RMN V4.5 (AND BEYOND) FEATURE IDEA LIST

author=Dyhalto
The solution to the first problem is simple enough : No star ratings on demos, unless it's so promising (like Homework Salesman) that 4-5 stars is indisputable.
As for "stale" reviews, have a way in which the gamedev can appeal to the staff (Soli) and request that an old, negative review be removed for reasons of . Problem solved.

My issue here is I don't like the idea of developers being able to appeal to get reviews they don't like removed. It creates bias; reviews that give good ratings would stay, while developers could petition to get bad ones removed.

The whole "Star ratings" vs x people liked, y disliked" has been debated many times before. The problem is that people believe this will be a perfect solution, because they'll see "456 liked, 7 disliked" and say "wow, this one must be good!" or "7 liked, 168 disliked" and say "wow, must be bad!"

On this site, you're far more likely to see "2 liked, 1 disliked" and that really tells you nothing. There was also concern it would give the various 'cliques' too much influence over their ratings of their friends' games.

It is a quick visual representation of the esteem that others placed on the game. Without it, you are left with what the developer tells you about the game (almost invariably "this is the coolest shit ever") or what RMN tells you about the game (here are some statistics and numbers with viewcount and shit).

It is not a perfect system, and I am tired of getting into over and over again, but it does have value.

I am not removing star ratings from reviews. Reviewers have (and always have had) the option of not using star ratings. If you want to effect change, talk to the reviewers and convince them not to use star ratings.

Reviews are too much information to be quickly consumed in a list format. Removing the rating means that players, when browsing, only sees what the developer is telling thing (and the few tidbits stats that RMN shows).

If only there was a way to balance out what the Dev is telling you and what RMN is telling you with something... like, say, what other players think... in a quick way... I know let's add a rating to reviews that we can display in a compact format!


I can refine the way the average is calculated (I outlined that some months ago in the last topic - it just didn't make the cut for v4.4).

Game pages, screenshots, and review content sell the games - a game rating helps them get into the store in the first place.


PERSONAL ANECDOTE: I hate game sites that don't have user ratings, because without them the list of games looks like one big undifferentiated morass of schlock. It makes it hard to find anything worthwhile when you are given no independent information of what others think was worthwhile.


Also, on the subject, I want MORE ways and ideas to help people find games they might like! I am toying with the idea of making a Favorite button, and using information like that to:
1) publicly display who favorited this
2) generating lists like "people who favorited this also favorited X, Y and Z"
3) track upward trending games
4) ???
5) profit!

PERSONAL ANECDOTE: I hate game sites that don't have user ratings, because without them the list of games looks like one big undifferentiated morass of schlock. It makes it hard to find anything worthwhile when you are given no independent information of what others think was worthwhile.

On the contrary, I never look at user ratings because as the number of ratings increases, it converges on 3.5-4/5, thus telling me no useful information whatsoever. When I'm buying something online, I always read the two or three highest reviews and the two or three lowest reviews, since that gives the broadest picture of what people like/dislike about it.

2) generating lists like "people who favorited this also favorited X, Y and Z"

See, now there's some useful information. The primary way I discover content I might like is by looking at what people who enjoy what I like also like.

author=Sailerius

PERSONAL ANECDOTE: I hate game sites that don't have user ratings, because without them the list of games looks like one big undifferentiated morass of schlock. It makes it hard to find anything worthwhile when you are given no independent information of what others think was worthwhile.

On the contrary, I never look at user ratings because as the number of ratings increases, it converges on 3.5-4/5, thus telling me no useful information whatsoever. When I'm buying something online, I always read the two or three highest reviews and the two or three lowest reviews, since that gives the broadest picture of what people like/dislike about it.

Actually, it has been studied and it found that most reviewers are people who either really loved it or really hated it, and it's the middle scores where there is a lack of reviews/information. People who have a more extreme reaction to something (positive or negative) are more likely to take the time to review something, than someone who found the experience mediocre or merely satisfying. This is just human nature and no review system can "fix" this.

2) generating lists like "people who favorited this also favorited X, Y and Z"

See, now there's some useful information. The primary way I discover content I might like is by looking at what people who enjoy what I like also like.

Yeah, I love this shit too. That's why a long (long) time ago I tried starting that voluntary "People who liked this also liked..." gamepage thing. (You can see it in action on some of my older gamepages).

If 15 minute dose not passed or someone else didn't posted after you, you can delete your post. it will deal with accidental double posts.

author=meisam
If 15 minute dose not passed or someone else didn't posted after you, you can delete your post. it will deal with accidental double posts.

I'd just like it if people could report their own posts. I'm not crazy about the idea of people deleting their own posts, but being able to bring things to the staff's attention when you yourself make a mistake sounds useful in general.

Star ratings are toxic for as long as there's no way to remove no longer relevant reviews. There's no way to implement a system that will remove no longer relevant reviews without it being exploited to inflate a game's review score. Getting rid of star ratings is the only way to resolve the problem.

Divorcing ratings from reviews also 'fixes' that problem, or at least it doesn't exacerbates it, which is what matters. Like Solytaire mentioned, removing reviews can also lead to abuse, so we'll have to live with that. (Or maybe if there was a yearly 'purge' to determine what reviews are still valid...) And getting rid of Star rating system is just too extreme. It may not tell YOU anything, but no piece of information on its own will. The point is to get the big picture with all the information available to you. In this case, the more the merrier.

Even a "You may also like" system has the potential of not telling you squat. It's not like people have different tastes or anything, and people who may have liked a Survival Horror may also have liked a Dating Sim. You'd have just as good chances of finding something you like just searching by genre or tags. (I still like the "you may also like" idea and would like to see it implemented. ;P)

On the contrary, I never look at user ratings because as the number of ratings increases, it converges on 3.5-4/5, thus telling me no useful information whatsoever.

I fail to see the problem here. If a rating converges on a 3.5-4/5, so what? In our case, wouldn't it be safe to assume that a game is at least of "above average" quality? (for amateur standards, that is) How is that bad? How is that useless information? That's purpose of ratings: it's just a metric to help you appreciate the merit of games, not a way to exalt some game above the others... Besides, assuming you could see who, and how many times a rating was given, that could help you further discern if a game is worth your time.
_
Ok, crazy idea time. How about we let people fill a questionnaire in order to rate games? For example, in the case of graphics, some of the more common concerns are whether if the tiles are properly used, or if the graphic styles clash, or is the maps are of reasonable size, etc. So, the questionnaire would ask you to rate all these items individually. At the end, the ones that score the best could be summarized in the game's profile as 'features' e.g. "This game makes good use of the RTP" ...This way not only the number would have a reasoning to back it up, but it would be less prone of bias than with some reviews that don't fully address all the aspects of a game.

author=meisam
For reviews you can add wait system, as already suggested. if a particular review was useful or majority of people disagree with a review, it have a lower effect on total rating.

At least, there ought to be something as they do elsewhere (amazon.com comes to mind) where people's reviews are voted as being useful or not by other users. Not sure, that the weight of the rating associated with the review on the overall rating of the game should be changed, though. Individual reviews would be scrutinized for their quality of their content. I still think there's too much emphasis on ratings ...

Actually, didn't that one poll indicate that most people base it off screenshots? Besides, if only one jerk is reviewing it and no one else is bothering to write out their opinion maybe the game deserves its score?

For me, if a game has a very low score I'm intrigued as much as a very high score. And perfect 3s get my attention too. Probably the one score least likely to pull me if I based it on only score is a 2.5.