Reviewing Linux Gaming Predictions for 2023


It’s already February 2024 and it’s time to review how our predictions fo Linux Gaming in 2023 did. This time we will start with our own predictions at Boiling Steam, and then invite our guests to comment on theirs. Each of us is going to comment below on their own predictions.

Patola’s predictions

Let me go back on my predictions from last year:

In the first half of the year, Valve will release a new non-VR game in the half-life universe

Verdict: Fail ❎

The VR market has had a lot of up and downs in 2023, and way more downs than ups. The PSVR2 gave a good boost to VR gaming (and a number of excellent new VR games appeared on Steam as ports), but it was not enough, even on console it was not generally profitable.

PSVR2 from Sony

As such, Valve was probably right in not hitting anything on the VR market now. Even the SteamVR software was non-functional on Linux for a couple months, only having been fixed recently.

Linux will cross the 3% mark on Valve Hardware Survey until the end of the Year

Verdict: Fail ❎

I could dream, and we’re still not there. But our numbers have been rising. We also got close to 4% on desktop usage, according to statcounter.

Microsoft will release a XBox portable console by the end of the Year

Verdict: Fail ❎

They didn’t. I still think that they might do it this year.

A big AAA publisher will release a game that is native Linux with dxvk-native, so that it runs on the Deck

Verdict: Fail ❎

I didn’t see any new activity on dxvk-native this year. It might be that the project is stalled.

One of the big cloud publishers (Xbox Cloud, Playstation Now, GeForce Now) will start supporting VR via cloud

Verdict: Fail ❎

It did not happen, I think mostly because of what I already said regarding the ups and down of the VR market this year. And maybe it also relies too much on people having very low latency, high bandwidth internet on their homes.

Podiki’s predictions

I only had four predictions back in 2023:

The Steam Deck will be available through at least one major retailer, including in-person brick and mortar stores with demo units.

Verdict: Fail ❎

Although it is available in retailers in Japan, I meant more in the US or Europe, since I think we knew earlier they would be in stores in Japan (such as the launch of the Steam Deck in Edion in Osaka). But maybe I’m misremembering and I get a partial ✅. Would still like to see this though it seems unlikely.

Valve will announce a new VR headset by the end of the year, but it won’t be available in 2023.

Verdict: Fail ❎

Wishful thinking so far! Still the rumors continue and SteamVR 2.0 (including streaming to headsets) might be pointing that way for 2024?

Along with that, Valve will open source (most of) SteamVR for others to use. This will be announced with the new headset.

Verdict: Fail ❎

Utter fail, though some bugs were fixed, others added in the refresh of SteamVR. I would really like Valve to open source this, though I don’t see that happening in the near future. More from me on this soon.

Linux numbers on Steam will continue to increase, and we’ll break the 1.5% level by the end of summer.

Verdict: Correct ✅

Hard to get exact numbers and timing with the noise inherent in Steam’s data, but I was mostly correct from what I can tell. We ended near 2% for the year from what I last saw, so maybe I was a bit pessimistic.

All in all I’m terrible at this, as I would have predicted!

Eki’s predictions

Last year I did miserably in terms of predictions so it probably can’t get any worse this time around!

Valve announces a refresh/revision for the Steam Deck, that does not fundamentally change its capabilities

Verdict: Correct ✅

The Steam OLED was announced and released in 2023, which is essentially exactly what I described. It came with an extensive internals redesign, as well as a larger battery and a better screen.

Steam Deck OLED special edition

And it looks like it’s selling very well.

Linux passes at least once the 2% market share on Steam before the end of the year

Verdict: Fail ❎

It came pretty close, but so far it did not pass the 2% threshold! The latest figure in December 2023 was 1.97%… and in January 2024 it was at 1.95%, so it’s almost like there is some kind of hard ceiling…

Box86 and Box64 will have close to full support by the end of the year on the RISC-V architecture

Verdict: Correct ✅

There’s Super Hexagon, Stardew Valley already running on it, and probably a lot more. We talked a bit about it in our review of the VisionFive 2 RISC-V board, not long ago. Steam won’t run on it for now as there is no support for 32 bits applications, and you’ll need to find games that work with older versions of OpenGL as the Vulkan driver on this board is a limiting factor as well… but if you find titles that meet such requirements, they will run.

Valve will not venture into the stand-alone VR headsets for now

Verdict: Correct ✅

The year is now well behind us, and no such announcement was made, and not even a hint of anything really serious going on in that area.

Proton will become an additional build target for at least one middleware engine (for example Unity, or Godot?)

Verdict: Fail ❎

This was a pure gamble, and it did not pan out so far. Maybe this will become closer to a reality once the Steam Deck becomes more widespread in terms of market size, potentially.

Nils’ predictions

I like to think that I failed miserably on my predictions:

Verdict: Fail ❎

From my metrics: 3.8k on Reddit (not over 5k), under Garuda on both Reddit and Distro Watch. I thought that the fact Nobara comes pretty much ready for games, it would increase in popularity, but I could not be more wrong. It had some growth in reports on ProtonDB, even beating Garuda, but it is shadowed by the growth of other distros. If you would like to learn more about the evolution of gaming distros, check Eki’s article

No new handheld device will be released with SteamOS out of the box (but some might offer support)

Verdict: Fail ❎

When I wrote this, I was thinking about competitors. But Steam released a new Steam Deck (OLED) so I count it as a double fail - one for the prediction and the other for the poor choice of words.

For every single month of Steam Hardware Survey, Nvidia will still be on top of AMD if you discount the Steam Deck GPU

Verdict: Fail ❎

Nvidia dominates with almost 75% of market share in the Steam Hardware Survey(SHS). But this is a Linux focused magazine, and SHS is horrible to get reliable data. The total data on Linux GPU for December 2023 Amounts to 98% (rounding numbers?), and there are +8% of others. I will declare a Fail given the data I have at hand.

GPU Percentage
Nvidia 18.3
AMD 20.41
Intel 11.24
Others 8.34
Deck 39.81
Total 98.1

Some Ubisoft games will work on Linux

Verdict: Fail ❎

In my prediction, I mentioned For Honor and Rainbow Six Siege. For Honor is working, Rainbow Six Siege is not.

Rainbow Six Siege is still not working on Linux

It was interesting to see that other games also got their anti-cheat patched to work with Linux, like Hunt: Showdown and Dead by Daylight. But I am trying to live a life free of live-service games.

Play 100 different games on Linux, less than 10 will fail, and less than 5 will not have a fix

Verdict: Fail ❎

All those numbers are pretty close: I played more than 100, more than 10 failed and less than 5 were not fixed. Besides the live services that were patched, I had some problems with Unity games, some closed betas and some old games.

  • Rising Storm 2: Vietnam - despite people saying it works on non-EAC servers, I could not join a single server.
  • We Who Are About To Die - Stuck forever on a loading screen.
  • Surgeon Simulator VR: Meet The Medic - the game opens, but the headset and controllers do nothing.
  • Total War: Shogun 2 - Tested during the Linux on Mac article. While writing this in January 2024, I checked the ProtonDB and they might have a fix now.


As you can see, most of our predictions tended to fail because we were too optimistic in one way or another. We imagined that Valve would move faster, that SteamOS would get more traction, that the Steam Deck would have more influence as a whole on the market and towards developers and publishers. This is a great lesson in humility: first, it’s very hard to actually predict right for something that we know very well in the first place, and second, the market has a lot more inertia than we’d like to admit. At the end of the day, Valve is playing a long game and there will be visible successes now and there on the way but the path is (probably) long.