With the way things are going at Giant Bomb, I don’t have a good feeling that I will always have access to their data. I’ve been talking with the folks over at IGDB, and I think I’m ready to make the switch. They’re very excited to bring us on board. They have a dedicated team whose only job is to curate their data. I believe they handle the taxonomy of games much better than Giant Bomb currently does. Instead of “releases” being put under the base game, there’s different categories a game entry can be with a pointer to a parent game. For example, Final Fantasy 3 (SNES) is considered a Port of Final Fantasy VI (Super Famicom, Wii, PlayStation). Something like the Jak & Daxter Collection is marked as a Bundle, and then the games included in the bundle have a pointer to the Bundle entry.
I’m sure there are examples where what they are doing will not be perfect, but no system is. It’s much better than anything I probably have the capacity to do right now. The fact that they have a team dedicated to entering and keeping track of this data is a huge draw for me.
There are several things I want to discuss with everyone and how we should handle things. I’m also going to need help mapping games in our database to games in their database. This might be something I’ll be offering to pay a few people to try and help out. They have quite a few Giant Bomb IDs in their database that will help us map things, but there’s quite a few missing.
The first question I’m encountering is how to handle ratings calculations, particularly for the Top 250 games.
For a situation like Final Fantasy VI, the North American version, Final Fantasy III, is entered as a Port of Final Fantasy VI. Each entry will now have its own separate page, but there will be links to go back and forth between the two. This is great for people who want to differentiate which version of a game they have on their shelves! When someone rates Final Fantasy III, should its rating be rolled up into the Final Fantasy VI main entry page? Or should they just be considered two different games? I ask the same question about episodic games. The Telltale games from 10 or so years ago will have a main entry page with each episode also having entries with the appropriate links. Should the episodes just roll up into the main entry?
This transition is definitely going to take a while, but I think it’s going to make Grouvee so much better in the long term!
Gosh, I think an argument could be made either way.
On one hand, rolling the individual reviews up to the main review might achieve a “consensus” review that I could see being helpful. On the other hand, not all releases are created equally, and I could see a bad port tanking an otherwise solid game’s reviews, or points of confusion if a game marketed as a port is actually very distinct. (Thinking of 8-bit Sonic games, Aladdin on Genesis/SNES, the weird differences between Rayman 2, etc.)
Putting on my developer hat, the questions I have are:
Which option is more difficult to implement or maintain?
Is it completely either/or, or could the overall “top games” ranking use its own logic for factoring in releases?
Is there an option that makes it harder to change your mind later based on actual user behavior?
I feel the same way. I have to think about this for a bit because this is a fundamental change to the way games are organized and creates more distance entries and I’m not sure which approach to aggregating common data between ports.
I’ll kind of answer these together. I was thinking I’d be using separate logic for the “top games” ranking, and just leave the individual entries as their own thing. That way you could see the different port ratings individually. It is a tough question about what do you do if one particular port of a game completely drags down the overall game’s rating?
It would make the logic more difficult in calculating the top games score, but not really all that much more.
The way I store ratings and game entries makes calculating the top games thing pretty easy to change, so I think we’re OK there.
Hypothetically, could treating them as separate games result in both Final Fantasy III and Final Fantasy VI appearing in the Top 250? If so, my personal preference would be for them to be treated as the same game for rating purposes, simply because I’m not a fan of different versions of the same game appearing as separate entries in the existing Top 250 – e.g. at the time of writing, The Legend of Zelda: The Wind Waker is #25 in the list, and The Legend of Zelda: The Wind Waker HD is at #26. IMO, this clutters up the list with (sorta) duplicates.
As I said, just my personal preference. Thinking beyond that, I agree with @tylerisrandom that an argument could be made either way.
This is just going to be kind of a free for all for everyone posting questions and thoughts, but another topic that needs to be handled will DLC.
The way DLC is entered in the Giant Bomb database sucks! Each platform that a DLC is on got its own entry, and it clutters the database up. That’s going to take a lot of cleanup. I’ll have to look to see how many DLC entries are on people’s shelves. If a DLC isn’t on a shelf, I could probably just blow it away out of the database and it’ll be replaced with what’s in the IGDB database.
P.S. Happy to hear about the possibility of the switch. I only learnt of IGDB recently (as a result of its being the apparent source of GOG Galaxy’s data), and it seems to be very comprehensive. I’ve been contributing to Giantbomb for years – just as a means of getting data over to Grouvee – and am not the biggest fan of their moderation and approach.
I would like to avoid this too. I don’t think remasters should be eligible for the list personally. I don’t have a way to differentiate currently. The IGDB data would allow me to filter out the HD remasters. I would think if we had the Final Fantasy 3/6 split from the start, only 3 would make it on the list. I don’t think 6 would get enough ratings. I’ll have to look and see how IGDB has those entries mapped to the Giant Bomb entries to see what will happen.
I don’t think there is a clean solution to this, so it may just be best to go with the approach that most easily maps to the user ratings we already have, which is likely going to just be rolling ratings for cases like FFVI into the main entry. Sure there are problems with that in theory with releases of varying quality, but it’s not been like a cataclysmic issue up until this point.
OK, so here’s something about Giantbomb’s approach that feels limiting (to me) – the way they handle compilations. According by GB’s rules, compilations that package multiple games into a single product cannot be entered as a release of one of its constituent games – e.g. Sonic Gems Collection cannot be a release of Sonic R. I can understand the reasoning behind this (I guess?), but I’d like to be able to select Sonic Gems Collection as an release of Sonic R in a review or on my shelves on Grouvee. It’d be cool if the switch to IGDB could make this possible!
The remaster thing is a weird case because some are genuinely significantly different, past just being a different release of the same game. Similarly games that are just different versions of an existing game, like Persona 5 Royal or Diablo III: Ultimate Evil Edition, don’t seem to have an easy answer on how to handle them. Or say, something like Mass Effect Legendary Edition which can be evaluated as a whole package or just as releases of individual games… I’m kinda in favor of things released under different titles still being considered separate “games” for purposes of rating/review, with the exceptions being localization changes like FFVI → FFIII, or cases like Pokemon where they release 2 of the same game at once. While a re-release under the same title on a different console would just be a different release of the same game.
Yeah, that would seem to make sense for the Top 250. The current status of Wind Waker shows that different versions of a game can end up with similar ratings. I wonder if there are many examples of games with (a) one edition reaching Top 250 status as well as (b) other versions so poorly rated as to severely drag down a combined average rating?
I’ve got this noted down in my integration list now. I think what’s going to happen is that the Grouvee/Giant Bomb idea of a Release is going to go away. However, I can see it being possible using their data to write a review of a game, and allow someone to select a compilation from a drop down like we do the releases now. As for shelves, that’ll be an interesting one. They’ll still be separate entries in the database, but I’ll be able to link games to compilations now, so maybe if you have a compilation on your shelf, I can put a see all games thing that would show the games in the compilation if you click it. That way you wouldn’t necessarily have to add those games to your shelves if you don’t own them individually (unless you wanted to of course).
I wanted to look to see how they have Persona 5 Royal and Diablo III Ultimate Evil classified.
This is classified as an “Expanded Game.” This one to me falls into the, it’s a whole new game category, just let it stand on its own.
They have this one classified as a bundle containing Diablo 3 and the Reaper of Souls expansion.
Both of those seem like reasonable classifications.
I think this is reasonable, and that’s why it’s a good thing for the individual games and the compilations to be available as separate entries. We have that now obviously, but it’s going to be so nice to be able to link them.
That made me think that I hadn’t checked how they do Pokemon. They list them all as separate entries. I’m OK with it. If we’re going for a serious collector’s site, people probably want to list them separately.
Not much to contribute, but just wanted to say if IGDB handles special editions of games, that’d be awesome so I can switch over from DekuDeals to here to completely track my collection.
Just thinking about this separately. Tough one. If a series is of fairly consistent quality and/or most users rate every episode of a series, then rolling up all the episodes’ ratings into the main entry would seem fair/representative (I think?) – but if you have a situation that combines (a) a series of inconsistent quality and (b) lots of users rating few episodes, then a rolled-up average rating might not be representative of the series’ quality overall.
I wonder how portals like Trakt deal with overall ratings for episodic TV programmes?
Things like the Telltale episodes and Hitman WOA maps really don’t require separate reviews because they’re parts of the same game released/sold separately (at some point).
For reviews i would offer a template with optional fields:
Platform: PC, PS5, NES, emulator
Version: Pre-release, standard, some DLC, complete (e.g. with all DLC), remastered
Add options to filter by platform/version [include/exclude] and you solve all problems regarding different iterations. You could even generate different top/bottom lists so people could see what the most (un)popular games for various platforms are.
Other optional fields could include:
Modded: Yes/No
Major mods: Open field so you can tell people you used a big mod like Skyrim Requiem or FO4 Horizon
Playtime: 4h18m
Completed main content: Yes/No
Completed optional content: Open field
Playthroughs: Just a number
Also, in general i would allow the community to do more. Allowing people to add tags would help. You could also have some sort of button where they can suggest edits to an entry.
Cool. And @peter, I’d gladly volunteer to help out, this is right up my ally. Well, metadata and curation more than coding, but who can’t learn a little coding right?
I think having ratings reflect an entire title isn’t necessarily a bad thing, it would just necessitate that people think about the title from a more holistic view point in their reviews. But as @anarchistica detailed, having descriptor options to ID which version the reviewer played specifically, and how they played it, would be awesome. And even if the rating number isn’t reflective of every release (e.g. FF6 Pixel Remaster versus FF3 wouldn’t likely have the same rating) but the reviews themselves would accurately reflect the release. I hope that makes sense.
First - sincere thanks to @peter AND all who lend a hand to make these changes. Your time/effort is appreciated.
Second - sounds like IGDB is the way to go for sure. I’m not really familiar, but checked how they label the different Metroids; they’ve already got Metroid Prime Remastered listed/labeled correctly.
Third - in general, the path of least disruption to existing organization systems seems wise. Change, even for the better, is hard.
