds are designed with specific objectives in mind for specific game modes. It doesn't make much sense to take points off a build for taking it into areas it wasn't designed for, you just take a different build. Most builds are designed to handle most/all of what the meta/area (PvP/PvE respectively) throws at them, so it's rare for there to be more than a 1-point difference in effectiveness and universality.

To me, Uni feels like a relic from when builds were expected to be tagged for multiple areas and farming/running builds were very generalized. The SC section (which is entirely composed of hyper-specialized team farming builds) didn't even exist! The only category where Uni really would be making a difference is general PvE, since that covers so many different areas each with different threats, but the builds in it are structured with tons of optionals/variants anyway.

Splitting some sections
Random idea I had just now, general probably should have been split into PuG and heroway (some player builds like the Fevered Dreams bars we have make no sense for 7 hero, but were what you had to run in PuGs). GvG might have also benefited from an AT subsection due to certain builds that only show up when guilds predict their opponents (though it probably would have been a tiny section).