Best review scale?
 

My site staff has been wondering lately, what's the better review score scale? x/10 or x/5?

Right now, we use /10, but not in the same way other site use it. We've seen it as a waste to throw away half of the scale (most places, anything below a 6 is pure garbage, and there's no difference between a 5 and a 1), so used the mathematical average of 5 for our "average" game.

It seems like it works well enough for us and the readers who actually try to understand it, but sadly, this is the internet, and not everyone reads the "Verdict" underneath where it says the final word like "Great" or "Average" or "Classic" or what have you. And I'd assume even less people actually click "See our review criteria" before bitching about what seems like a low score.

Would an 0-5 score scale, sticking to (presumably sticking to 2.5 as the average) lessen the misunderstandings and decrease the complaints, do you think?

What do you see as the pros and cons of each scale?

Feel free to also express other ideas and why you think they'd possibly be good.

I used to wonder all the time why movies were on a 4 star system, but when you make it concise, I guess it helps some people draw a hard line between the cream and the curd.

Of course, if you don't have a long-running understanding with your readers, they might balk at a 4 or 5 star system. I.E. "Who is this guy, and what's an 'average' game for him?"

10 star systems help to do away with some of that, and I personally prefer a more detailed score. It's not going to be objective, by any means, but I feel that the more specific the score, the more specific the reviewer has to get to justify it.

I prefer a percentage scale myself. I feel that it is more precise that a /10 or /5 rating system, that is unless the /10 system allows for marks such as 8.2, then it is just as good as the percentage system.

I also feel that the reviewer should break down each element of the game (graphics, gameplay, sound, etc) and mark these each with a score of it's own, something like Gamespot do.

I also like the way that Gamepot sums up it's reviews with a information box at the start of the review that covers the best and worst points in a nutshell.

Thanks for the 2 posts so far.

Domino: we break ours down in a way you suggested, I think. Here's one of our reviews. Tell me if it's what you had in mind:
http://www.rpgland.com/games/marvelalliance/review.html

Side: our 1-10 scale goes in .5 incriments. When we just barely started doing reviews, we'd do all decimals, but it got tedius trying to figure what separated a 7.3 from a 7.4, so just simplified it a tid bit.

Yeah that's the sort of thing that I was on about. The review seems to cover all the important elements of a game, and has the summary points nicely laid out in a system that is clear and easy to understand. I like it.

Just a suggestion though. How about having the overall score at the top of the page next to the game title, or at least on the same row as the title, while keeping the summary of points down the bottom of the page.

The score at the top of the page will then give visitors to your site a quick indication of what you thought of the game, rather than having to scroll down to the bottom of the page just to see what score the game got.

That seems like a good system. To make it more immediately apparent that 5 is an "average" score how about putting a line right above the reivew #'s box that has a scale from 1-10 and lists 1 as painful, 5 as average, and 10 as the crowning ultimate achievement in gamingdom or some such thing... When I see a 1/10 scale on a gaming site I assume that few if any games ever get 5 or below, even if they are horrible.

I write reviews and I use a system which has a rating x/5 where x is a positive integer.

I've never liked percentage systems. Rating a 98% seems odd that it be 2% away from supposed perfection, but take a game call it A which scores 90%, I tend to find I in comparison to others, it's hard to put a finger on why it should score 40% more than that or 7% less than that one. Personally I can't use it. Five suits me fine.

As long as it isn't 5 star bullshit or anything with letters, it's good. I prefer percentage scale myself.

Right, but, like I said:
We covered that for the most part.

I use the 10 star scale doing reviews... It just gives you a little more to say as opposed to the 5 star scale.

Since when was 5 average? I always drew the line at 7... I wouldn't play alot of 5/10 games...

Gamespot Wii Reviews

Would you honestly play some of those games at that rating?

This was mentioned as a problem with the /10 system in the first post.

5 is average because it's halfway to 10. Or at least it should be, but for whatever reason, a lot of places now use 7 as average, like Gamespot. If 7 is average, then anything 5 and below is crap and interchangeable, and essentially your scale just goes from 5-10. And in that case, a /5 system is what it pretty much is.

Sorry for repeating what you said. :riiight:

I modeled my ratings from a PC games mag can't remember which:
5 - Classic, rare, near-perfect
4 - High quality, succeeds in most areas
3 - Flawed but successful development
2 - Suffering from problems
1 - All wrong

I find this seriously makes it easier for me.

I really like the /10 scale, especially when decimals are involved, but the point about 7 being average makes me think about that. Any game that is less than 7, I probably would not buy, or even try. However, I don't like the /5 scale since 3/5 is average, but you can't really tell if it's slightly above average (since 4 makes me think it's a good game, and 3.5 still gives the impression of average), or on the lower end of it. Breaking down by components is definitely a great way to go and do it, but overall I think I would go with the 10 scale still. Most games should never have to deviate from 7 much, and you can really justify the bad ones, however low their score maybe. In addition. good games don't seem to need as much explanation with descriptions as bad ones do.

Exactly, which is why I agreed with the original poster:

I can see what you mean. Games that score 5 I think I would keep and usually I would love to play it again. For 4 it's good that I would finish the game at least once. Three is where I enjoy it but maybe I get bored, consider it for trade-in, 2 is bad, 1 a disaster.

Exactly my thoughts if I ever see a 5 point scale used, although for personal enjoyment, I need those excess numbers as a point of comparison. Well. not like I let exact numbers sway my thoughts. I look for key gimmicks of a game that are mentioned and how I personally picture them being used. Surprisingly enough, one of my favourite things to do is to read bad reviews and see exactly how 1/10 is justified by the reviewer.

I'd go with the x/10 feature because there's a clear low point (1), middle point (5), and high point (10). It's also akin to a x/100 system, which is what I'd use, if your x/10 uses the tenth place. But it hardly matters as long as you have a convincing text body to back up your rating.

As scores go, it's all about first impression.

The problem with the x/5 scale is that there's not as much room for differentiation, IMO. I'm not that picky when it comes to games, but a 4/5 and a 5/5 are more or less the same to me. They both represent great games. Hell, even a 3/5 would stand as a decent game. Personally though, a rating of a 2 or lower registers as awful in my opinion.

With a x/10 system, I see anything above 8 as a damn good game. Anything from 5 up to 8 seems pretty decent at first glance. Anything below 4 would register as a bad game when I first see it.

I say do away with numbers completely and use language to convey the score. Numbers are far too ambiguous and easily misinterpreted. It's too difficult to understand how 7.8 is superior to 7.3 or an 85% over a 79% because numbers themselves just don't carry enough information.

Let good argumentation be the decider.

Over the past while I've been finding that I usually only play games that receive good reviews or are made by developers I like. Once I got my PS2 modded, I tried out a lot of games that I'd normally ignore, but found a few that I was genuinely impressed with, despite their flaws.

Since everyone values aspects of games differently, I figure the simplest review score would be out of three:

Buy it
Try it
Avoid it

I'm thinking of avoiding reviews entirely these days, and making my opinions on games based on previews and developer logs. It makes things more interesting.

Yeah I guess I'm not one to compare games against each other, five is enough to give me peace of mind. So far in my games log I don't think I have given a 1/5 score to any game, I haven't played too many bad games but I can't really vision how bad a game must be for it to score such a low mark.

The problem with using 5 (out of 10) as an average is that most sites don't. I like the gamerankings.com links on some sites, that give you the average score from all of the review sites. If one site gives it a 5 and another gives it a 7.5, and in spirit they mean the same thing, it's going to drag down the overall score for a game. Personally, I prefer a percentage.

Well, the danger a media outlet gets into there is that somewhere, someone's going to take that recommendation and regret it. Even a big fan of, say, the RPG genre is going to have at least a handful of games that get near-universal praise, but just aren't Example Gamer's cup of tea. The "buy it/try it/avoid it" is generally what one tries to describe in the text of the review.

Giving it an 8.5/10 is one thing, but telling someone to straight up spend their money on it is a somewhat different territory. :eagletear:

But even with percentages, you get people who review things with 50% as average, and other people who review things with 70% as average. Obviously, every source for reviews is going to have a different scale, and the best way to go about thing would be to learn the scale of the source you are viewing before trying to follow their advice. Or, actually read the article included with the number.

Instead of putting a numerical rating, why don't game review sites just use a descriptor for the rating and avoid the confusion of deciding on what number represents an average? Seems like that's the easiest solution that allows for the greatest amount of clarity.

So instead of rating something X/5, and declaring some arbitrary number to be "average" they can use a scale that says something like: Masterpiece, Excellent, Good, So-so, Bad, Awful, Complete Garbage.

I'm aware that rpgland uses that in their verdict section, but more people will notice the numerical value and ignore the verdict altogether, which makes me in favor of removing numerical scores altogether. People put too much of their decisions into the raw number as opposed to figuring out what the number actually means.

I am particularly fond of the 1-5 scale, with .5 numbers also being used. I don't like anything that has to do with partial numbers unless it's half, such as 9.2 or 3.8.