author: Bart Swiatek
Steam Labs - Interactive Advisor and New Recommendation Systems
Steam developers have announced further experiments that are being conducted under the Steam Labs programme. The entry published on the blog concerns an intelligent Steam advisor, as well as a new way of recommending games based on user reviews. There was also an interesting idea of an in-depth analysis system.
A note appeared on Steam's blog informing about further experiments that are being conducted in Steam Laboratories. The aim of the tests is to improve the platform's recommendation system. Developers are testing the so-called in-depth analysis system and the new community recommendation mechanic. It was also informed that an intelligent advisor has become a permanent part of the platform.
Steam Labs Intelligent Advisor
The second experiment, i.e. an interactive advisor, has become a permanent feature on Steam's homepage. To use its help, just log in to your account - under the games "Trending among friends" there is now a section "Recommended for you by Steam Laboratories", where we will find twelve titles similar to those we have played so far.
Clicking on the bar or the "Discover" button takes you to a page where you can customize your search results.There you will find your play time in individual productions, filters for adding or excluding tags, and two sliders. The first one determines the popularity of the presented games (if you want to see more niche titles, you have to move it to the right). The second allows you to adjust the time at which the recommended productions came out (the default setting is two years).
Deep Dive
The other two experiments are ongoing - for the time being it is not clear whether they will become a permanent part of the platform. The first - deep dive - uses the tags defined by game developers and the activity of the Steam community to offer recommendations to customers. How does it work? Quite simply - just go to this page and enter a title (or choose one from the visible grid that presents games from our library).
Steam will then show us the so-called analysis map, on which we will find a number of items that may be of interest to us. If what we see does not meet our expectations, we can add more titles to the list on the basis of which the list will be built. For example, according to the deep dive mechanics, fans of Baldur's Gate II, Pillars of Eternity, Divinity: Original Sin II and The Witcher 3 may be interested in such productions.
It is also worth noting that by moving the cursor over the displayed tiles we can watch a short video, read a description and learn how similar the game is to the selected titles. It's quite a convenient way to explore new productions, although the interface and speed of the map still leave a bit to be desired (for example, we can start the search with a manually entered title, but by completing the list we are limited to the displayed games - if we enter something new, the whole process will start again).
Community recommendations
Another experiment is the mechanic of community recommendations. On this page we can see the titles that are currently recommended by different Steam users. Each item is accompanied by a review, which allows us to learn a little more about the production. Move the cursor over one of the tiles to display the video and additional information. We also have the ability to filter reviews based on tags, game time and date of publication.
