In defense of the number: a note on video game review scores
- Thread starter StreamerDarkly
- Start date
Really depends on how a publication uses their scale, so always check. IGN is not Edge and Edge is not the college you attend.
If a mag or site says a 6 is a decent score, you should still pay attention to 'sixes'. I've handed out plenty of sixes to games that were flawed (or just very bland), but not to the degree where they wouldn't be fun for anymore. Atleast, I kinda enjoyed them and I doubt I'm that unique.
If a mag or site says a 6 is a decent score, you should still pay attention to 'sixes'. I've handed out plenty of sixes to games that were flawed (or just very bland), but not to the degree where they wouldn't be fun for anymore. Atleast, I kinda enjoyed them and I doubt I'm that unique.
The problem with marking 7 as average is that everything below 7 would amount to grading how shit something is.Scow2 said:
Which seems like a lot of wasted effort. And usually it is, since most ratings below 7 seems almost entirely arbitrary. The only thing I think a review should get across, clearly and concisely is this - Do I (The reviewer) recommend this game?
The whys and buts should then be covered in the text. If a game is better or worse than another game... Doesn't even matter. If it does, the number of recommendations or non-recommendations will indicate how many like or dislike it. Now how much, but just how many - Which is a reliable and consistent metric.
The more people that like a thing, the likelier that more other people will like it too. But even so, it does not tell you anything specific about the game or if the game will be good for any one specific player.
So yeah, I hold to my stance that the Steam model of reviews works best. That is - As much text as you want, and then a recommendation or non-recommendation. That the total number of those is tracked is good, for the reasons stated above. And also good is the fact that you can filter to lists of either category.
Then having other people up-vote detailed and informative reviews (This being the factor that you should look for in any given review), to make them more visible.
Then the best way to find out if the game is good for you or not is to read the top voted positive and negative reviews, apply your own personal tastes, and then make a decision.
If it sounds like I make a big deal out of this it's because I make an effort not to pay good money for games I personally won't like - Regardless of how many other people like them. I don't believe in wasting money on games you might not like, regardless of how critically acclaimed they are.
Like say, Far Cry 2 - Which reviewed wonderfully. But it's a piece of tedious crap to me.
Exhibit #1 is a deceptive one, since it only states the existence of complains. But it doesn't indicates the existence or lack (much less the percentage) of readers who accept the high ratings, because usually they don't give feed back. Why would they need to challenge the claims if the official rates already agree with them? And yet, when controversies happen, the "gamers entitlement" topic appears in some places.StreamerDarkly said:
Exhibit #2 those are gamers who write reviews. I'm talking not only about them, but also about those who just read reviews and play games (who, I think, encompass a higher percentage of the gaming community than those who bother in commenting or writing reviews themselves).
Out of topic, this brings me a question. What's the goal of someone reading reviews/scores? To find out the best games from the bunch? Or to find out how good a specific game is? In the first one, that person has no idea of which game to buy (or even which unplayed games exist right now). In the second, he already heard about the title before and probably made already an opinion based on whatever he heard.Mutant1988 said:
For me it's mostly just to find information about it on things that would make it unplayable for me specifically. I have a pretty good idea of what I like and if I see a game that does things I like, I check the reviews for any specifics that I do not like.CaitSeith said:
Rarely do I ever just look at one game at a time. What decides my purchases is primarily - What do I want to play right now? The quality of a title really only has any bearing on how much I would be willing to pay for it. But I can figure out the quality from the information written about the game far easier than I can from nondescript numbers.
No, it doesn't make sense. School is the only thing where 7/10 is average because you should fucking know over half the shit you're taught. Having average at 7/10 causes all the good games to be all bunched together making it hard to tell which ones are better. 8s, 9s, and 10s should be reserved for the best of the best, not just good games. That's exactly why no other medium uses 7/10 as average. Even IGN and GameSpot state that 5 is average on their scoring systems even though neither use 5 as average.Scow2 said:
Aggregating likes and dislikes is even worse that aggregating scores because then FFXIII would have a 98.7% rating vs an 83 aggregate score. How is that more helpful?Mutant1988 said:
As a rough estimate for how likely it is that any given person would like it.Phoenixmgs said:
This is assuming a whole lot of likes and dislikes. The more people that vote, the more accurate it is as a "rough" estimate for how likely it is that someone might or might not like it.
Whether they do or not requires more than just that (i.e. reading the reviews). But it's probably enough to make them give it a second look, if only because there's a lot of positive (Or negative) votes.
My point is that the actual amount of people voting is factored into determining this, rather than the sum of individual ratings that are skewered by how much any given reviewer like or dislike a game (And by the importance they put in specific elements) - Which I think is completely irrelevant in this context and not even something that need or should be quantified. Why? Because it's not fact.
If we are supposed to look at games critics as especially trustworthy we should do so for the quality of the written review, and for how well it conveys the opinion of the reviewer and the information about the game.
Not because a game is given a "proper" numerical rating.
Expecting or demanding that all games be judged by the same standards is effectively nothing but confirmation bias. You want others to agree with what you have already made up your mind about. That or be told that you should like something more than you do (Because it's rated high). Too much faith and too much importance is put into these damn numbers.
You either recommend it or you don't. Buts and ifs are covered in text.
As for your example - How would that be less accurate? If more people like the game, then more people like the game. That's the one and only purpose of that metric. That is how you determine the likelihood of something (Given enough samples, of course), by looking at the pattern of recorded fact.
Is it invalid because not everyone leaves feedback? Then I'm sorry, but we can just scrap democracy then because not everyone vote.
Whether you will like it or not won't be dictated by numbers anyway.
Note: I'm not against summaries. A text summary that quickly details why you would want it or why you really wouldn't is a great way to end a review and a perfect tool for skim readers to determine if the game is interesting or really not good for them specifically.
"I recommend this game for anyone that likes X or Y and don't mind Z."
Or if it's a bad game, I'll just say that "X or Y are good, but it's ultimately ruined by how bad Z is." If someone don't mind Z then they can still buy the game.
Details are important.
Here's what I find most amusing about debating the validity of Steam's rating system versus that of Metacritic.
Whilst dismissing Metacritic's user rating system as a war between trolls and fanboys who (dishonestly) vote at only 0/10 or 10/10, the PC Master Race crowd apparently fails to notice that this perfectly mirrors Steam's system. Only two possible ratings exist, and the game is reduced to a straight popularity contest of which side can accumulate more votes. From a technical standpoint, the Steam approval percentage becomes exactly the Metacritic UserScore multiplied by 10.
Pure democracy for the win, or an unearned sense of superiority by those who are sexually aroused at the mention of Gabe Newell?
And while I suppose you might argue the point, I'm not so sure there's a huge intelligence gap between the typical user review comment you find on either service.
Whilst dismissing Metacritic's user rating system as a war between trolls and fanboys who (dishonestly) vote at only 0/10 or 10/10, the PC Master Race crowd apparently fails to notice that this perfectly mirrors Steam's system. Only two possible ratings exist, and the game is reduced to a straight popularity contest of which side can accumulate more votes. From a technical standpoint, the Steam approval percentage becomes exactly the Metacritic UserScore multiplied by 10.
Pure democracy for the win, or an unearned sense of superiority by those who are sexually aroused at the mention of Gabe Newell?
And while I suppose you might argue the point, I'm not so sure there's a huge intelligence gap between the typical user review comment you find on either service.
I'm only talking about professional game reviews. I'm not exactly sure what you ideal system is here but it sure seems like you're more so talking about user reviews here. The problem is user reviews are more useless than professional reviews because quite a large percentage of user reviews are "best game ever" or "this game sucks". Your system of "buy it" or "don't buy it" is going to be really skewed as well. If you use your system for professional reviews, then almost every major game release is going to be well over 90% for "buy it". That's even less helpful than what we have now.Mutant1988 said:
Numbers still provide valuable information like seeing what games your favorite reviewer gave 9s and 10s to. Numbers can still be used for your system as that's exactly what RottenTomatoes does with numeric scores. Reviews are going be aggregated regardless of if there's numbers are not, and a number aggregate is more informative.
If there's 3 FPSs that came out, I have money for just one, and you say A and B worth buying but C isn't; then how am I supposed to know whether you liked A or B more? If you rated the games based on a number, I'd know which one you liked more.
Personally, I don't see what review scores really add.
I'm currently working on an in-depth review of the Mass Effect series, are you really going to tell me that I'm going to add anything at all to those three reviews by slapping an almost entirely arbitrary number at the end? Especially in the era of "7/10 = It sucked, 8/10 = It was okay, 9/10 = IGN."
The review score adds nothing to the review, it's practically dead weight to the review. I'm reminded of an old, old, old quote: "A good designer has done his job not when he has run out of things to add, but of things to remove."
I'm currently working on an in-depth review of the Mass Effect series, are you really going to tell me that I'm going to add anything at all to those three reviews by slapping an almost entirely arbitrary number at the end? Especially in the era of "7/10 = It sucked, 8/10 = It was okay, 9/10 = IGN."
The review score adds nothing to the review, it's practically dead weight to the review. I'm reminded of an old, old, old quote: "A good designer has done his job not when he has run out of things to add, but of things to remove."
From the text? It shouldn't really matter which game a reviewer liked more anyway. What's important is which game seems the most interesting to you.Phoenixmgs said:
As for ranking games, I'm perfectly fine with top 10 lists and such. If given a context, it's perfectly reasonable for one person to rank games in order of how much they like them.
Like in month X, I liked these games in this order.
It's the obligation to stick to a rating system, a quantification of overall quality, that bother me. Especially when you mix in different opinions. Aggregates of that skewer the results of any one review, based on how many vehemently dislike it or loved it without reservation - That while conveying no information about why.
So for a professional review, of the same reviewer - I would suggest just having a list at the beginning and end of the review list where they put the games in falling order of "like".
And I don't think user reviews are useless at all, when their purpose is simply to gauge how "popular" a game is. That is, how likely it is that anyone "might" like the game.
It's rather hard to determine which game a reviewer liked more just from text especially considering reviews aren't comparisons. A review may come off as more negative due to the reviewer being more passionate about a game so the flaws/negatives seem worse. Whereas a game that's just good all around but does nothing great will come off as more positive. Adding a number at the end fixes all of that.Mutant1988 said:
I don't see how putting a number at the end stops whatever you want reviews to be. A number is just another tool for both the reviewer and reader to use. Just because scores become aggregated doesn't make them bad.
Popularity is a horrible way to try to determine if you will like something. Most popular things aren't very good. Most of my favorite works from any medium aren't very popular (like my favorite TV shows usually get cancelled). Having broad appeal isn't inherently bad but the more popular something is, the less chance an individual will passionately like it. That's what I'm trying to find, not stuff that I'll probably enjoy an alright to decent amount, I want to find stuff that I'll love. I'd rather play a game like Majin and The Forsaken Kingdom that Jim Sterling loved vs some FPS that almost everyone liked. Chances are higher I'll probably dislike Majin vs that FPS everyone liked but chances are higher that Majin would be a "special" type of experience that I might love. That's what I try to find.