TL;DR: Some way to differentiate canonical franchise entries from noncanonical ones would be awesome.

First off, I love that IGDB accepts and maintains data about ROM hacks, fan games, and other "unofficial" titles. It really helps the platform's position and usability as a database for all games of all kinds.

That being said, when looking through the search results for a franchise like Pokemon, there are an incredible number of these unofficial games mixed into the results, many of which are built off of the existing, official games. There's nothing (as far as I can tell, anyway) in the metadata outside of the game's free text description that separates something like Pokemon Blaze Black (a ROM hack) from Pokemon Black or Pokemon Black 2 (both official games).

I think it would be great to have a tag of some kind that denotes the work as unofficial or fanmade. I see that ROM hacks are commonly listed as "mods" of their base game, but this doesn't encapsulate games made in tools like RPG Maker or Unity or games made using their own custom engines. I would also argue that ROM hacks and mods as they're known in the PC gaming scene are two different concepts, but that's a subject for a longer debate.

As it stands, there is a lot of confusion and mixed up data between Collections and Franchises, all relating to the series in which a game belongs. Some games will be within a collection without a franchise, while other series titles such as "Legend of Zelda" has 2 different database ID's for its collection and its franchise.

Because of these differences, it's hard to utilise the series distinction in searches, as it might be in one category or the other, or both. If both are merged, there would be a more consistent way of determining a "series", with many games being a part of multiple "series".

Hi,
when I check my own channel on Twitch, I like to see the colouredful section name something like "Recently broadcast categories" (in Italian "Categorie trasmesse di recente), with all Game Covers from recently streamed games/categories.

Anyway, seeing I bring lot of indie (expecially less less known ones) I often find they have no Game Cover at all related to their category. And they show and ugly grey empty cover box.

But most of them actually can show a Game Cover on IGDB... or if they don't, well I often upload a cover by myself, if I can.

The fact is...: doing this, doesn't update Twitch's category cover. :(

My suggestion is:
seeing that covers are first of all verified and accepted by IGDB admins, why not making Twitch use directly Game Covers from IGDB database?

I wish for a more colorful "Recently broadcast categories" section on my channel for the future!
Thank you!

Games' ratings may be skewed or conflicting if they have wildly different ports.

For example, "The Legend of Spyro: A New Beginning" has its console ports, which are fundamentally the same, but it also has a 2D side-scroller port for GBA, and an isometric port for DS; both vary greatly (especially the GBA port) from the console release. Another example is Shrek 2, where the PC port is effectively an entirely different game from the console ports (and the matching PC port was called "Shrek 2: Team Action!")

Metacritic already separates game ratings by every port, and imo that's how it should be.

Several games have interesting or obscure caveats in terms of their release or their playability, and it's worth mentioning them for the sake of completeness. For example, Donkey Kong Junior is listed as a Gameboy Advance release, which is technically correct, but it was only playable via the e-Reader, which is not obvious. Additionally, Donkey Kong Junior also exists as a playable game within Game and Watch Gallery 3 and 4, which may technically not count as a release but deserves mentioning somewhere. There are several examples of games with unique scenarios like this, and this data doesn't really have a section where it could easily be integrated. Furthermore, some games have memes or cultural phenomena around them which also lacks a section. A trivia section to serve as a catch-all for data like this would be great, so long as said trivia can be cited and verified.