We've had some internal discussions about whether to change the current game comparisons feature to just be "plain" links to "similar games". In practice, this would mean that we strip away all the +-= stuff and simply retain a picture and the game title.

There are two reasons for this:

First, we would like to make this more attractive for contributors to actually use. Currently, many game entries are being created as "dead ends", with no links to further game entries.

Second, there is even an editorial inconsistency right now. The comparisons concept comes from the time when there was still just one "editorial" review per game and it was clear whose subjective view the comparison text represented. Now, with some games having several reviews, it's not clear; in some cases, what the comparison texts say is vastly inconsistent with some of the reviews, and there is no attribution who wrote the comparisons.

So, what do you all think? Vote, and if possible, also say why!

And third, it would also provide better rendering for touchscreen display.

P.S. Here is the thread discussing the introduction: /forum/topic.php?id=303.

In case the verbal description of the plan is too abstract, have a look at the latest game entry.

This is for illustration purposes only.

Please don't make the discussion about styling. I don't want to discuss the "how", but rather the "if"!

In concept, I really like the expanded information about how the similar game is the same and also different.

In practice, I've never added an entry like this to any of the games or reviews that I've posted so far. I think this is for two main reasons. First, I want to get the game and the review posted to the gallery and this always tends to come at me as an after-thought so I skip it. Second, the similar games that I know about are usually not in the database themselves so I'm never sure about the protocol for creating a "stub" just so that the entry I'm creating can link to it.

I'm not sure that simplifying in this way is going to actually encourage more use of the feature. It does make it easier since you don't have to enumerate the differences but you still have to find or think of what games are similar which requires some effort. This also doesn't address my second issue where I don't know what to do when the other game doesn't have an existing entry.

I definitely see the point about the editorial inconsistency since this can certainly be colored by personal opinion and the current form doesn't show who that opinion belongs to. I would say that eliminating the details though doesn't completely solve this issue though since exactly what games are similar and those that are not is still a matter of opinion to some extent and so still inherently editorial.

At face value, I like the idea of attributing the comparisons to the contributors just like we do for the reviews but I think that could get messy; a simple set of links is more tidy than a break down by contributor. Another idea would be to have a similar links section and then have a comparisons section with more detail by contributor but that also adds complexity.

I think I lean toward keeping them because it's potentially useful information and I'm not sure that removing them really resolves the problems.

Thanks for this balanced feedback.

The "mode" to put comparisons to games without entries is simple: you just do it. As for example here: /game/id%2C198/Walls+of+Rome.html. Cohort is not on the site (yet). Nevertheless, I put it into the comparisons. The site recognizes that this cannot be resolved and just displays the text without any link. It also automatically puts it on the requests page. Once the game is added, the comparison will automatically resolve and the "request" will be gone.

Entering stubs just because a game has been referenced somewhere, however, is not an option (unless one of the regular stub rules applies).

Is it possible to move the game title completely outside the frame either above or below? I think that would make it look much better.



Most of the time you don't have to find anything, you simply recall a handful of similar games, that is, if you have played more than a couple in your life. Of course, i can't account for your experience, but i personally do tend to have at least a few references in mind and i'm sure other people do just as well. And even if you can't remember anything, simply leave it to others, problem solved. Just like comparisons aren't mandatory to fill, you either do it or not.



As a matter of fact, we did start off by considering this as a third solution, however we ended up having two, because, as you have noted yourself, it makes things much too complex.

Ehm, problem not solved. I thought we were discussing how to make filling this out more attractive, rather than accepting the "fate" that half of the game entries are dead ends.

Well, as expected, feedback is rather low. We've had a second visitor voting "no", but without saying why. Which is fair enough, but hard to interpret how important it is to him or her.

The thing is: if we go for "no", i.e. keep things as they are, we currently have no alternative solution (attempts) other than appealing to people to please, please fill this part of the submission out – which hasn't really worked for the last 13 years, either.

So, even though we have a slim majority against the change, the absolute number of votes is so low that I tend towards accepting the policy change as apparently, the couple of thousand people who visited since the poll started and did not vote do not care. I.e. they seem to have nothing against it.

I'd say, few people visiting the site really care about forum activity, so i doubt they even saw the post. So yeah, wise decision right there.

Personally I am a fan of the comparisons with all their plusses and minusses. I think I added them to each of my reviews and at times even to games I inserted into the database for other people. It is just that I love comparing games or at least weighting them against each other, because for one it gives them context and for two it gives their flaws and strengths a bit of perspective. E.g. It is easy to say Sierra games are the best ever… until you compare them to LucasArts titles and see that while their graphics, scenarios and music are slightly better, LucasArts hit the nail a whole lot better on the head as far as gameplay and humor is concerned. And while you could put those thoughts in the reviews, I guess it is just bad style to directly compare one game against another in a text that should be dedicated to one of them. Plus: The comparisons often tend to be more tongue in cheek, which makes them a lot more charming.

Still, I have to admit, that there is just a small minority which actually makes decent use of them, and yes it does feel like an extra effort, which might not be needed. Browsing through the reviews you get to see a lot of missing comparisons, which gives the feature a kind of broken feeling. I do not like throwing away the effort, but I have to agree that for consistency's sake it might be a better idea to remove them. And maybe a new system might move some reviewers to make use of the similar games section more often. Therefore: Cautious yes on my part.

I know the how is not really the issue right now, but I would suggest putting at least a short description of the game on the links. Just the name and a (at times missing) screenshot is lacking something. Linking missing "similar games" directly to the request page might also be worth a thought.

Hi Herr M.! Good to see you again and so great that you decided to take part in this!



True, i've had my share of fun too



I think just in case comparisons should be saved for possible future utilization. So the effort won't be in vain. I suggest we keep them.



What about curiosity factor? Also, don't forget, descriptions much like comparisons can be subjective, inadequate and simply misleading to people whose priorities are manifold... so i'm not sure how useful that can be. Not that i strongly object to it, but it's a remnant of comparisons which, in my opinion, should be relinquished along with comparisons.

I would never "throw it all away" in the sense of deleting the data we already accumulated. I would simply not display it anymore. Like Herr M., I believe displaying some "comparisons" (with full text) and some "similar games" (with no/little text) is not a good idea. We should go one way or the other as far as the visible part of the site is concerned.

First one thing. in the current state it seems that every comparison is stored in each game. Otherwise dead ends would not be possible. For me it would be more logical to have an new entry for that or no entry at all but therefor later more.
At first we have the game entry and an comparison entry.
GAME(name)
COMPARSION(game:ref.1,game:ref.2,difference)
Thats is nearly like the currently existing system. Now something you could do is having a page showing two random games and an visitor must write down what is the difference between two games or what they have common. For example Freddy Barkas is an Adventure and Outlaws an shooter but both have an western setting. Until every comparison is done.
Adding this into our model gives:
GAME(Freddy Barkas).1
GAME(Outlaws).2
COMPARSION(1,2,both are western games, but 1 is adventure and 1 is shooter).3
Now adding for Example adding Wolfenstein 3D to this list
GAME(Freddy Barkas).1
GAME(Outlaws).2
COMPARSION(1,2,both are western games but 1 is adventure and second is shooter).3
GAME(Wolfenstein).4
COMPARSION(1,3,they have nothing common).5
COMPARSION(2,3,both are shooters but 1 has western setting and 2 has Nazi setting).6
But this gives (n*(n-1)/2) entries. One thing what could be done is to not store entry that have nothing in common, but better seems to be to not having comparison entrys at all. That means not to have no comparison simply move the comparison data into the game data.
GAME(name,setting,type)
Then with the following data:
GAME(Freddy Barkas,western,adventure).1
GAME(Outlaws,western,shooter).2
GAME(Wolfenstein,nazi,shooter).4
You could generate the original comparesion entrys. So in short manual comparsion is not needed when game classification is improved but can be used to improve classification.

Heh. Well, I guess there is no choice now. The whole topic actually started based on prototype code I made to perform automatic similarity matching based on the meta tags we have (themes, genres etc.). This works very well for games with a strong definition, but not so great for games with few tags.

I.e. we concluded it still needs to be done manually unless we go back and systematically "reclassify" everything. Which was where I was reluctant, because of the experiences of similar projects where (slightly exaggerated) in the end it was me alone doing about 95%. Second thing is that the guiding goal is really to make the extensive game submission process easier rather than harder (which would be the case with more refined tagging).

This is not to say it's impossible. It's just that this route, although amazing in theory, is undesirable workload-wise (or: perceived to be).

Hmmm… I am not so sure if only the name of the game makes most people all that curious. The description would not necessarily need to be that long (on the contrary the shorter the better). At least there should be some kind of background graphic even for missing links, so that they do not look that glaringly empty. But I guess those are details, first we have to agree on whether to drop comparisons or not.


I would suggest putting them either in the review archive or open up an exhibition for it. Put I guess it depends on how hard it would be to transfer them there.

Indeed, the "emptiness" is an issue, as illustrated here: /game/id%2C1149/Cohort%3A+Fighting+for+Rome.html. A default placeholder image could be one option. Alternatively, what does everybody think about actually requiring a short description text if the game is not on the site yet (which would then automatically disappear once the game is there)?

A short description for missing games would also help identifying them later on, because at times just the name can be ambiguous (e.g. remakes).

Eine möglichst breite Auswahl an vergleichbaren Spiele ist wichtiger,
als eine direkte Orientierungshilfe mit den "+", "-" und "=" Attributen
anzubieten. Eine reichhaltige Auswahl bringt dem Leser meiner Meinung
nach mehr, als eine "verstümmelte" Beurteilung über ein paar kurze
Stichpunkte.
Damit nicht so viele "dead Ends" entstehen, sollte man es darüber hinaus
dem Ersteller der Rezension auch einfach machen. Denkbar wären
beispielsweise auf Tags basierende Vorschläge mit aus denen per
Mausklick manuell ausgewählt werden kann.

Anmerkung:
Sorry for writing in German, but my Englisch is only peripheral.