We’ve all been there, casually browsing a forum with nary a care in the world. The sun may be shining but you’re safe in a darkened room absorbing top-tier video game chat when boom – some little scrote called ‘XXb0ngsm0kA98XX’ throws a hand grenade into the room. “Bloodborne is overrated emo trash, like all of Miyazaki’s games.”
And so your day, which had began so well, is now consumed by thoughts of vengeance.
Yep, one of the things we fleshbags just can’t get our head around is the fact that opinions are like bumholes – everyone’s got one. Rarely are they so pronounced or hotly contested as in the video game sphere, however, where shadow armies with silly usernames battle each other with interminable 3000-word screeds about the genius of Kingdom Hearts.
One of the strangest things about the industry is that, in terms of perception, this makes reviews seem very important – for a day or two, at least. It’s not just about whether the latest expensively promoted game gets 9s and 10s from the usual suspects, but the feeding frenzy when a high-profile title doesn’t deliver – and the raging arguments and prevailing counter-narratives when critics just can’t agree.
Anyway. We’re sitting here with a beautiful big dataset from Metacritic, and decided it was time to see what games divide critics the most – as well as the games they agree on. First, a note on the numbers.
How We Did It
Our dataset is taken from Metacritic, the review aggregator that collects game reviews. We downloaded data on over 350,000 reviews covering over 18,000 different games for analysis (yes, it took ages to do). Obviously, given that there are a number of important caveats, you should take this data with as many pinches of salt as necessary: for example, the data will only be as accurate as Metacritic’s database when we took our snapshot. Also, the list is limited to games with 30 or more reviews on Metacritic to stop any individual reviews skewing it, and to filter out any weird or obscure games that didn’t receive normal releases.
One other factor is that Metacritic ‘weights’ scores, but we haven’t – we’ve taken all the individual scores and worked out a pure mean. This is why you may notice our means differ from those of the source.
Most of all, please remember that this is for fun. It’s interesting! The results are points of information and curiosity that might spark good conversations – they’re not a truth-cudgel with which to batter your foes. Unless those foes have slagged off Bloodborne, in which case let ‘em have it soldier.
What The Numbers Say
We’re going to look at the most controversial games overall, then we’re going to break down the data by platform. One note: I know I said ‘all time’ but what that really means is ‘over the time where there’s been a huge games media and Metacritic has existed.’ But Most Controversial Games Since 2002 just didn’t have the same ring to it.
Without further ado, the most controversial video game on the whole of Metacritic is…
Well that was unexpected. When I first saw the name of this I thought it was some platformer about a sentient spring. In fact it’s a pretty cheesy dating sim on DS where, apparently, “everyone scores.” Apart from the game itself #rekt #belter
Game Revolution gave it a 0 while frothing at the mouth. “No skill is required to play this game!” As any fan of Eurovision knows, no skill means NUL POINTS. But then DS Central called it a “fantastic mess” and awarded it 80. That’s two of 30 critic reviews – no, I’m not going to go into this detail on every entry.
One might guess that Sprung’s controversial status has rather more to do with being an unusual style of game in an era (2004) where critics and fans alike could be a little sniffy about ‘non-traditional’ games. But that’s speculation.
The numbers show that Sprung has a mean (average) score of 47.83, and the standard deviation from this is a whopping 20.16. This tells us most of the scores are between 27.67 and 67.99 – standard deviation is a measure of consensus, so the higher the number, the less consensus there is over a given game.
2. Kung-Fu Live! (PS3)
Now you see why we’re gonna break it down by platform soon enough. Kung-Fu Live is almost certainly here because it was a PlayStation Eye game that, not to put too fine a point on it, didn’t control all that well. If you ask me it’s a disaster but the fact it’s here means that at least some people liked it!
Kung-Fu Live’s mean score is 51.73, but again the standard deviation is huge – 20.01. I guess it all came down to whether you thought it was working or not. Respected outlet Da GameboyZ say that Kung-Fu Live allows you to “actually be in the game,” which is nice.
3. Lux-Pain (DS)
I feel a bit bad for this one. Lux-Pain is a 2008 visual novel, and that’s probably why it’s here. Visual novels aren’t really my thing, so I just don’t bother with them, but there was an unfortunate period when the genre was first becoming popular and would be reviewed by strict gaming criteria – e.g. the lack of meaningful interaction was held against it. Or hey, Lux-Pain might just suck.
Anyway… Lux-Pain has a mean score of 49.97 and a standard deviation of 19.20. The outlet Modojo said you should “bury it in the dirt with all those worms,” which seems a little harsh.
Here are the rest of the top 20 ‘most controversial’ video games across all-formats, and after this we’ll move onto the platforms.
Subsumed is a walking simulator, while Entwined is a rhythm-action game – the former genre has been divisive among critics and players, while weirder examples of the latter often come down to taste. The real surprise here is perhaps Ubisoft’s Assassin’s Creed: Unity, though this could be explained by the game launching in a rather embarrassing state (it was fixed post-release, to be fair, but that’s not going to be reflected in reviews.)
It’s all getting a bit tense at Redmond! Lococycle was a launch title that had been ‘upgraded’ from an XBLA Xbox 360 original, which may explain why some were so disappointed with it. It’s worth noting also that Yooka-Laylee’s mean is remarkably high (74.51) for such a poor game (CONTROVERSIAL!) AssCreed Unity’s face pops up again, minus any skin textures of course.
Yowch! We may have found 1-2 Switch pretty great in the right situation, but Nintendo’s launch software definitely wasn’t for everyone. No surprise to see ports of older games appearing there, which often divide critics, but more interesting to see both Xenoblade Chronicles 2 and ARMS on the list.
It’s also notable that the standard deviations here are all quite low, relatively speaking, which suggests Switch’s software has less score controversy overall than that on PS4 / Xbox One – but this has to be seen in the context of it being on the market for much less time.
The sexy old PS3! My man. We knew Kung-Fu Live would be up there, but surely the take-home message here is the point where critics finally got sick of David Cage. Beyond: Two Souls in there at number 7, with a mean of 71.06 but a seriously chunky standard deviation of 16.92.
God love you, 360, the days when a titan roamed the Earth. And look, it’s our old friend Sonic the Hedgehog! That may not surprise anyone, and that goes for this list really: Tony Hawk Ride was an abomination, Steel Battalion turned the onetime Capcom luxury dream into a Kinect game (enough said) and Superman Returns… well hey, at least the flying was alright. But spare a thought for the excellent Fusion Frenzy 2, sadly underrated and hanging in there.
The OG motherfucker! The little clamshell that could! Given the sheer size of the DS’s software library you’d expect a few crackers here and sure enough, as well as the top two covered elsewhere, there’s SEGA’s disastrous Super Monkey Ball: Touch & Roll, as well as rubbish like Asphalt: Urban GT. But hold! Electroplankton? Surely not. It’s painful to see a thing of such beauty traduced, though at least the mean of 72.10 is decent.
Ah, the Peace Walker and Monster Hunter Freedom Unite machine. How I miss the claw, by which I mean thank god I never have to put myself through that again *strokes Monster Hunter World and DualShock 4*
Nice to see Dead or Alive: Paradise where it belongs, though I have a soft spot for Dead Head Fred (as it appears do some critics). That at least has the highly respectable mean of 76.03, while the top spot is taken by Rengoku – and with a mean of 42.6 and a standard deviation of 17.54, critics were only really divided over just how bad it was.
Hold me close and rub my nipples, you sexy purple box. Let me use the wireless wavebird and lift you by the handle. Let me… Killer7 at number 1? Hahahaha nice – what an overrated game. A good mean of 73.91 rubs up against a whacking great standard deviation of 17.20, so folk generally quite liked it (I love the aesthetic) but opinions otherwise varied wildly.
I weep a solitary tear for Donkey Konga 2 and Chibi-Robo! I cry buckets at P.N. 03’s appearance, Mikami’s solitary misfire – though of course, that’s why it should be up there.
The big daddy. The real McCoy. The Playstation 2. Biggest selling console of all time, absolutely amazing software library, and Leisure Suit Larry: Magna Cum Laude stinking up the place.
That’s not the only offender, of course, with the disaster that was Tomb Raider: Angel of Darkness managing both a shockingly low mean of 52.89 alongside a huge standard deviation of 18.04. Good to see Killer7 in there again as well, get rekt son.
The fat man, the brave new entry, and where it all began and ended for Tao Feng: Fist of the Lotus. An interesting one here is Enter the Matrix, which really is hard to pin down and a great example of a controversial game – it’s awful game in many respects, but also had an ambition to it and ideas that so few movie licenses have subsequently approached. 66.58 isn’t a terrible mean, though standard deviation of 17.19 tells the story.
Well lookie here, if it isn’t the one category that readers consistently complain about us omitting, usually on a story about comparing consoles. Yes, it is the turn of the Personal Computer. I feel a bit bad about this one because I might have reviewed X3 in a past life and I may possibly have hated it and given it a bit of a kicking. So basically I am the most influential man in PC journalism.
The Beginner’s Guide is up there, you need to have a brain to appreciate this one so no wonder lots of people don’t dig it so much. Otherwise it’s relatively straightforward – garbage like Leisure Suit Larry and Duke Nukem alongside tired gags like Goat Simulator.
People really do love X3. I hear it’s great. I’m sorry.
Ah, if ever a console was misbegotten yet beautiful all the same… come over here, Wii U, no need to cry. I will always love you.
But I certainly don’t love Devil’s Third. Some people do, however, which is why it sits with a mean of 46.22 and a standard deviation of 16.89. Star Fox Zero split folk down the middle on those who found it inspired and those who found it uncontrollable, though a mean of 70.04 is still respectable. Sad to see NES Remix so controversial also: a smarter way to enjoy nostalgia, I thought.
If ever there was a system synonymous with shovelware, unfairly or not, it was the Wii – and critics knew not what to make of the stripped-down version of Michael Jackson: The Experience. More amazing is that anyone gave Tournament of Legends or Castlevania Legends decent reviews.
The sad one on this list is undoubtedly Wii Music, a personal project of Shigeru Miyamoto’s that was beautifully realized in many respects – but almost too unusual to find success, with many players bouncing off it. It’s one of the system’s most controversial games, but here that indicates a good thing.
Another beautiful machine that deserved better, RIP in pieces Vita. Instantly I’m thinking, why is Ninja Gaiden Sigma 2 Plus on there? Which fool gave that peerless gem of a game a bad review? By god I’ll hunt them down and… oh, it’s got some performance issues apparently. Well there you go: mean of 65.59, standard deviation of 12.74, and richly deserved.
Criminal Girls: Invite Only is apparently a cross between an RPG and elements of BDSM. I thought the latter was what turn-based battles were all about anyway? Low mean (54.61) and big standard deviation (16.92) indicates it may be a bit of a stinker.
Very little surprising in this 3DS list, beyond the fact that a lot of folk apparently enjoyed Sonic: Lost World. A little odd to see the excellent Crimson Shroud in there, which has a very nice average of 75.75 but a sizeable standard deviation of 14.42. Probably because it’s got dice in it.
Oh go on then, the pocket rocket, the king of hard-to-see screens, the duke of dinky cartridges, give it up for Game Boy Advance! What’s interesting here is just how many big-hitters make it into the list – no less than four Nintendo games, including two mainline Pokemon entries.
The Most Consensual Games
I’ll tell you one thing about consensuality. It’s HOT. Given the above reams of data, I know you’re probably flagging by now. We’re not going to split it out, let’s just do this: the 10 games, across all platforms, where critical opinions align.
It’s-a me! And who’s surprised? Galaxy 1 and 2 probably are the best games of all time, and the stats are incredible. Galaxy has a mean score of 97.42 and a tiny standard deviation of 2.98, meaning most critics thought it was within 3 points of what is already a pretty cosmic number.
To be absolutely clear here, this is a measure of consensus, and not an indication that Super Mario Galaxy is ‘objectively’ a 97.42 on the megawott scale. This data shows that most people agree it should be around there, and few think otherwise.
Lovely to see Resident Evil 4 with a sexy mean of 96.23, and standard deviation of 3.67. My only disappointment with this whole experiment is that Bloodborne isn’t up there somewhere. Maybe XXb0ngsm0kA98XX had a point, but now we’ll never know.
This is a simple point, but it’s worth drawing out. Both the online games media and Metacritic itself are fairly new. What this means is that we think the further back you go, the less reliable the data is – there were fewer outlets, to name but one factor, and many of those that crop up don’t seem to be around anymore. So the further back we go in terms of platforms, the more likely that we’re not getting a full picture. But hey, it’s better than nothing!
I’d hesitate to draw firm conclusions from this data without drilling-down a little more, but am happy to offer a few speculations on what this surface-level look at controversy shows.
The most notable aspect, as we get closer to the present day in terms of platforms, is the increasing presence of certain genres – such as the so-called ‘walking simulator’ – which outside of any one game have been part of a wider argument. To me this would suggest such games are victims of their own status as pioneers (though this is arguably no longer the case in 2018). The same argument could be applied to stuff like visual novels and anything that’s a little unusual or outside the envelope – it was a real shock to me to see that Electroplankton was one of the most controversial games on DS. It’s weird, yeah, but it’s also super-classy and beautifully thought out. But clearly that kind of thing just didn’t fly with some people.
It was also interesting to see both Assassin’s Creed Unity and Battlefront II being quite controversial – in both cases the games were the subject of much (justified) negative coverage while reviews were appearing. We can wonder about a correlation there, but of course it’s impossible to measure.
Finally it strikes me that the games which gained the most consensus among critics were also heavily marketed. I don’t think any conclusions should be drawn from this, and we have to acknowledge that something like Super Mario Galaxy might just be a brilliant game that appeals to everyone. Well, it is. But it is also meet, in an age of overwhelming information, to remind ourselves that consensus can often be a social phenomenon, as well as any kind of reflection of the arguments.
James O’Malley (@Psythor) helped us crunch all this data.