This is a split board - You can return to the Split List for other boards.

It seems like game reviews should be distributed in a normal distribution, no?

#1BendoHendoPosted 8/15/2013 9:05:00 PM
Serious thread, skip to the TL;DR if you have attention span problems.

Or at least something similar? Approximately a mean score of 5.0/10 and normally distributed?

But they aren't.

By that, I mean the average game it seems should be around a 5.0/10, and most games (say 68%) of games fall between 3.0/10 and 7.0/10, and around 28% are either between 1.0 and 3.0 OR 7.0 and 9.0, and the final %4 falling between EITHER 0.0/10 and 1.0/10, OR 9.0/10 and 10.0/10.

I know I didn't articulate that well, but basically it seems like game reviews should follow this distribution

http://upload.wikimedia.org/wikipedia/commons/8/8c/Standard_deviation_diagram.svg

But most game reviewers do not use the lower numbers very often and far, far too frequently give out 8, 9's and 10's.

In fact, given sufficient hype, a game is almost guaranteed to score at least a 7/10 even if there are significant number of complaints about its quality.

For instance, Mass Effect 3, which granted I have not played, was universally criticized by gamers upon release....yet magically has a 92% rating on gamerankings. This isn't a thread about ME 3, but the general problem of not-so-outstanding games receiving outstanding scores all too frequently.

It seems like games should be reviewed something similar to this where:

1 - horrendous
2 - terrible
3 - bad
4- below average
5 - exactly average
6 - above average
7 - good
8 - great
9 - excellent
10 - outstanding

Given this, we'd expect most games to fall between 3/10 and 7/10....but they just don't. They are skewed towards higher, inflated scores.

TL; DR Why aren't game review scores distributed as we would expect - normally distributed with mean of 5.0.....and instead distributed with a strong skew towards high scores?


--
#2DoramiPosted 8/15/2013 9:07:57 PM
It's a normal distribution. The mean is 7.5 like the grading system in the U.S.

Not sure what's so hard to understand.
---
"Why do ppl say Steam/PC has DRM? The only thing steam does is prevent piracy, and there's nothing wrong with that." - Oil_Rope_Bombs
#3GabrezuPosted 8/15/2013 9:09:00 PM
The problem with the current situation is that a 7 is considered as the average and the inflated scores are just stuck there. Maybe it would be better and more intuitive to have 5 as the average, but like the American length measurement system, changing it would cost too much.
#4BendoHendo(Topic Creator)Posted 8/15/2013 9:12:43 PM
Dorami posted...
It's a normal distribution. The mean is 7.5 like the grading system in the U.S.

Not sure what's so hard to understand.


But that proves my point. Why is 7.5/10 average, when intuitively approximately 5.0/10 should be average? If 7.5 means the game is goodish-great, that is saying the average game is good/great....but that's not true.

--
#5DoramiPosted 8/15/2013 9:29:45 PM
That doesn't prove your point. A normal distribution is over an infinite set. On a finite interval, whether the data distribution fits a normal distribution depends on how well the data points would fit on any normal distribution curve, not the one that is specifically centered on the center of the data interval.
---
"Why do ppl say Steam/PC has DRM? The only thing steam does is prevent piracy, and there's nothing wrong with that." - Oil_Rope_Bombs
#6Babyfur_RigbyPosted 8/15/2013 9:36:30 PM
rating systems like these x/10 things are lame ass hell in general because what a 9/10 represents to one person will almost never mean roughly the same thing to someone else, which sorta defeats the whole purpose imo but w/e
---
http://i.imgur.com/JBesOj0.png
#7Devil_wings00Posted 8/15/2013 9:37:26 PM
Boiling down a game review to a score is arbitrary and pointless. Taking a reviews score without actually reading the review is absolutely no indication of anything. Every reviewers numbers mean something different then every other reviewer and vary greatly between game to game. They all hold differing opinions so it's absolutely impossible to apply a generic scale across all reviewers. It would make the scores even more pointless then they already are.

One game can get a 7 despite being a fricken awesome game but it had one major flaw that held it back or an ending that sucked. Another game can be an above average game in all areas but standing out in none getting it a 7. Another game can get a 7 and be an awesome game but it was plagued with technical issues that bring it's score down. All games got a 7 but without taking each individual review's content into account that number really means squat because WHY it got a 7 is far more important then the number and without the WHY you, at best, get a pretty loose generalization of what a website/reviewer thought.

Honestly reviewers just toss numbers at things because it gets them web traffic. If people actually started reading their reviews instead of wanting some arbitrary number to fight over maybe we wouldn't be stuck in a world where meta critic somehow has power over game development houses and sales projections, even though it's complete BS.

That's never going to happen because people are obsessed with the number itself instead of why the number was used.
---
3570k @ 1.260v - 4.6Ghz, GTX 580 @ 980/2106, ASUS Sabertooth z77, 8GB G.Skill Sniper 1600MHZ, Intel 335 SSD (boot).
#8YoungAdultLinkPosted 8/15/2013 10:01:09 PM
Because game reviewers tend to review games that their company thinks they'll like. There's good reason for this, too. Look up IGN's review of God Hand, on the PS2. They gave a beatemup to someone who doesn't like beatemups, and the wonderfully difficult game was given a 3 out of 10 and a poorly written review that is missing basic grammar and spelling.
---
Once we perfect building Mechs, I expect there to be a factory in Mexico pumping out a brand call Mechsican. I wanna buy me a Mechsican.
steamid: Rohchya
#9AskeladdPosted 8/15/2013 10:03:31 PM
Game review scores? Who cares?
---
TWSS hes speken end the wenner es E. There cen enle be ene.
"If it can fit in a Pokeball it can fit in a Gatorade bottle. I'mma catch it." NobleRoar
#10GreenMage7Posted 8/15/2013 10:04:56 PM
BendoHendo posted...
Dorami posted...
It's a normal distribution. The mean is 7.5 like the grading system in the U.S.

Not sure what's so hard to understand.


But that proves my point. Why is 7.5/10 average, when intuitively approximately 5.0/10 should be average? If 7.5 means the game is goodish-great, that is saying the average game is good/great....but that's not true.

--


He told you: it's like the grading system in the US.

100%-90% = A
89%-80% = B
79%-70% = C
69%-60% = D
59%-Below = F

People have been using that system all their lives. I don't know what is suddenly wrong with it now. If you can't do more than 60% of things correctly in your video game, it is a failure.

The only thing that is off is that my school had a slightly different grading scale where 100%-93% was an A, and the rest curved like that.

Switching to a bell curve system at this point would be counter-intuitive, since all previous reviews use the current system.