Places that properly gives out the correct review scores?

NotMSRP

Member
Using the 1-10 scale, looking for reviewers who know that 5 is the average and should be the most common score. Most review scores are positively biased, which are scores bunched towards the 10 side. Negatively biased is near the 1 side. Anyone who has taken statistics knows about the negatively biased and positively biased bell-shaped curves. So any place that gives out a normal distribution of scores?
 
sometimes EGM and for sure igncube.com (which actually takes into account outside Nintendo business decisions), but don't count on ignxbox.com or ignps2.com

edit: actually, Edge and GMR also come to mind as knowing a 5 is average, also, forget about GameInformer, which thinks a 7 is an average sometimes...
 
belgurdo said:

aren't they always fighting back about 5-6 scores for games that were average, that's what I remember, and they keep on reminding people that 5 is average, not 7...
 
I know when I write my reviews at GameDaily (using a 5 point scale), I have no problem giving an average game 2.5, and most of the other reviewers are the same way as well.
 
NotMSRP said:
1-10/Letter grade scale

10: A+
09: A
08: A-
07: B+
06: B
05: B-
04: C+
03: C
02: C-
01: D
00: F

How's this?

Good. It's the exact scale we use at the college newspaper (Florida State).
 
They also gave VJ a 5.5, which I don't agree with. But hey, it's videogames. If I ever agree with someone 100% I'll marry HIM or her.

Haven't played Sudeki yet, sorry if it was just righteous bashing.
 
aerodynamik said:
They also gave VJ a 5.5, which I don't agree with. But hey, it's videogames. If I ever agree with someone 100% I'll marry HIM or her.

Haven't played Sudeki yet, sorry if it was just righteous bashing.

Yeah, and without places with scores we can't comprehend (Edge), there wouldn't be a need for four page threads discussing them.
 
Just use the Conanza Interpretation Scale to figure out review scores.

For reference, when the score = x, the formula is...

x^2

So...

10 = 100
9 = 81
8 = 64
7 = 49
6 = 36
5 = 25
4 = 16
3 = 9
2 = 4
1 = 1

...and so on. If the score is not a number less than ten, or on a ten point scale, convert it before running it through the formula. It usually comes out about right with fanboys.
 
I don't know who does, but I can tell you who definitely doesn't: RPGFan.

Check this out:

Q: How should I rate the categories?

A: Rate them on a scale of 0-100% based on your opinion of them. The rating scale is as follows:

100% = Perfect. You should probably stay away from this rating since nothing is perfect.

99-90% = Great. This is the equivalent of an A grade. Reserved for those categories that you believe are so outstanding that they are nearly perfect in every regard.

89-80% = Above Average. This is the equivalent of a B grade. Given to those categories which are more than satisfactory, but not the best quality ever.

79-70% = Average. This is the equivalent of a C grade. Given to those categories that are merely average. This does not mean that the category is bad, rather, it means it is on par with the majority of RPGs out there, and doesn't do anything out of the ordinary to catch the eye, ear, imagination, etc.

69-60% = Below Average. This is the equivalent of a D grade. Categories that merit this rating are sub-par or below that of most other RPGs out there. It says that there is a flaw in the category which hampers it somewhat.

59-0% = Poor. This is the equivalent of an F grade. For a category to receive this grade, it must present itself in such a manner that it severely detracts from enjoyment of the game, or is a source of aggravation. Care should be taken when giving anything below a 60, for that is an indication that the category is shameful or reprehensible in some way. A more moderate rating would be a 69-60%

Fucking idiotic, if you ask me.
 
Matlock said:
Just use the Conanza Interpretation Scale to figure out review scores.

For reference, when the score = x, the formula is...

x^2

So...

10 = 100
9 = 81
8 = 64
7 = 49
6 = 36
5 = 25
4 = 16
3 = 9
2 = 4
1 = 1

...and so on. If the score is not a number less than ten, or on a ten point scale, convert it before running it through the formula. It usually comes out about right with fanboys.

Genius! G33kalicious! Thanks for the laughs.
 
I don't see why everyone loves 5 being average. It seem to me that every outlet that uses this scale basically has about 3/4ths or more of its reviews as being above average...which is simply not the case.

Take any issue or site's scores and look at the data and you will most likely find a curve that is extremely high.

A lot more games would get 5s if it was indeed average, which it never really is.
 
Hemmdog said:
I don't see why everyone loves 5 being average. It seem to me that every outlet that uses this scale basically has about 3/4ths or more of its reviews as being above average...which is simply not the case.

Take any issue or site's scores and look at the data and you will most likely find a curve that is extremely high.

A lot more games would get 5s if it was indeed average, which it never really is.
I think EGM does a pretty fine job of defining 5 as average. In fact, it's hard to argue otherwise.

GameSpy also has an interesting curve with its 5-star system, where 3 is "good" and 2 is "fair." It ensures that the majority of games fall around the middle of the curve, with a few ascending to the heady heights of 4/5 stars and a few deserving pieces getting trashed with 1. I'm going to go out on a limb and say that I think it's a pretty well thought-out vg ratings system, and not just 'cuz I work there. Unfortunately, some people / sites can't be bothered to understand that 3 stars >>>> 60%, so some changes are being made soon.
 
This is why I like the five-point scale we use. I initially fought it coming from using a 10-point scale and all the decimal points. But it didn't take me long to realize that it's a much clearer way to score games. Each number in the five-point scale is clearly defined, with a three being average and both points on either side used to denote scores that are above or below average. For me, it's the easiest scale to understand and you don't have to worry about what part of a 10-point scale really means a game is average.

I dunno. Can anyone really tell me the difference between a game that scores a 7.8 and another that gets an 8.0? A few more posts and I'm sure we'll see the obligatory "review scores don't mean anything because everyone has their own taste in games" post, which I definitely do not agree with. There are plenty of quantifiable, technical aspects of games that make them very easy to score with a reasonable consistency on a five-point scale. It's certainly more true of video games than of movies, which consist almost entirely of subjective material. Yet I don't think I've ever heard someone complain that movie reviews don't mean anything.
 
GS's scale is still positively biased and you see that in their distribution of scores; 7-8 being the most weighted section.
 
Agent Dormer said:
Yeah, and without places with scores we can't comprehend (Edge), there wouldn't be a need for four page threads discussing them.

I hear that reading the fucking reviews helps.

Well, sometimes.
 
It doesn't matter if a mag or site gives out a lot of 5's and 6's to "average" games. He's asking if there are any sites that will give the below average games less than a 5 and the above average games more than a 5. Unfortunately, this just doesn't happen. For a 5 to be an average game, it should also be the average score.
 
Matlock said:
Just use the Conanza Interpretation Scale to figure out review scores.

For reference, when the score = x, the formula is...

x^2
...snip...
Actually after much research, we found the perfect scale for us in France where grades are usually on a 20 scale :
N = exp ( X / Pi^(Pi - 0,077))

Yes, some people have waaaay too much time ;)
 
satterfield said:
This is why I like the five-point scale we use. I initially fought it coming from using a 10-point scale and all the decimal points. But it didn't take me long to realize that it's a much clearer way to score games. Each number in the five-point scale is clearly defined, with a three being average and both points on either side used to denote scores that are above or below average. For me, it's the easiest scale to understand and you don't have to worry about what part of a 10-point scale really means a game is average.
funny that you say that -- as soon as gamespy switched over to the five star scale, publishers and readers could not fucking begin to understand what the hell it meant. it seems so obvious to me, but it was like we were scoring games in a completely alien language. i still shudder to think about the mass reaction to it.

worth nothing that i think 2.5 (which doesn't exist on the scale) is the dead average. 3 is a good game, 2 is a bad game.

I dunno. Can anyone really tell me the difference between a game that scores a 7.8 and another that gets an 8.0?
not really. i like GMR because we use the 10 with no halves. it makes you really have to commit to the score. makes me really think about it. i find that the .5s let you kind of wimp out, but on the other hand they do allow you to lend some nuance to the score. it's a tough argument. anything broken down further than .5 is just masturbatory nonsense, though.
 
Dagon said:
I don't know who does, but I can tell you who definitely doesn't: RPGFan.

Check this out:



Fucking idiotic, if you ask me.

Uhh, that's the correct scale to use. NotMSRP's scale is absolutely terrible. A C is considered to be a pretty decent grade.

F = 0-49
D = 50 - 54
D+ = 55-59
C- = 60-64
C = 65-69
C+ = 70 - 74
B- = 75-79
B = 80-84
B+ = 85-89
A- = 90-94
A = 95 - 99
A+ = 100

That's pretty much the scale that critics go by. A 50 or a 5 is by no means "average". It's a failure.
 
Why shouldn't five out of ten be average? It is the average of all the possible scores, after all...

X-Play gives average games a 3 on a 1-5 scale, which I like.
 
IMO 1-100 or 1-10 scales are stupid. Too much variance. With 1-5 scales you typically see the most balanced scores.
 
As I've always said, a fruit-based scoring system is the only way to go for truly balanced reviews.

As for '5 should be the most common score', has anyone considered that this shouldn't be the case? We have quality control systems in place to ensure that truly bad games don't normally slip through, with the developers, publishers and console manufacturers all having the ability to reject or hold back the game if it isn't good enough. So with a ranking system that uses the full range of scores to represent the full range of game qualities, I'd still expect the average score of released games to be higher than 5. Most of the 1-3 games shouldn't even get released - these scores should be reserved for when they are (*cough*Godai*cough*).
 
Bog said:
It doesn't matter if a mag or site gives out a lot of 5's and 6's to "average" games. He's asking if there are any sites that will give the below average games less than a 5 and the above average games more than a 5. Unfortunately, this just doesn't happen. For a 5 to be an average game, it should also be the average score.

Not true.

I think the big problem lies in using the word 'average' to describe the quality of a game. When magazines say that a 5 is 'average', it's pretty obvious that they're trying to say it lies at the midpoint between awful (1) and superb (10), which would make it 'playable enough but thoroughly unexceptional'. Since most people consider things that are OK but unexceptional to be 'average', that's the word they attach to the rating(s) in the middle of the scale.

Unfortunately, some people (particularly the number-fixated folks who frequent GAF) hear 'average score' and immediately assume it's a numeric average, rather than the adjective that's intended. So you get these off-kilter assumptions that if a 5 is 'average', it should be the most common score handed out, or that such-and-such score is the 'new average' because it's the one that most games are receiving, etc. And it's all because of the baggage attached to that one word.

Maybe magazines and websites should stop using the word 'average' in conjunction with 5's, and switch to a word that won't cause so much confusion amongst the more literal-minded, like 'acceptable' or 'unremarkable'.
 
AlphaSnake said:
Uhh, that's the correct scale to use. NotMSRP's scale is absolutely terrible. A C is considered to be a pretty decent grade.

F = 0-49
D = 50 - 54
D+ = 55-59
C- = 60-64
C = 65-69
C+ = 70 - 74
B- = 75-79
B = 80-84
B+ = 85-89
A- = 90-94
A = 95 - 99
A+ = 100

That's pretty much the scale that critics go by. A 50 or a 5 is by no means "average". It's a failure.

That's a positively biased/skewed curve.
 
To me, it makes a lot of sense to have 5 be an average score. Why have 7 be "average"? Do we really need six (or seven) different numbers assigned to below-average games and only three for above-average games?

The main purpose of numerical ratings is to give people a quick idea about the quality of a game, right? Most people who read reviews won't be trying to decide between games with ratings of 2 and 3... they'll be looking for 8s, 9s, and 10s. Why not widen the high end of the scale to give people a better idea about games they'll actually want to buy? Besides, on a scale where 7 is "average", anything below a 5 is practically useless.

Not that any of it really matters, of course... but it's a lazy Saturday afternoon.
 
All of the bunching in the 6-9 range caused the reviewers to add a decimal point to the score; they needed to expand.
 
As said before, it really depends on what the benchmark is. If 1-10 is being used with respect to all other games as a comparison, then yes, 5 will be average, but if it's being used with respect to some scale of innate quality, then a passable game should have more than half the points needed. This is why school grades are marked only on the top end of the scale--no student should be getting the majority of their answers wrong.
 
It should be noted that the average score on a scale from 1-10 is 5.5... Which makes 5 slightly below average and 6 slightly above.
 
We're only considering 1, 2, 3, 4, 5, 6, 7, 8, and 9 since no game will likely receive a 10 or a zero. We need an odd number of terms so we can have a real middle score. As you can see, five is our middle man.

This gives us the low-tier block(1,2,3), middle-tier block(4,5,6), and the top-tier block(7,8,9), and each block having low, middle, top.

Guess we can attach a 0 and a 10 at each end if you really want them although they will be rarely used.
 
Top Bottom