Thread:Gsimenas/@comment-36433523-20190415175447/@comment-6697950-20190416223415

Thanks for the information. Didn't know that was a thing. Though, that doesn't change the fact that barely any of the regulars voted on it.

And yet in all of the conversations we had, as well as the posts that I looked over on Discussions, I never saw any real arguments to support the change other than the fact that it's an anime image, so sorry if I got the wrong impression. If we decided to change infobox images due to subjective reasons like "I like this image", we'd be having infobox image changes at least once a week. To avoid such scenarios, we require objective arguments to prompt an infobox image change - there has to be something objectively better about the new image to consider a change. The fact that I didn't hear of any such arguments from you is one of the main reasons why I didn't support your proposed change.

For the record, Alice had the same issue at first. Some people tried to push an image from her first appearance (https://swordartonline.fandom.com/wiki/File:Alice_Synthesis_Thirty_gazing_at_Eugeo_at_Kirito_during_her_introduction_-_S3E10.png), but we rejected it due to various issues (shadows on her face, bland background, lack of expression etc.), which made it no better than the one we used from the novel. On the other hand, the current image didn't have those issue + it portrayed her expression better than the equivalent image in the novel, thus there was sufficient reason to make the change for her. Meanwhile, anime images for Eugeo don't really offer any significant improvement over the LN image (minus the fact that abec sucks at getting colours right), seeing as the LN image is pretty good on its own, so it basically comes down to subjective aesthetics. The eye colour is pretty much the only objective argument I can think of to support the change.

I only undo content edits when it's either off topic (e.g., when someone starts writing about relationships or just a plot summary on a personality section)/irrelevant (e.g. digressions about non-important plot events), or there's too little useful content in them to bother fixing them. And the latter criterion is mostly only used when someone spams poor quality content without showing any effort to improve. When someone doesn't bother to wait for a review yet churns out content with many blatant issues left and right, I have to take more drastic action to ensure that I can keep up with the edits, or I won't have enough time to review all of that content (I already have experience with where unsupervised content additions like that end up - gigantic walls of text with little to no focus on what's actually important - which is why I've become stricter on that). In all other cases, I try to fix the issues and make improvements to the edits, rather than undoing the edits. If you want to contribute like that, I recommend starting small and letting me review your edits to see if there are any issues with them (e.g., incorrect tense, style/register issues etc.) that you need to improve on before undertaking larger projects. Most new editors experience the same issues until they get used to the standards, after all.