On that note, improving feedback online is nearly impossible. Even big companies cannot get a good reviewing system going, a small forum is therefor even less likely to improve it. You are facing several behaviour issues that you cannot enforce to change:
- Either people always review afterward, in which case it ends up mostly being negative reviews as positive ones are less likely to review, for this forum this would be bad.
- Or in this case, people review before testing based on the video that was scripted. They like the video, so they want more of such content. A basic thumbs up does this job very well, and is to me currently still the best reward for a good script.
- Having multiple upvote options isnt going to be doing what it should do. In almost all cases i noticed it ended up causing a few side effects that often are undesired
- People want a negative option to select then, as it can highlight something bad (which you should never give in to). This feedback will always happen, and is the main reason to never even make more options possible.
- More options will cause a certain option to be deemed a flagging option. A ‘funny’ option can be used as downvote for regular posts, while only on shitposting it might have the proper meaning.
- No matter how much i would like a ‘liked video’ and ‘liked script’ option. I think its just not going to work well. because people cant realy judge good scripts from average ones from just trying it once. They can notice bad ones easily, but the small quality portions are hard to identify. Most people arent experts, they just enjoy content. These votes are most often just going to be aligned equaly and dont say anything. Especialy since people will still blindly click it anyway.
- Again, people arent experts, so dont make the expectation that they have to be one. Letting them say ‘i like this content’ by an upvote is more than enough. If the content is realy bad, comments will make this clear quickly. And yes, average content is fine enough for 90% of users.
- Numeric ratings are most of the time useless, its mostly 10 and 1 ratings you are going to get, because it makes people feel like their vote has the biggest influence. This is why sites like imdb already use a weighted system. And the only reason they can do that is becaue of a huge sample size. The 0, 1,2,3,4 and 10 votes on that are almost always deemed as binary votes (yes, a 4 is deemed binary, as movies rarely are actualy this bad). And users build a reputation on how accurately they rate things. This is far too complex to implement. Stars or other systems dont change anything.
- even steam which has just a yes/no system, has such a weighting system in place, that just the plain number of votes is generaly already an indication of good, even if 10% of the votes say its bad. The sheer number of votes says it does something well, so it gets 5 stars regardless. This works because feedback of bad content is still given, but it does rely on a lot of votes (although this weight is decided per game).
- imdb/rt both have ‘expert’ ratings as separate category. which works there because movie making regardless of category has a lot of things to which you can see that they showed skill or not. Flaws can sometimes be used in an intended way to set a joke, or because the alternative is too hard/expensive to implement (visible camera crew in gladiator is an example of this). Experts can judge this objectively.
- google showed that most of the time the sheer number of reviews is a far better indication of quality, hence a lot of sites/stores just generate a lot of AI reviews. A 4/5 stars with 1000 reviews is far more trustworthy than a 4.5/5 stars with 90 reviews. This being a free sites makes this false reviewing easy to do, so its best to not realy do anything with it, rather than just giving the creator a message of ‘someone liked your content’. Good content will get a lot of these over time.
So yeah, dont bother implementing anything as improvement for rating, its not going to work better than what we have now. And what we have now does exactly what it should do: it tells the creator they made something good. And even if its just 10 votes, thats at least 10 people that will enjoy your content.
That is still a reward, as the niche you made something for showed they are there. Sure 100 sounds nicer, but if you made a rare niche thing, thats unrealistic to expect. But at least you know you arent alone. And who knows, maybe someone will also start making more of such content.
Being reliable as scripter will sometimes cause someone to follow you, which then indicates you realy are doing a good job. A single 1000 upvote thread isnt very likely to get someone to follow you, multiple threads are. And that is the best metric to use.
And yes, if a creator is falsifying that, well, good luck… there are no leaderboards with heavy competition going on here. And if you are trying to go for it and people notice it and use AI accounts, this will eventualy be noticed anyway. And corrections here are usualy more devestating to you, as this most likely just means clearing all upvotes (including legit ones). Leaderboards are meaningless, since if something doesnt align with your preferences, its not going to attract attention anyway.
Sure, it might exist as sorting option, but its very clear that most of the time these are time tested scripts, and its highly unlikely you are going to exceed them, becaues it takes time to build the numbers. AI isnt going to help here by boosting numbers. Tag filtering does, which we already have, and if your content is good, its going to show up high in that list anyway.
But from my experience, my default sort order is new/activity. As good content gets an occasional bump anyway. And this shows me a lot more threads over time, allowing me to find more good scripts faster.
