It seems that you're using an outdated browser. Some things may not work as they should (or don't work at all).
We suggest you upgrade newer and better browser like: Chrome, Firefox, Internet Explorer or Opera

×
The mass public is, and will always be, more polarized in their opinions than some objective specialist that gets paid, because he has the rare ability to not be like the mass public. While it is understandable that someone would desire the public to be more level headed or rational about their opinions, it is ultimately asking something of it that it isn't really capable of.

This does not invalidate the usefulness of an imperfect ratings system. While you may have to re calibrate what those stars really mean, it is far better than just not having any public feedback at all, and while the system may have difficulty telling you if you will like or love a particular game, it can warn you if something is wrong. Beyond that it is tough for anyone to really guarantee if you will like vs love something, and at some point you just have to find out for yourself.

With that said there is something I like to do when looking at ratings that I feel really helps. It works best with music on Amazon, but I think it helps here as well. I look at how many people rated the item. Low numbers often means most people are apathetic about the item, while high votes with high stars are a strong indication that an item isn't just getting votes by a handful of people with an acquired taste. In my opinion it tends to be a better predictor of quality that the stars themselves, and put together you can get a reasonable amount of information to help make judgments on the quality of an item. (this isn't true for all kinds of items, but with highly subjective entertainment experiences it tracks well)
Post edited May 26, 2013 by gooberking
avatar
solzariv: User-based rating systems are never to be trusted anyway, because different people have internalized different ideas about what the scale of the ratings is. Some people take it to be linear, while other people interpret it like school grades ("anything below an 80 is utterly awful!").
Not a problem for user ratings. Due to sample sizes and the general lack of correlation, it all averages out; I find it highly implausible that fans of one game are statistically more likely to lean toward a particular interpretation of the rating scale in a way that affects comparison.

avatar
solzariv: GOG should do away with stars and number-based ratings entirely, and make it entirely text-based.
People like to rate games; each game has many more ratings than it has reviews. Text-based rating would mean a flood of "teh bestest gaem evar" reviews; and in the absence of stars, it would legitimize them. Take Anodyne, for example. I love that game, I paid upwards of $100 for it, and I don't think the GOG reviews are doing it justice. At all. But I don't have the time or the skill to write a useful review, so I had to express my sentiment by rating it 5 stars and leaving it at that. Without the stars, a "teh awesome" would've been actually useful information expressed in an inconvenient form.
I think the star rating system is some sort of first orientation for me. Also, the scale probably adjusts accordingly. So let's say most people dish out 5 stars for their favorite games. If there's a game with way less stars than 4 on average (just an example) I probably should start reading the user reviews on here. If those aren't really that informative or in the case I want even more info, I'm looking for scores of old reviews that were done back in the day when the games got released. Luckily I have played a lot of the games on here myself back in the day or at least read about them in various magazines, so in most cases I just buy what I like and don't need to rely on more or less objective ratings or reviews.

Anyway, if you check out modern professional reviews, there's not much use of the full rating scale either. How many high profile titles are there that only get like 20-50% or even 60%? Not many. And since a lot of the good old games were high profile titles, the same rule applies.
Post edited May 26, 2013 by Dragonfly2012
Individual user input. You'll never get it completely right. Might as well get used to it and do your own research not centered on a 1-5 point system.
But Dungeon Keeper 2 is a 5/5 game for me. Can I rate it that way? :-(
avatar
gooberking: The mass public is, and will always be, more polarized in their opinions than some objective specialist that gets paid, because he has the rare ability to not be like the mass public. While it is understandable that someone would desire the public to be more level headed or rational about their opinions, it is ultimately asking something of it that it isn't really capable of.

This does not invalidate the usefulness of an imperfect ratings system. While you may have to re calibrate what those stars really mean, it is far better than just not having any public feedback at all, and while the system may have difficulty telling you if you will like or love a particular game, it can warn you if something is wrong. Beyond that it is tough for anyone to really guarantee if you will like vs love something, and at some point you just have to find out for yourself.

With that said there is something I like to do when looking at ratings that I feel really helps. It works best with music on Amazon, but I think it helps here as well. I look at how many people rated the item. Low numbers often means most people are apathetic about the item, while high votes with high stars are a strong indication that an item isn't just getting votes by a handful of people with an acquired taste. In my opinion it tends to be a better predictor of quality that the stars themselves, and put together you can get a reasonable amount of information to help make judgments on the quality of an item. (this isn't true for all kinds of items, but with highly subjective entertainment experiences it tracks well)
You make some excellent points there. I especially agree with you about the significance of how many star votes there actually are.

I noted elsewhere in the sale thread for this weekend that it was interesting to me that the game Sniper Elite got 3.5 stars here which is just generally speaking not a good sign. There was 203 ratings. I went and checked Gamespot and found they gave it a 7.6, 1,971 users there gave it an 8.0 on average and Metacritic gave it a 76. Also, 100 Metacritic users gave the game a 8.5 on average. The "Most Helpful" review on GOG gave the game 3 stars and is titled: "A great idea badly executed."

So, what to make of all that? Well, I can't be sure but for whatever reasons this title is not the GOG crowd's cup of tea compared to many others. On the other hand, everybody else I checked with seemed to like it, faults and all and it does have some.

I said in the other topic about this stuff and I'll repeat it. Cast a wide net. You catch more fish that way.

I should clarify, that is just a general statement and not directed to any one person in particular. I thought of this after posting this as a reply that you might think I was telling you this specifically. I just mean we all do well to check multiple sources. I think probably most of us agree about this anyway.
Post edited May 26, 2013 by dirtyharry50
avatar
solzariv: GOG should do away with stars and number-based ratings entirely, and make it entirely text-based.
Actually, GOG should base ratings entirely on like-dislike system.
One problem with critiques like these is that we are dealing with collective decisions here. So every single person who has ever given a rating or written a review may agree with the criticisms put forth here and it still wouldn't change a bit if they all thought the problem was everyone ELSE having rose tinted glasses.

I mean surely there ARE some games that are worthy of high ratings and it is quite likely that everyone who has rated a game on GOG believes the rating was merited. Even if they happend to think there are too many 5 star ratings. Everyone who has given a 5 star rating could potentially think there are too many of them and so we are stuck with the current situation.
avatar
keeveek: But Dungeon Keeper 2 is a 5/5 game for me. Can I rate it that way? :-(
I don't really like Dungeon Keeper but this is the second post in here in this vane and I think it speaks volumes; namely is says that keeveek and Licurg have no taste. *RIMSHOT* I'M JOKING! XD

Seriously, even if it's only one in a group of one thousand that think something is 5/5 then that one legitimizes the right for that review to be posted and seen/read/heard. I don't want to be told my opinion doesn't count just because I'm the only one that has it. If I give a 5/5 that you object to so much, then counter my review with one of your own. Tell everyone my review can be used to fertilize the lawn and then let the reader decide.
Post edited May 26, 2013 by tinyE
avatar
solzariv: GOG should do away with stars and number-based ratings entirely, and make it entirely text-based.
avatar
Fenixp: Actually, GOG should base ratings entirely on like-dislike system.
Thumbs up!
I don't understand how would you like this to work (I mean the older titles, not the new ones obviously)... if somebody played a game when he/she was a kid and now couldn't write a "I LOVED THIS BACK THEN!" review, but would buy the game, play it again, find out it sucks now and write a "NOT AS GOOD AS I REMEMBERED / EVERYBODY THINKS" review?

I mean, how's that different? they're both absolutely individual opinions, every single person can have different experience, maybe now somebody don't like the genres they liked before (and you might now) etc.

I never look at the stars to find out about the game, and any reasonable person shouldn't - if you loved the game as a kid and are thinking whether or not to buy it now, NOBODY can tell you if it's still as much fun as back then, that's completely up to you... so you either do some research, watch some videos to remember how does it play, or you just risk it.

for me, the stars here are kind of like a "honor" mark - if they're remembered as being great, they have a high mark, if not, they have a lower one, but it still won't (and can't) tell you whether you'll like it or not.
avatar
jjsimp: I have been criticized that I never give out the perfect 5 out of 5 or 10 out of 10 by friends and peers. It's a pet peeve of mine that nothing is perfect, thus no perfect scores are warranted. Even when grading myself for my boss' annual review of me, I am brutally honest.
Well, if 4 is the best grade anything can get from you, then you're not using a 5 scale but a 4 scale. But then you shouldn't give a 4 because perfection is unobtainable, so you're down to a 3 scale. See where I'm heading?

The error lies in your thinking that a 5/5 means perfection. It becomes obvious when you transform the numbers into percentages.

1/5 -> 1%-20%
2/5 -> 21%-40%
3/5 -> 41%-60%
4/5 -> 61%-80%
5/5 -> 81%-100%

So everything above 80% actually deserves a 5/5 (obviously I'm using a linear approach here and not that approach from school, where a 5/5 would end up somewhere around 90%-100%).

The star rating here at GOG actually is useless for me. But that's okay, because I've been a PC gamer for 20 years and I know most of the old games, even if I haven't played them yet. But the reviews sometimes are helpful when they actually mention pros and cons of the game.
Post edited May 26, 2013 by xy2345
Perhaps we could eliminate the stupid star system and instead go categorical with 1-100 rating system?

You know, true oldschool when reviews could be trusted and made sense?
I'm afraid where old classics involving childhood nostalgia is concerned, people vote with their hearts instead of their brains. They simply remember how much fun they had back in the day playing the game, while most of the details are a blur. Removing a star seems difficult unless you remember some specific annoying detail, but most of the time you'll only remember the good moments and forget about the bad.

I wouldn't trust the reviews you read here, but most of the time you can easily tell the ones based on nostalgia from the ones based on recent play. You'll just have to wade through the reviews.

Also, keep in mind that as children or teenagers, most players had a different taste or critical sense than they do now.

(This reminds me of the time I tried to fly a kite as an adult in a huge park I have next to my apartment. I used to fly kites there as a child and figured I'd be reliving some old memories from my childhood. I bought a kite and when I went out...

1- I was the only adult trying to fly a kite without a little kid nearby.
2- Getting a kit to fly involves running. As a kid, that was awesome, as an adult, I just wiggled my arm and tried to get that damn thing up somehow.
3- The kite fell a few times, rolled around with the wind and the string started to tangle up. A flood of memories suddenly invaded me, of me trying to untangle string as a kid. My brain immediately went 'Ain't nobody got time for THAT!' and I gave the kite away to a neighbor's kid.

The memories of kite flying are still awesome, but I can no longer enjoy reliving them.)
How about a system where the possibilities are something like this:

I would not recommend you get the game.
I would recommend you get the game if you are a fan of the genre or franchise.
I would recommend you get the game.

You could do all kinds of variations on that including "You absolutely should get the game" and the like. You could have as few as two levels with the middle option above taken out.