Skip to main content

Indie game storeFree gamesFun gamesHorror games
Game developmentAssetsComics
SalesBundles
Jobs
TagsGame Engines
(+1)

Whats your pov about game tech, its age and status?  Your example of TPS and FPS being close, is legit, but what happens when the first game is 3D f2p android game and the other is 2D browser furry nsfw game? There are lots of variations to even start comparation, imo..

As the opinion above, itchio gives untied hands for everyone without any harsh rules to follow, combined with its huge content, it's a big pile of data to analyze.

The hidden gems you mention, what would you do with them? How would they serve a purpose in your project? Let's say you found 10 games (hidden gems) which are 99% close to the first game, how would that help your goal? Another example, if all those games are $20 each, i doubt people will happily buy games around.

Whats your pov about game tech, its age and status?

Sorry, I don’t understand this question. Do you mean game recommendations systems or gamedev as a whole?

but what happens when the first game is 3D f2p android game and the other is 2D browser furry nsfw game?

I believe this is possible to build a recommendation system that can handle such complexity. It’s hard, but it’s doable. I think the main tactic here is divide-and-conquer - you split the game’s metadata into different, clearly defined characteristics and work with that. From your example, you can describe a few separate parameters:

  • dimension (3D or 2D)
  • platform (Android native game or PC browser)
  • price (F2P or P2P)
  • age rating (NSFW or SFW)
  • style or whatever you call that parameter (furry or whatever)

After that, you can find a way to correctly classify every game in those categories so you can deterministically compare those games and find their similarity weight.

As the opinion above, itchio gives untied hands for everyone without any harsh rules to follow, combined with its huge content, it’s a big pile of data to analyze.

It will be hard to correctly classify every single tag, but I think it’s possible to classify the most valuable.

The hidden gems you mention, what would you do with them? How would they serve a purpose in your project? Let’s say you found 10 games (hidden gems) which are 99% close to the first game, how would that help your goal? Another example, if all those games are $20 each, i doubt people will happily buy games around.

My current idea is to create a non-biased recommendations system that is based on users’ experience with other games, not on popularity. This is also a good selling point because, from what I know, big recommendation engines like Steam have based their recommendations on popularity and traffic, not on personal experiences. The idea I have is to find not a “good” game but a game that the particular user will like. Some people may enjoy games that are marked as “trash” by the majority and vice versa (some people mark popular games as “trash”).

(+1)

the game games are made in different technology, and game from 10 years ago (with exact same tags and content) is not the same with game released this year. As for the status, some games are work in progress, some are completed and some may be demo. Such diversity may add another set of complexity layer.

The current tease with popularity and sales is working because it's causing viral effect and marketing serves it's purpose. 

The idea you have is interesting, but with the potential walls and complexity, it may not worth enough or be hard to execute and monetize. I bet you heard about youtube/netflix/amazon recommendation system, that it is bad. On the other hand, however, tiktok rocks but it's algos is based on millions on data and it's content can be consumed within seconds/minutes and it's unlike games.

Regardless, good luck and keep us posted!

(+1)

Correct, creating a good recommendation system for anything is a very hard task. I’m for sure can’t do that alone, especially because I’m not a data science guy, I have a bit of expertise in gamedev only. What I hope is to be able to create a prototype and see if anyone is interested. If not - I will just move to another idea.