What do you think about the way game reviews are interpreted differently to other mediums by Metacritic?

I've always wondered why, if a movie ends up with, say, a 61 rating out of 100, its colour is green on Metacritic, suggesting a generally positive reception. It has to hit a score of 60/100 before it's coloured yellow, suggesting an average or mixed critical reception. This is the same for everything from music to TV, but not games. With games, 75 is the lowest green score possible. A 74/100 or lower is where a yellow colour is used, which is an average/mixed score by this metric, I guess? I thought we considered that a decent to good score in the gaming community, and I thought 61/100 for a movie was considered pretty middling to say the least.

Metacritic actually addresses it in their FAQ, which interested me:

WHY IS THE BREAKDOWN OF GREEN, YELLOW, AND RED SCORES DIFFERENT FOR GAMES?

The reason for this special treatment for games has to do with the games publications themselves. Virtually all of the publications we use as sources for game reviews (a) assign scores on a 0-100 scale (or equivalent) to their reviews, and (b) are very explicit about what those scores mean. And these publications are almost unanimous in indicating that scores below 50 indicate a negative review, while it usually takes a score in the upper 70s or higher to indicate that the game is unequivocally good. This is markedly different from movies, TV or music, where a score of, say, 3 stars out of 5 (which translates to a 60 out of 100 on our site) can still indicate that a movie is worth seeing or an album is worth buying. Thus, we had to adjust our color-coding for games to account for the different meaning of games scores compared to scores for music, movies and TV.

I use Metacritic as a place to have quick access to a lot of reviews at once, I'm not a review score whore, but I'm not sure if this makes a whole lot of sense to me. I was just wondering what you all think? Not a serious topic or issue or anything, I'm merely a curious person.

I think it sorta makes sense. A bad movie is 1½-4 hours of more or less passive digestion.

You have to actively process and participate in a bad game to see it to the end, and often for far longer than 4 hours.

While a game might, on aggregate, be above a certain threshold, that threshold demands much more of you than a bad movie, so the judgement of good/bad reflects that.

I think thats a pretty good reasoning to do it the way they do, even if some websites claim that 50 or 5/10 is average they certainly dont seem to use the bottom end of the scale unless a game is really shit.

I feel like grading on a scale of 1-5 gives you a better sense of what a game is like than the 7-10 most websites use.

Well, part of this probably has to do with how scores are traditionally divided. Giant Bomb is an outlier in that they only use a 5 point scale. So, a 3/5 on GB doesn't mean the same thing as a 6/10 on Gamespot. And that was after the big Gamespot redesign a few years back. It used to be a hundred point scale.

It was kinda a thing in the early days of the site where something would be trending on Metacritic at a 9.2 and the great 4/5 GB score that was just as positive on the game would "only" register as an 8.0, so it would bring down the average. It's what has always been the problem with Metacritic, where you can't just shove an 8 point scale into a hundred point scale and say it's the same thing. There are different expectations there.

But the vast majority of film reviews use a 4-5 point scale, sometimes with half points in between. So it makes sense that the average for a film reviewed on a 5 point scale would be lower than the average of a videogame that is reviewed with 10-20 point scales.

tldr; review scores are meaningless, and Metacritic numbers aren't arbiters of anything.

All review systems are arbitrary nonsense at the end of the day. It cant objectively be broken down into numbers.

As I have gotten older I realise that averaging all criteria just doesn't work. Hence why a 1-5 star or even 1-10 system based on the reviewer is best. Then just find reviewers who have tastes that align with yours.

While this video is largely about RottenTomatoes, its content is still relevant when discussing Metacritic and immediately came to mind when I saw this topic.

@zombiepie: That's the very aspect that was confusing me, the lack of consistency with the numbers/reviews themselves and the visual language of the top two aggregator sites. Not that I could think of ways to improve them, and not that I particularly think it matters, but to express how jarring it is to parse the info you want through this extra subjective glaze. It's definitely messy when that one movie or game straddles the line between those arbitrary markers of quality and passively deconstructs the entire concept.

Thanks for sharing that, it's just what I hoped to see here.

The couple of sentences in Metacritic's explanation after the "(b)" are basically the answer here. A while ago, all the major game sites started giving scores around 7.5/10 for games that many users considered marginally above average, and that even in the text of their articles, outlets would often suggest weren't anything special. Arguably, that's fair when most video games that are going to be widely reviewed are explicitly positive experiences, but the kind of reaction that would be summarised into a 5/10 on a film website would get expressed as 7.5/10 on a gaming website.

There have been attempts to fix that problem, but they're complicated by the standard that previous reviewers set. You could say that you're going to give the "pretty good" games 5/10 instead of 7/10 now, but because many readers have been trained to expect 7/10 to mean "fairly good", and anything below to mean "bad", adopting another scheme risks confusing users. So now we're stuck in this slight limbo where one site's 3/5 is another's 7.5/10. All a review aggregator like Metacritic can do is stir those scores into the same pot so the combined result is never going to be all that accurate. However, it still makes sense for them to consider a 7.5 to be the cutoff for "good" because that's the way most sites still do it.

I've noticed that their color coding is different for games, but never thought too much about. Very interesting discussion!

I think it makes sense. If you have a scale, you want the audience to actually see it in use. If they put the boundary at 61/100, it might as well not exist. For games released this year, you need to be at the 7th & last page to come across that score. The problem with putting the boundary at 75/100 is that signaling that a 74/100 game is not very good is hard to swallow. But hey, the reviews are right there. And metacritic's users by now generally understand why the color turns yellow at the point that it does. If Jason likes Godeater, he's not going to drop Godeater 3 for having a 74 score and play 75 point City of Brass instead. And with the reviews right there + youtube + streams, there's ample opportunity to inform yourself.

We're also in such a saturated market nowadays that outlets choose to review games that are noteworthy. That filters out a lot of the bad games already, so the average review score is naturally going to be quite high. I'm honestly not sure what the movie review industry is like in that regard. Are most movies getting reviews or are there quite some that appear on metacritic but sit without a score? I'm looking at games like Unturned & 100% Orange Juice sitting there pretty with no score attached to it. A cursory glance at the movies page does tell me that there are more negative reviews given to movies than to games.

The 5-to-10 scale in video game reviews has been pretty well documented, I think. @gamer_152 explains it pretty well.

I think most of that stems from the old way video games were reviewed, with partial scores given for graphics, sound and gameplay. A shitty videogame that didn't look or sound like total garbage would still get some points in those categories. Ie. it would score 2/3 for graphics, 2/3 for sound and 1/4 for gameplay, which is a pretty garbage score but still nets it a total of 5/10.

Movies and other media have always been reviewed for the total package. If it's a bad movie, nobody cares if the audio mixing is done well. Therefor, it is traditionally way easier for a movie to score 2/10 than it is for a videogame.

I'm still trying to peddle my idea for a 3-star review system:

3 stars: you should buy this

2 stars: you could buy this if a certain aspect specifically appeals to you

1 star: you shouldn't buy this unless you really know what you're doing

Everything else is bullshit, anyway. Consumer experiences are so subjective that having a 100-point scale is just nonsense. It might help ranking one game over another, but that's largely academic and not really relevant. Which is why I kinda like RottenTomatoes's system.

Doesn't matter too much as long as they're transparent about it (and since you can see the raw number anyway...meh)

I've already considered aggregate scores completely worthless and basically nonsense, for reasons already mentioned here.

The only thing a site like Metacritic is good for is gathering all those reviews in one place to easily look at each individual one. It always bums me out when I hear that publishers still treat those aggregate scores as a definitive indication of a game's success and worth.

Yeah, I go for it for a consensus. Also, as a reviewer for a smaller site, we cover a lot of games that wouldn't even reach metacritic in the first place. We use a 10 point scale but I rarely dip into the lower 5. There's a quality to games in general that the industry has grasped.

I do hate how 7 being categorised as 'mixed' has come about. It's broadly positive. 6 is mixed. 5 is bang average and so on.

I feel like if i grind away at a mediocre game for 30+ hours i'm going to be more annoyed than a 2 hour movie.
That's the only reason i can come up with.

Metacritic is considering the source. Good for them. That’s some really basic statistics work, but better to do it than not.

Also, any common rating scales beyond a 0-4 or 1-5 system are inherently garbage. People aren’t great with details.

All a 100 point scale does is help two dudes fight over fractions of a point on the Internet, for some reason. No reasonable person says “oh, this averaged 87.8, I guess I should watch it before the 84.6 film.”

It mirrors how the public sees games. I think andromeda was rated at 70 right? It was lauded as the worst game of all time damn near lol.

@oursin_360: I agree, even if some games straddle the line in an awkward way. I looked up every game with a 70 rating before making the thread, and the only one I would fight for being shaded green instead of yellow (worth playing instead of not so much, I mean) is the PS2 version of Killer 7.

From 71-74, however, I saw a bunch of games that most consider worth playing at least or even outright good, like Radiata Stories and Binary Domain. Their scores are accurate, but they're still considered good I thought. Exceptions aren't that damning, though; this was more a case of me finding the disparity interesting where games have this 25-point window (75-100) to be considered absolutely worth playing by Metacritic's colour-coding system, whereas other mediums have a 39-point window (61-100). Even if that is the result of two very different scoring systems, it interests me, probably more than it should, ha.

99% of the time I think their system is successful, I just wondered how other people viewed it, and the replies have been great so far.

That's because people can't math...just kidding, sort of.

A lot of it probably has to do with hold over from another era where 7 was consider average because they simple borrow what a school grading scale tends to look like and never gave it much thought. And created a system where there's a least 6 degrees of badness and 3 degrees of goodness so an "average" score is usually inflated past the mid point of a scale.

Then there's the determination of the minium and maxium score a scoring system, where the real average score is at, and how it scales up and down. GB ranges from min 1 to max 5, with 3 as average I believe. This is a numerically sound system. There are equal steps above and below the average.

But it gets lost in translation as you scale up and down. In a 10 point scale it is tempted to put a GB 3/5 as 6/10 as average...but that's only true if the 10 point scale has a min of 2 and max of 10. If a 0/10 and a 1/10 game should exist in a 10 point scale it has no real GB equivalence while you can say that a 0 and a 1 out of 10 game is probably pointless and not worth mentioning, it does create a discrepancy.

A whole number only scoring system of 1 to 10 no "real" average you have 5 steps of below average from 1 to 5 and 5 steps of above average from 6 to 10. While you can certainly work with such a system if you want each review will have to lean on one side or another. A 0 to 10 system allows 5 as the true average. There must be an odd number of steps in a scale to correctly represent an average score.