The amount of stars that defines above or below average is defined by the audience.

In this case there's no money involved, so the results will be honest, as opposed to pro reviews which have a scale of 1-10 but use 4/5 as terrible and 6-7 as average.

Having values adjust dependent on how a user has rated in the past can't work, because people's opinions change over time, same as a group is made up of individuals. 

Rating systems only make sense if there are guidelines of what they are highly visible on the web page/site and the ratings are controlled/submitted by a small collection of respected individuals. 

ijed: I would love to visit your planet! :/

both for the honest people and the professional reviews.

why would 6/7 be average for you? mathematically it would be at 5.5 for a 1-10.
---

this about the user ratings, not the "official" rating from quaddicted. I will never make the range wider than 1-5. I like 1-5, it works well and is granular enough to allow honest and helpful ratings without becoming esoteric 76.2/100. 

oh, misunderstood probably. what do you with honest?

a histogram is not something with a user's history. it would show the rating options and how many people used each. 

maybe if you start all maps off with 1 star, them if people feel it's better, they will feel more compelled to vote to bring it up. 

Why not let people create their favourite lists of maps and post/link them to the various maps? I know Lvl does this and it works really well because you get to see what people like and if you recognize the nick it means more. 

i would personally prefer to see the 'official' ratings removed; they kind of null the effect of the user ratings, and really they are just one other person's opinion (so ideally each one should get added to the personal ratings as individual votes, if that is in any way possible without a ton of work). i appreciate this may have already been suggested and dismissed and i won't kick up a fuss if it's shot down; it's just an opinion

regarding scoring, a map's current score might affect how i rate it if i'm being honest. if it has 5 stars and i absolutely hated playing it, i'm probably going to give it 1 to knock the overall score down even if it technically warrants a higher rating. 

oh, misunderstood probably.

Yep. 

i've noticed on the iPhone App Store, and on Yelp, people will explicitly say "giving this a 1 to counteract all those 5s" -- which is of course bullshit reasoning but people thing the idea behind voting is to get the outcome YOU want instead of measure aggregate opinion. 

also because when you dislike something, and then find out everyone else loves it and sings its praises, it kind of makes you dislike it even more. as illogical and dumb as that seems. maybe it's just a subconscious refusal to admit to having backwards taste... 

Official ratings won't go away because exactly that stupid "I do not vote honestly but trying to influence the rating" behaviour. negke and I know better!.

The histogram seems like the best idea to show those.

sock, that seems like a nice idea. The lvlworld interface makes me rage any time I use it so I keep my time there on a minimum. Could not find any lists on a quick glance, can you link me? I'd make it free lists where people can create it, give a title and then add maps. So it could serve as bookmarks too, which would be handy. Best with a private/public toggle too. 

negke and I know better!.

looking at some of the ratings, i'd question that claim. :P

ok, i'm half-joking there. but still, there is no-one alive capable of convincing me this is a 5 star map. honestly, without any bias i'd have given it a 2. the official and user ratings actually dismayed me enough into lowering my vote to a 1...

(i guess the above answers why i generally don't vote...) 

Some people just hate things regardless of logic, common sense or if 99% other people on the planet like them! Down voting something because you feel the rating should be lower is a spiteful thing to do. The best thing anyone can do is post why they think something is wrong to them.

spirit, the lvl site is ok, but yes it can be difficult to workout where things are sometimes. Here is a perfect example of what I mean, someone reviewing maps, has something useful to say about them and offers good feedback. Over the years I have come to respect the views of this user and when I see a comment/vote by them I take notice.

http://www.lvlworld.com/#p=member&l=1059

One thing I really like about this system is that user accounts are linked to what they vote so you know if they are being honest or just an idiot. I like the accountability of this because it makes people think twice about down voting and vote rigging. 

http://xkcd.com/1098/ 

http://goatkcd.com/1098/sfw 

I rated every single map not made by me to be a 4 or lower, and rated my own 5.

GLHF everyone else. 

that map was pretty good too 

...and maybe a few others...


...and I likely got bored before rating all of them... 

The reason I asked is because there seems to be a conflict between the editor's rating system and the way users interpret their own ratings.

There was a particular user who rated maps in this typical undifferentiated "all or nothing" fashion that's so common all over the internet. It was especially annoying, because he downrated maps when he was just too dumb to run and play them correctly, if there was a heapsize error, for instance, or he got lost, killed by a trap, blocked by an awkward clip brush. And at least one point rigging, too.

Though it also made me wonder about the systems. I never really thought about it and just assumed people would more or less base their ratings o similar criteria like we do, which tries to go beyond personal preference to a certain degree. But then, it suddenly seemed more likely that they interpret the user rating more like "How much did you enjoy this map?" - "Very much" (5)... "Not at all" (1), which could well result in low ratings even for 'good' maps. Hence my question without further information.

Of course, we know that the editor's rating is a questionable thing. The idea behind is to give people a rough indication of the overall quality of a release, so that they won't have to take their chances and end up with only the worst maps. Ideally, the whole system would be entirely made up of user rating like on DW, but the problem is that each map would need to have a minimum number of ratings, ten or so, for it to work (to become somewhat credible and counter individual troll votes). It's no problem on DW because of its age and the fairly high number of users. However, I doubt it would be easy to establish for Quake, considering how few people are willing to participate. Even on Func, it seems many of the regulars can't even be bothered to leave some quick feedback on a map they've just played, so it's even less likely to work on an external site.

How would it influence your ratings if the system wasn't about numbers (of hearts or stars), but corresponded to our text ratings ("poor", "average", "nice")?

Also, I believe it's about time the Quake Injector gets updated with features like rating and commenting (account-baded), maybe even something to emphasize the desirablity to rate after playing (pop-up, red outline, or the like). 

Perhaps you could set up some guidelines first. For example use a 1-100 point system with say 1-20 points for texturing and lighting, 1-20 points for the layout and clipping, 1-20 points for item and enemy balance, 1-20 points for game play and pacing and 1-20 points for enjoyability (or some mixture thereof with less weight given to some aspects and more weight given to others). Then you can divide and average the final number to the range that you prefer (1-5 or 1-10).

That way you might be able to get some feel for where the overall opinions (good and bad) come from and where a particular level failed to meet others likes and dislikes. 

I think that would be way too complex. even with limiting each to a reasonable 5 scale it would be even harder to get people to rate.

the user ratings are meant for random Mr Smith. 

metacritic weights different reviewers based on how reliable they are, i think... (reliability = closeness to norm, i think?)

Also, Some review aggregators will ignore the outliers... so a bunch of 4s & 5s with a single 1, the 1 would be ignored.

Both of these require enough raw data to even have outliers and norms, of course. 

sock: Those things exist at Quaddicted (comments, ratings and tags history of users) but they are not linked and only available for registered users. https://www.quaddicted.com/reviews/user.php?username=Spirit

I think I should probably through my privacy concerns away and link them from comment pages.

I really like the idea of user-created lists/selections so I wrote some thoughts on implementation: https://www.quaddicted.com/forum/viewtopic.php?id=84 

it would even be relatively easy to measure how long players spend with a map. But at least until oct, i have zero time to spend on it. It would really be awesome if someone could take over. I'm here to answer questions. 

nah, that would need to be in-engine. you could not know about level changes, pauses etc from measuring the time quake was running from the outside.

full offline support and comments and ratings (and tags) would be my favourite issues.