Despite the tongue-twister title of this thread, it actually has a serious topic: what is the best way to use reviews of a game to decide whether or not you should buy it?

I've been doing a survey of the user reviews of some of the games I've purchased here on GoG.com, paying especially close attention to the games I found disappointing and the games I found to be much better than expected. My goal is to figure out what to look for in a user review, and what to ignore, in order to get an accurate picture of a particular game. Here are my preliminary ideas.

1) Throw out all 5-star reviews, and all 1-star reviews.

Now, this is not perfect. There are lots of games out there that deserve 5 stars, and a few that deserve 1 star. However, there are simply far too many 5-star reviews, because it seems to be a natural human tendency to overrate things. Try to discern which 5-star reviews are accurate when 85% of them for any particular game are 5-star. It's impossible. On the other hand, 1-star reviews, although fewer in number, tend to be the most thoughtless reviews of all. If you spend some time reading 1-star reviews, you'll see things like, "I started it up once and it didn't work. One star." That's not helpful at all. Many of them are written by individuals who had technical problems of one kind or another, but unless it's a technical problem that affects everyone across the board, it doesn't say much about the game itself. Finally, based on my short survey, 4-star and 2-star reviews both tend to be more thoughtful and accurate than 5- or 1-star reviews.

2) Try to differentiate between kinds or praise and kinds of criticism.

What I mean is, if a positive review starts with, "I played this game all the time when I was 12!" that's not a very sound basis for an opinion. It doesn't really tell you what the game will be like for an adult. Similarly, you might see a negative review of a game that complains about the graphics, even if the game was made in 1996. That's also not very helpful. On the other hand, if you see a positive review that says, "This game has been on my hard drive for 15 years, and I still play it," that's probably a good sign. On the whole, judging the different kinds of praise and criticism is more subjective than my first suggestion, so I recommend figuring it out for yourself. But you should do it.

Hopefully others have more ideas to contribute. Thanks.

I never read the user reviews at all, precisely because of the problems you mention. If I want a review, I try to find video footage of gameplay. That tends to be the best indicator for me when it comes to old (pre-1998) games.

I like looking at what reviews tell me about the game, bad or good; if some review sounds like a shill pitch in disguise, I will disregard it regardless of what score it gives to the game. Also, I think watching gameplay videos is more important than reading reviews, but I seriously dislike youtubers who prattle on without focusing on gameplay, which is why I prefer to watch normal people play than some journalist.

Point 2 is important. Digging out what specifically made a review get its score and how that relates to how you tend to feel about games is going to have a large impact on if you should weigh a review or not. People giving 1 star because they couldn't run a game are generally worth ignoring unless there is an overall chronic theme of bugginess. I generally discard anyone's opinion who gives extremely critical marks but can't give any real specifics about it, or is reacting negatively to a subjective point. For example giving 1 star for something being too hard, which someone else might find appealing.

I personally watch out for repeated complains on a game's controls, or glitchy mechanics that take survival out of the hands of the player. Negative marks are best when they appeal to something with as little subjectivity as possible. When it comes to selling a game that stuff only helps sell it, but in the end someone needs to draw comparisons to similar items so that the reader can have some sort of frame of reference for what the experience will be like.

Commenting on the graphics is usually unhelpful unless there is something that cant be seen in screenshots and videos.

And as a general rule for life, anyone trying to sell their point with a box full of exclamation points instead of actual points, should be ignored.

This is how I take user reviews. So far it's worked out well for me. ^^

On IMDB, I ditch the 10's and 0's... the fanatics and the haters. After that, the layout generally gives a good idea about the movie.

I wish GOG had a way to show how many votes of each amount, and maybe differentiate a bit more by showing "owned the game at time of vote" vs "did not own the game at time of vote".

But generally, if I actually look at user reviews, I just see if there's some sort of trend in the reviews.

I always do my research by reading reviews before purchasing a game and in nearly all cases, I read opinions from different sites to collate the information I need.

So, for instance in order of importance to me:

- Has to be a genre that I like

- The games strengths and weaknesses (for instance I can cope with the odd glitches if it's gameplay is highly rated)

- Does the game cater to my ability.....which means it has to be relatively easy as I tend to have a low tolerance of difficult or frustrating games

So, a review specifically tailored for me would have a few lines for each bullet point with a quick mention of graphics, voice acting.

The problem with this on a wider scale, is that each person will have their own perspective on what makes a good game.

Finally, I have strayed and impulsively purchased a title that doesn't fit my simplistic requirements. However, in the vast majority of cases, I have uninstalled these prematurely....so my system does work....I just don't always pay attention.

That's been unreliable for me. Consistently high review scores have contributed to purchases I found disappointing. That's why I'm trying to refine the process.

In a perfect world I would not use reviews to influence my decision whether or not to buy, but the fact is they do influence me.

Well, it depends on the trend you find and whether you place any value on it.

I generally start with a positive view of the game, giving all games the benefit of the doubt. I look at the graphics, maybe some video of game play. If it reaches this point I may or may not check out user reviews.

When I do, I look for negative trends that mean something to me and ignore the stars completely,, although, at least on Amazon, that generally means starting with the 1 star reviews. I don't check so much on GOG. Rarely do I even look at the user reviews here because I wish they were divided better here.

Here are my thoughts about certain "trends"
Graphics suck - ignore
Technical problems - ignore unless it's a big trend involving setups similar to mine.
Difficulty handling the character, vehicle, etc - these might have some weight.
Nonsensical solutions to puzzles - these can be turnoffs in adventure games
These are what come to mind at the moment, the first two you mentioned.

If there is a pervading feeling that the game sucks, I tend to listen, but not always. I have some very crappy games in my collection and I'm not really ashamed of that fact.

There is no formula to personal taste. You win some, you lose some.

They're best with mustard .

I agree for the most part with adambiser, but there's one more thing I look at.
Taking two games that have been listed in the catalog for about the same amount of time, how MANY votes and reviews are there?
If one has 3000 and the other 102, that 3000 game is probably a pretty good bet.

On GOG I first look to see if it is on sale. If I am paying full price then it better be a game I already know I am going to love. That means watching youtube vids, or reading Let's Plays, or having already played it a few years before, or whatever. If, however, the game is on sale, well that is entirely different.

First I look at the overall star count, 4-5 means seriously look at it, 3 means it may be misunderstood, 1-2 means it's probably not something I want. Next step, I read the most helpful rated reviews. And I do mean read them, I don't care what rating they give, I care what they actually say. For instance one of the games I enjoy is Sniper Elite: Berlin 1945, I picked it up during the Insomnia sale. The star count puts it at 3.5, the screenshots look decent enough, the reviews though are generally quite abrasive towards it. One of the critiques it receives again and again is the fact that it allows a third-person perspective instead of forcing first-person; this just so happens to be one of the things I really like in shooters. I also look over at the GOGMixes and try and get a general feel for recurring themes (there's a difference between awesome amazing games to play before you die, or play it for the plot, or avoid if at all possible). If I am on the fence I go poke around the GOG game specific forum, looking for things like bugs or just general impressions. If I'm not in the mood for a frustratingly hard game it is a good warning to see something like 10 different threads saying "stuck on impossible boss #7." Finally, if I still can't decide I simply ask in whichever sale thread is currently on for impressions. The GOG community has a pretty amazing communal wealth of knowledge, and the chances are high that someone will respond back comparing the game in question to other games that I may have actually played, which is very helpful.

I don't think you can really make use of ratings. There's a distinct lack of data from the people that would have an interesting opinion. I can't recall ever having rated a game here, and I doubt I'm the only one.

With a self selecting sample, it's incredibly hard to turn that into something that's representative of what the game is like.

This is true but extremes do usually show themselves. An extremely well regarded game like Baldur's Gate is going to have tons of reviews and five stars. Whereas a loathed game like Master of Orion 3 is only going to have a couple.

Of course the number of stars doesn't tell you what the game is actually like. It just gives you a rough idea of what the popular opionion is.