Latest Steam experiment aims to improve Search on the platform
Valve unlocked a new experiment on its Steam platform with the aim to improve search functionality on the gaming platform.
The company launched Steam Labs in July 2019 on Steam to give Steam customers the opportunity to test certain features and provide Valve with feedback and data at the same time.
Valve Software launched three experiments in July that added micro trailers, interactive recommendations, and an automatic daily show that highlights new and popular games to users.
The latest experiment is designed to improve Search on Steam. Steam's search has never been that good; while it supported some filters, e.g. by tag, number of players, or platform, it lacked basic filter options such as price. Search results can be sorted by price but that would change the order of results in the process.
The launch of the Epic Games Store in 2018 introduced a new tough digital PC sales competitor and it is likely that the launch played a role in Valve's recent attempts to improve the Steam client and platform in several ways.
Steam Search experiment
The new Search experience is not all that different from the old. In fact, when you launch it you may have difficulties finding lots of changes.
The most visible change is the new price filter; it is a slider that you may use to narrow down results by price, e.g. to only display search results that cost less than $50. There is also an option to filter by free so that only games that are free to play are returned. A checkbox attached to the price filter module named special offers returns only deals when checked.
Tags list the number of titles next to them in the new interface. The old search interface does not list any numbers and it is often not clear how popular a particular tag is on Steam, especially if a less popular tag is picked. Steam lists tags by number. The numbers provide insight on how popular certain genres are. The indie tag has more than 40,000 titles associated with it, action 27,000, and strategy 14,000.
Not all tags are listed with numbers next to them. When you search for a tag, only some tags may be listed with numbers. It is probably a bug though as you get the numbers of the other tags returned to you when you select the major tag as these may also be listed on the results page.
A search for strategy returns the 14k results figure only for the main term but nothing for turn-based strategy, grand strategy, or strategy rpg. When you check strategy however, you will see that there are 2043 turn-based strategy games listed on Steam currently as the number is returned on the results page.
The results listing uses endless scrolling in the new interface. The older layout separated results by page. Endless scrolling does away with the page flipping but it may lead to issues such as a lack of bookmarking a certain page to continue browsing results at a later point in time or having to restart at the very beginning when you experience issues.
Closing Words
Search is definitely an area that Valve needs to work on as it is in dire need of improvement. The new options are a step in the right direction, especially the new price selector is handy. I wish the slider would support the selection of minimum and maximum and not only maximum.
Now You: What is your take on the Search experiment?
The Steam client is weirdly bloatedly bad and slow. 99% of the time I just start it to play a game. But it is incredibly sluggish, always checks for updates and wastes CPU cycles and screen space on a bunch of stuff I’ll never use, like worthless and non-disableable notifications about “achievements” or whatever that is – cut it out, just let quickly get into the game!
Steam should at least release an alternative simple frontend for just playing the games. One that fully loads in <1 second with no wait for updates at start. Instead do updates silently that for after the game, when the user closes the application.
Google has already solved the search algorithm. Steam simply needs to stop tweaking it to show what they want to sell, instead of showing what we want to play, to improve it for us.
Even so, it’s still a difficult task as what game I might be looking for on any given day depends on what mood I’m in, how much time I have to play at that moment, what I played last, and many other personal factors your going to have a hard time predicting without vast amounts of personal information about me.