Forum Settings
Forums

"Confusing" reactions to reviews shouldn't count towards how high up they are for an entry

New
Pages (2) « 1 [2]
Dec 17, 2022 6:08 AM
Offline
May 2016
26
leuscius said:
why shouldn't they? they still count as engagement, and whichever gathers the most would naturally assume a top spot due to high traction.


I think the review system should be promoting quality reviews, not reviews that generate "engagement". A confusing review is a bad review by definition.
Dec 18, 2022 3:52 PM

Offline
Feb 2021
766
random_savage said:
leuscius said:
why shouldn't they? they still count as engagement, and whichever gathers the most would naturally assume a top spot due to high traction.


I think the review system should be promoting quality reviews, not reviews that generate "engagement". A confusing review is a bad review by definition.


quality is an ambiguous term and a confusing review can simply mean reactionary which, by definition, is still a determinant to generate "engagement" — negative or otherwise.
Dec 19, 2022 1:00 AM
Offline
May 2016
26
leuscius said:
random_savage said:


I think the review system should be promoting quality reviews, not reviews that generate "engagement". A confusing review is a bad review by definition.


quality is an ambiguous term and a confusing review can simply mean reactionary which, by definition, is still a determinant to generate "engagement" — negative or otherwise.


Don't try to tell me that "confusing" is a reaction that indicates a review is worthwhile. Just because the concept of engagement is easier to quantify does not mean it is a good or better benchmark. If the city asked feedback on two designs for a new road sign and 300 people said option A is confusing and 200 people say option B is good, that does not mean it is better to put up sign A because the quality of a sign is ambiguous but it is clear more people felt sign A was worth engaging with.

Reviews serve a certain purpose. The purpose is to review, not to cause the greatest degree of "engagement". Nobody is earning ad revenue off these horrible reviews.
Dec 19, 2022 5:06 AM

Offline
Oct 2018
170
I have been thinking about this ever since they changed the old review system. Confusing is used as a downvote which should not be promoted over more upvoted reviews. I could write a completely nonsensical review about an anime attacking things which didn't even happen, and people will downvote that leading it to come up to the top reviews. A lot of users will see the not recommended on the top and make false assumptions about the show. While mass downvoting by rabid fans is possible, a confusing review is more usually than not an indication of disagreement by the general fans over that review. Having those reviews on the top where a casual fan would expect the reviews most people agree with to be is very disingenuous. It would be best to stop counting the confusing votes for the top place, and if people do want to see not recommended reviews in particular then have a sorting option which shows them the most upvoted not recommended reviews and the most confusing reviews.
Dec 19, 2022 4:03 PM

Offline
Sep 2018
5281
Just look at the top reviews for Isekai Ojisan and Bocchi The Rock lol, almost 1500 and 2000 reactions respectively, yet the positive reactions (upvotes) are miniscule in comparison. These two reviews possibly wouldn't even had made the front page before the update. Just seems counter intuitive.
Dec 19, 2022 8:10 PM
Offline
Jan 2021
1
Agreed, all the most controversial reviews aggregate to the top.
Jun 1, 2023 6:55 AM
Offline
Oct 2021
417
Absolutely agree!

Oshi no Ko has two “not recommended” and one “mixed feelings” for its top reviews despite being the TWELFTH HIGHEST rated entry on the site.

If you expand the reactions, the majority of them are confused, which is extremely misleading for people who don’t know the site as well, and might think the anime is bad when in reality, many people love it.

At the very least, I wish the engagement would be more explicitly stated on the reaction pages, given that many people (myself included, before figuring out the truth) treat it as a downvote when in reality, it means just as much to the system as a “love it”.


Candy 2023:



Candy 2024:
Jun 1, 2023 10:21 PM

Offline
Oct 2013
8743
How about just getting rid of emoji reactions and going back to the good, old system with marking reviews as either helpful or not helpful?

What's confusing the most about the "Confusing" reaction is the exact meaning of it. From what I noticed, a lot of users have been using it to express their disagreement with a review's contents. Is it fair to not want to return to "not helpful" functionality, but implementing something trying to imitate it, and failing in doing so (after all, any reaction given to a review makes its positioning higher).

About the suggestion from the original post, I wouldn't want to exclude any emoji from being capable of bumping a review. By doing so towards "Confusing" reaction, some folks could have stood one day against other reaction emojis, like dunno, "Funny".
✨Saint Seiya Club🌠
Aug 13, 7:17 PM

Offline
Jan 2015
357
Up.

By now, all of us should've realized that the "confusing" button is used as the substitute for "dislike" button.
If we find a negative review to be reasonable or well-written/accurate, we would choose the other more positive-sounding buttons.

I know that the current random display review system makes this more tolerable, but i think separating the count for like/positive reactions and dislike/negative reactions for a review would be better.

How about 2 buttons: "well-written/accurate" and "poorly-written/inaccurate"? (and only count the "well-written/accurate" for the ranking).
For each review, the current count for "confusing" button will go to new "poorly-written/disagree" count/button, and the current count from all other buttons (nice, love it, funny, informative, well-written, creative) will go to new "well-written/agree" count/button.

Doing this will also allow the top 3 display system to be re-implemented since all of them will be the most objectively agreed upon. We can still keep the current blue / gray / red indicator (recommended / mixed feelings / not recommended) to differentiate the nuance of each review.

To prevent review voting manipulation, we can also make it so that there are some requirements for each user to be able to vote for a review, like minimum 30 active days on MAL for example (there will be no backlash for implementing this requirement except from those who manipulate votes).
On top of this, we can also still hide the voting button behind the "read more" button to make it a hassle for voting manipulators.

(the terms "well-written/accurate/poorly-written/inaccurate" are chosen to make the voting ecosystem more objective and dependable, instead of the terms "like/dislike" or things similar to them that are oriented towards the reviewer instead of the review a.k.a. elevating the ego or hurts personal feeling).

Example implementation:
base display:

when we click "read more":


The two buttons will not be visible or cannot be clicked if the account hasn't been on MAL for at least 30 days, for example. And if it's possible (i don't know much about web programming), we can make it so that it's 30 days of activity, not 30 days since account creation (x days since account creation is easy to abuse). The daily activity doesn't have to be consecutive, just record each day that the user logins or do something in MAL and sums them up.

Thank you for your consideration.

---

@Adnash Thank you for your kind thought/feedback. I hope that if they decide to make changes to the review system again in any way they see fit, the outcome will be for the betterment of the MAL community. Cheers :)
XavionAug 15, 12:36 AM
   

• reminder to do something today that will make your life better •
• true failure is when you stop trying •

Aug 14, 6:54 PM

Offline
Oct 2013
8743
Reply to Xavion
Up.

By now, all of us should've realized that the "confusing" button is used as the substitute for "dislike" button.
If we find a negative review to be reasonable or well-written/accurate, we would choose the other more positive-sounding buttons.

I know that the current random display review system makes this more tolerable, but i think separating the count for like/positive reactions and dislike/negative reactions for a review would be better.

How about 2 buttons: "well-written/accurate" and "poorly-written/inaccurate"? (and only count the "well-written/accurate" for the ranking).
For each review, the current count for "confusing" button will go to new "poorly-written/disagree" count/button, and the current count from all other buttons (nice, love it, funny, informative, well-written, creative) will go to new "well-written/agree" count/button.

Doing this will also allow the top 3 display system to be re-implemented since all of them will be the most objectively agreed upon. We can still keep the current blue / gray / red indicator (recommended / mixed feelings / not recommended) to differentiate the nuance of each review.

To prevent review voting manipulation, we can also make it so that there are some requirements for each user to be able to vote for a review, like minimum 30 active days on MAL for example (there will be no backlash for implementing this requirement except from those who manipulate votes).
On top of this, we can also still hide the voting button behind the "read more" button to make it a hassle for voting manipulators.

(the terms "well-written/accurate/poorly-written/inaccurate" are chosen to make the voting ecosystem more objective and dependable, instead of the terms "like/dislike" or things similar to them that are oriented towards the reviewer instead of the review a.k.a. elevating the ego or hurts personal feeling).

Example implementation:
base display:

when we click "read more":


The two buttons will not be visible or cannot be clicked if the account hasn't been on MAL for at least 30 days, for example. And if it's possible (i don't know much about web programming), we can make it so that it's 30 days of activity, not 30 days since account creation (x days since account creation is easy to abuse). The daily activity doesn't have to be consecutive, just record each day that the user logins or do something in MAL and sums them up.

Thank you for your consideration.

---

@Adnash Thank you for your kind thought/feedback. I hope that if they decide to make changes to the review system again in any way they see fit, the outcome will be for the betterment of the MAL community. Cheers :)
@Xavion Out of all suggestions regarding the review voting system I have read pretty much since the current emoji system was implemented, this one is the best option idea. It combines both systems (up/downvote and reaction-based) by mixing them in a satisfying way, taking the best from each without making any of those models visibly dominant.
✨Saint Seiya Club🌠
Pages (2) « 1 [2]

More topics from this board

» Add "author" to manga DB entry page

kuroneko99 - Dec 17

2 by kuroneko99 »»
7 hours ago

» Reviews

x_scolopendra_x - Dec 16

8 by ACasualViewer »»
Today, 7:52 AM

» More Two Factor Authentification Options

The-Demiurge - Dec 17

12 by mdo7 »»
Yesterday, 9:10 PM

» Option for draft mode on anime reviews

ggol - Dec 17

1 by The-Demiurge »»
Yesterday, 7:13 PM

» A new badge for when you buy someone a gift

_cjessop19_ - Dec 19

1 by Rosy_Rose »»
Dec 19, 9:49 AM
It’s time to ditch the text file.
Keep track of your anime easily by creating your own list.
Sign Up Login