5 star rating scale > 10 scale

This topic is locked from further discussion.

#1 Posted by drekula2 (1872 posts) -

Two issues I have with the 10 scale.

10 implies a difficult to attain gold standard. I look at the 10's of this gen like Super Mario Galaxy 2. Yes, an amazing game that I loved a lot, but I never felt it (or the other games I would consider a 10) were in a league above most of the 9's.

Also, a 5-scale is better for when the reviewer has a mixed opinion. Let's be honest. A 3/5 for an AAA game isn't nearly as repugnant as a 6/10. Because AAA games deserve at least a 7 or an 8. I'd rather a reviewer give a good (but heavily flawed) game a 3/5 than go with the Western grade-inflation of the 10 scale.

#2 Edited by kingoflife9 (1987 posts) -

Like the old gamepro reviews.....but then a 4/5 is basically an 8/10.....so what do we give games that are genuinely a 9/10?

#3 Posted by drekula2 (1872 posts) -

okay. lets say you give a game 9 out of 10. what if the game is genuinely a 8.6 or a 9.5, what do you give it.

to me, 5 stars is simple

1 = bad (not recommended at all) - why do we need 4 spots on the 10 scale to say the same thing?

2 = mediocre (recommended in exceptional circumstances)

3 = good (recommended for the fans or on sale)

4 = great (recommended)

5 = excellent (Strongly recommended)

#4 Posted by lostrib (30942 posts) -

what? 3/5=6/10

#5 Posted by kingoflife9 (1987 posts) -

@drekula2 said:

okay. lets say you give a game 9 out of 10. what if the game is genuinely a 8.6 or a 9.5, what do you give it.

to me, 5 stars is simple

1 = bad (not recommended at all) - why do we need 4 spots on the 10 scale to say the same thing?

2 = mediocre (recommended in exceptional circumstances)

3 = good (recommended for the fans or on sale)

4 = great (recommended)

5 = excellent (Strongly recommended)

10 scale is more accurate, but I see what you're saying. 5 scale will put a lot of 8s in the same category as 9s though.

#6 Edited by drekula2 (1872 posts) -

@lostrib said:

what? 3/5=6/10

the 5 scale is not simple division. because the 10 scale is inflated. for example, we have 4 numbers (1-4) that mean the exact same thing "Do not buy this game".

A 7 point scale would be cool tho, but BLARGH WE NEED EVEN NUMBERS!!!

A 3/5 is like a 6 or 7. Sure, 6 and 7 do strike a stark divide on a 10 scale. But on a simpler scale, they both mean they mean they're worth buying but there's a lot better options out there.

#7 Posted by psymon100 (6138 posts) -

I like percentages, they're easy to understand, and a large percentage of people understand them. Sometimes a game is an 82% game -not an 80% game. The 5 point scale is like the Xbox One - it has insufficient resolution.

#8 Posted by drekula2 (1872 posts) -

I like percentages, they're easy to understand, and a large percentage of people understand them. Sometimes a game is an 82% game -not an 80% game. The 5 point scale is like the Xbox One - it has insufficient resolution.

but whats the difference between 82% and 80%. LOLGN

#9 Edited by psymon100 (6138 posts) -

@drekula2 said:

@psymon100 said:

I like percentages, they're easy to understand, and a large percentage of people understand them. Sometimes a game is an 82% game -not an 80% game. The 5 point scale is like the Xbox One - it has insufficient resolution.

but whats the difference between 82% and 80%. LOLGN

To attempt to answer your question .... first a disclaimer. This is all about subjective reviews, so I already accept that I've lost and you can just poo poo what I say. Besides, but surely differences cam be as subtle as 2%? Surely they can be more sensitive than just 20%?

Think of maybe sequels, or similar games, sometimes a game can be a fraction of a percent better or worse. Like, if I had to rank this gen's 3 Forzas now, I suppose I'd rank them 88, 90, 92% in order of release. I don't think it's be fair to say 4/5, 4/5, 4/5 because I can't discriminate numerically, to state which one is best. Or a game could fall a fraction short of perfection, like Haze on PS3, 98%.

Not a fan of percentages? By the way interesting thread.

#10 Posted by drekula2 (1872 posts) -

i agree there are subtle differences and games could be rated on either system.

i just prefer the 5 point scale because of the cultural assocations of the 10 scale.

"7 is average ; not 5. 10 is perfection or a league above all 9's, AAA's deserve 8+, and anything below 8 shouldn't be played"

#11 Posted by getyeryayasout (6799 posts) -

I like a 5 point scale, but Hustler's limp to erect scale makes the most sense to me, and I would like to see it become the industry standard.

#12 Posted by Gxgear (10425 posts) -

If you're using the 5-point scale then mentally converting to a 10-point scale, then you've completely missed the point.

Stay away - Poor - Okay - Good - Great

#13 Posted by drekula2 (1872 posts) -

@Gxgear said:

If you're using the 5-point scale then mentally converting to a 10-point scale, then you've completely missed the point.

Stay away - Poor - Okay - Good - Great

something like that. though i'd moreso go for recommendations

HELL NO! -- No thanks. -- Yeah, but... -- Yes! -- HELL YEA!!!

#14 Edited by Gxgear (10425 posts) -

@psymon100 said:

To attempt to answer your question .... first a disclaimer. This is all about subjective reviews, so I already accept that I've lost and you can just poo poo what I say. Besides, but surely differences cam be as subtle as 2%? Surely they can be more sensitive than just 20%?

Think of maybe sequels, or similar games, sometimes a game can be a fraction of a percent better or worse. Like, if I had to rank this gen's 3 Forzas now, I suppose I'd rank them 88, 90, 92% in order of release. I don't think it's be fair to say 4/5, 4/5, 4/5 because I can't discriminate numerically, to state which one is best. Or a game could fall a fraction short of perfection, like Haze on PS3, 98%.

Not a fan of percentages? By the way interesting thread.

Why should a game be judged based upon its predecessor? A good game is a good game is a good game, regardless of its prequel(s). If a series is guilty of being stagnant in its evolution, then it's the reviewers' job to reflect upon that within the review itself. A newcomer to a game series is not going to recognize the significance of a score difference if you don't give it context in the text. The series itself, as a whole, should be reprimanded in its own light on a separate occasion (original journalistic content instead of copy-paste news, imagine that right?).

Percentages doesn't work simply due to how it's so intertwined with the school grading system. There's so much connotation implied within the ranges of numbers that you end up only using a portion of the scale that the public finds "acceptable" (ie. anything above 60%). This is the exact same problem that we see with a 10 point scale and clarifies nothing.

A 5-point scale is simple and elegant: the entire scale has practical applications, every rating is meaningful and distinct from one another, and with personal preference/bias or disagreement with review you can shift 1 rating either way and it'd still make perfect sense.

For example: if a superhero title managed to swing an 'okay' (3/5) because of its licensed content, you can personally knock it down one rating and it's now something that you'd pass on or only as a bargain bin pickup; on the other hand, if you are a pretty avid fan of said superhero license, maybe you'd bump it up 1 rating and it's now something you see yourself buying.

#15 Posted by Netret0120 (1852 posts) -

i prefer 10 point scale

#16 Posted by psymon100 (6138 posts) -

@Gxgear said:

Why should a game be judged based upon its predecessor? A good game is a good game is a good game, regardless of its prequel(s). If a series is guilty of being stagnant in its evolution, then it's the reviewers' job to reflect upon that within the review itself. A newcomer to a game series is not going to recognize the significance of a score difference if you don't give it context in the text. The series itself, as a whole, should be reprimanded in its own light on a separate occasion (original journalistic content instead of copy-paste news, imagine that right?).

Percentages doesn't work simply due to how it's so intertwined with the school grading system. There's so much connotation implied within the ranges of numbers that you end up only using a portion of the scale that the public finds "acceptable" (ie. anything above 60%). This is the exact same problem that we see with a 10 point scale and clarifies nothing.

A 5-point scale is simple and elegant: the entire scale has practical applications, every rating is meaningful and distinct from one another, and with personal preference/bias or disagreement with review you can shift 1 rating either way and it'd still make perfect sense.

For example, if a superhero title managed to swing an 'okay' (3/5) because of its licensed content, you can personally knock it down one rating and it's now something that you'd pass on or only as a bargain bin pickup; on the other hand, if you are a pretty avid fan of said superhero license, maybe you'd bump it up 1 rating and it's now something you see yourself buying.

"Why should a game be judged based upon its predecessor?" - Why should it? - agreed. I was just giving an example of where a 2% difference might be relevant. Apologies if I mislead.

" A good game is a good game is a good game, regardless of its prequel(s)." - Agreed, each game stands alone. Don't know how I could have inferred different? Also a good game can have a bad 'metascore', or similar. Ironically I place little importance on published scores.

"Percentages doesn't work simply due to how it's so intertwined with the school grading system" - an opinion you're entitled to share, but I reject it. I'd say - preposterous, percentages are ubiquitous. And I never think about school. But I accept it's possible for some people, what you say rings true.

But your other point regarding a tendency to stick to a certain part of the curve, I agree with that also, but for different reasons. My reason is because before releasing the game they QA it for the level of the curve they expect it to fall on, and delay and improve until QA is satisfied. Thus shifting game quality towards the higher end. How many 0-20% (1/5) games are there vs 80-100% (5/5)?

"A 5-point scale is simple and elegant..." - agreed. My point was that the scale lacks in resolution, and this doesn't change that. Games are complex and I could probably handle a 500 or even 1000 point scale when discerning them. Is this practical? of course not.I feel like 100 is a good number, there is no science to this just my opinion.

Yeah, look that's all great and I can agree with most of that. But despite the accessibility of the 5 point scale I stand by my opinion in post 7, that I find percentages better, and they are widely utilised, and sometimes when discerning titles there can be little differences and some people ... they might find this helpful to know numerically at a glance.

#17 Posted by DraugenCP (8469 posts) -

I prefer a well-written review and I don't really care what the rating scale is. I wouldn't have a problem with reviews not having scores at all.

#18 Edited by Gxgear (10425 posts) -

@psymon100 said:

...

...

"Percentages doesn't work simply due to how it's so intertwined with the school grading system" - an opinion you're entitled to share, but I reject it. I'd say - preposterous, percentages are ubiquitous. And I never think about school. But I accept it's possible for some people, what you say rings true.

But your other point regarding a tendency to stick to a certain part of the curve, I agree with that also, but for different reasons. My reason is because before releasing the game they QA it for the level of the curve they expect it to fall on, and delay and improve until QA is satisfied. Thus shifting game quality towards the higher end. How many 0-20% (1/5) games are there vs 80-100% (5/5)?

...

Yeah, look that's all great and I can agree with most of that. But despite the accessibility of the 5 point scale I stand by my opinion in post 7, that I find percentages better, and they are widely utilised, and sometimes when discerning titles there can be little differences and some people ... they might find this helpful to know numerically at a glance.

This is an engaging conversation we're having.

Methinks your reasoning of QA doing their jobs properly is more farfetched than mine - a cultural perception when dealing with percentages. Plenty of games have skated by with plenty of game-breaking bugs, non-functional features, or other technical issues when reviewed (Skyrim, D3, SimCity, Pokemon X/Y, to name a few). The scoring trend itself is evident enough, I don't think there's any arguing that, and it's fairly consistent with the grading theory. The fact that you pointed out how games seems to be scored on a curve is pretty revealing; other than academic excellence there's almost nothing that's comparable, and certainly no other creative works is evaluated like this as far as I know. I look forward to a better explanation(s?) as to why games are always rated using a limited portion of any rating scale.

Really, in some sense, people perceives the larger denominations scales into a simplified one. From what's seen in the forums, people won't touch anything below a 50%, 60-70% is ridiculed except for those with strong feelings for the games in question, 80-90% is what most have come to expect as a minimum to warrant purchases, and anything above that are must-buy's.

At the end of the day, I'm not saying percentages isn't viable, it just faces a lot of the same problems large scales also face. You've made a case for it being better than a 10-point ( based solely on the fact that it's more universal), but I don't think you've established that it's better than a 5-point.

#19 Edited by psymon100 (6138 posts) -

@Gxgear said:

This is an engaging conversation we're having.

Methinks your reasoning of QA doing their jobs properly is more farfetched than mine - a cultural perception when dealing with percentages. Plenty of games have skated by with plenty of game-breaking bugs, non-functional features, or other technical issues when reviewed (Skyrim, D3, SimCity, Pokemon X/Y, to name a few). The scoring trend itself is evident enough, I don't think there's any arguing that, and it's fairly consistent with the grading theory. The fact that you pointed out how games seems to be scored on a curve is pretty revealing; other than academic excellence there's almost nothing that's comparable, and certainly no creative works is evaluated like this as far as I know. I look forward to a better explanation(s?) as to why games are always rated using a limited portion of any rating scale.

Really, in some sense, people perceives the larger denominations scales into a simplified one. From what's seen in the forums, people won't touch anything below a 50%, 60-70% is ridiculed except for those with strong feelings for the games in question, 80-90% is what most have come to expect to warrant purchases, and anything above that are must-buy's.

At the end of the day, I'm not saying percentages isn't viable, it just faces a lot of the same problems large scales also face. You've made a case for it being better than a 10-point ( based solely on the fact that it's more universal), but I don't think you've established that it's better than a 5-point.

Yeah This is fun. I'm giving it 110%

I don't intend to establish percentages as better, I just like them the best for the reasons I gave.

When I mean QA, I mean the QA throughout, not just the technical term for bug hunting. Sorry this probably stuffs things up. Essentially I mean the overall total level of polish of game they're going for, this might mean doing a high quality job of everything, the devil is in the details - every step of the way through development through to release. It might mean going back and doing things twice.

But a good point about bugs specifically, and Skyrim. Well, what is Skyrim's score with it's bugs? I notice on Metacritic the PS3 version scores 4% lower than the 360 version. Considering it's the same game, could those 4% be the average influence of bugs? a 1/25th penalty. With the 1/5 system, what would Skyrim 360 vs Skyrim PS3 receive? I think a cow would suggest they'd get the same, and with 1/25th I'd mathematically agree with them. What about you, man?

I don't necessarily have a better explanation, but I can take influence from your comments on academic scoring and maybe provide a different explanation and include them.

Assuming the grades aren't predetermined by 'grading on a curve', I'd suggest students don't aim for D's, they probably want A's, or C's, I'm not sure many aim specifically for B's but obviously people accomplish them. Obviously motivation for a grade is just one thing, intelligence and other things would factor in also. This should skew the mean 'up', which is very similar to video games, yes.

But yeah, I have two different explanations for similar patterns in different things. Perhaps the similar thing is trying to achieve a high score? Perhaps it is the critic who introduces the similar scoring? I don't know. But yes I can agree the skew towards the high end is there, but that doesn't mean there aren't low scores ...

Quite interesting, on metacritic for the PC games range from 29-96% So 35% untilised, almost 2/5.