Had a lot of conversations about the Steam Deck and I figured I should put my figurative money where my mouth is. I’m quite convinced it’s going to be a serious flop. The thing has a host of issues which I’ll lay out below.

Before that though I would make one overarching point: pre-launch excitement and sellouts do not make a product a success. The Ouya and Wii U both had excellent pre-sale / launch weeks and went on to be modern flops.

So with that out of the way, here are four major reasons I think the Steam Deck will ultimately fail in the market and why I think you’d be better served spending your gaming dollars elsewhere.

1. Hardware issues

Much is being made of how much more powerful the Steam Deck is than the Nintendo Switch. And that’s an accurate statement! The hardware is absolutely beefier in terms of both CPU and GPU than the Switch. The problem is that it’s not “enough” for all the supposed uses of the device. To be specific the AMD APU is probably roughly as capable, overall, as a stock PS4 / Xbox One. Not a PS4 Pro or One X, both of those machines are considerably beefier.

The GPU in particular has 8 RDNA2 CUs clocked at 1.4ghz. For comparison the Radeon 6700 XT (a decidedly mid-tier card in 2021, supply shortages notwithstanding), has 40 RDNA2 CUs clocked at 2.4ghz. Put differently this thing has maybe 1/6th the chops of the 6700 XT. This is basically in the GTX 1050 performance space. The GTX 1050 came out as a budget card in 2016, so that’s not a terrific starting space. To be sure some RDNA2 features are going to impress over an old Pascal card, but the starting position is frankly anemic. Even were you to generously bump it up to a GTX 1060 this isn’t a world-beater by any stretch.

The device has a 1280x800 screen, which will help a lot to paper over the weak GPU (this works well enough on the Switch). The problem is that the Switch is already four years old, was underpowered when it came out, and is tolerated in no small part because of the software cachet of Nintendo. Valve no longer has that same mindshare. They famously can’t release a third version of any of their successful games, and their last major software releases have a spotty history. So while Nintendo can skate on brand and a stable of popular IP to move devices, even when underpowered, Valve has no such first party strength.

The battery life commitment is “2-8 hours” which is frankly terrifying. The Switch claims “3-8” (a better floor) but the reality is that even modestly intense games like Breath of the Wild give you closer to 2.5 hours of life. In addition to this the Switch runs on a hardware architecture better known for excellent low power usage (ARM) as compared to x86/x64. Expect the standby time of the Deck to be measured in tens of hours, not days.

Finally we come to the USB 3 docking story. Much has been made of this capability, and I’m sure it works … in a sense. However, given the poor hardware in the device, users docking it to a gaming setup (1080p/1440p monitor) are going to be extremely disappointed in performance. It’s absolutely not going to replace even a mediocre budget gaming rig. There aren’t even any claims (yet) of better performance enabled by being on A/C instead of running off battery.

2. Device Software

Valve is shipping this with steamOS 3.0, which is a Linux variant using Proton (i.e. WINE + mods) to run Windows games. By all accounts Proton is decent at what it does, with carved out exceptions for things like anti-cheat software (prevalent in most popular FPSes and other multiplayer games) which tend to simply not work as they expect to be running in a real Windows environment where they can do all sorts of questionable things. I’m not a believer in the anti-cheat arms race, but it is what it is, and not being able to play games like Fortnite/etc is going to be frustrating for people.

Valve claims they will work with software makers to improve the experience, but the stark reality here is that they’ve been trying and failing to make “Steam on Linux” a thing since 2015. More people use 64 bit Windows 8.1 on Steam than all Linux variants combined (0.92% vs 0.89%) and a graph of historical usage shows that this hasn’t changed meaningfully in over 5 years of attempts to make Linux a thing.

I like Linux and I work with it every day, this isn’t me trying to say Linux is bad or even necessarily unsuitable as a gaming platform. But the sinple reality is that effectively nobody uses Linux for Steam games. This matters, because software vendors need to determine what platforms to target. PC/Linux being a continuous drop in the bucket is going to make it a continuously unpopular choice. The fact that Valve has moved to a “get it working well with Proton” strategy over native ports is proof enough of this reality.

3. Gaming Software

My partner and I took a stroll through our Steam catalogues yesterday to see what we think would make for good handheld fodder. We both absolutely had a pile of indies (bought during various sales) and some classic ports (I’ve repurchased a hell of a lot of Final Fantasy and other SNES-era RPGs from Square) that would probably actually be good fits. The problem in all of these cases is that the same software is already available on the Switch, or on iOS/Android, or both.

To be sure if you somehow don’t own a Switch and you’ve got the money for a Deck this might be enough to entice you. It will run indie/AA type games reasonably well for a while. If your Steam library is replete with these things and you’re desperate to play on the go and Steam Link to your phone with a $50 controller grip isn’t enough already then… yeah, good fit.

The problem is that most other game software we both own on Steam is incredibly unsuited. They’re broadly games which are built for keyboard+mouse input, or games where their PC variants are built for KB+M and have afterthought controller support. While the Deck has little touchpads (and a touchscreen) you can attempt to use as a facsimile for a mouse the simple and awkward truth is that these games were never built to be run on a handheld.

Even those indie games with Switch/mobile versions? Their PC variants were not built for a handheld experience either. Many of them have little launcher apps which are going to be miserable to interact with on a low-resolution touchscreen with poor UIs never designed for touch. The games themselves often make UI changes expressly because they expect to be on a non-touch device class.

Quite frankly, it’s going to be a mess. Lots of software won’t work well, or work at all. The rest of it will already be available on a more performant or more task-suited device.

Finally, as alluded to above, the underpowered hardware is deeply concerning for game developers as well. As the 2013-era gaming consoles are sunset and as massive jumps in GPU capability have arrived / are on the horizon (supply chain allowing) developers will want to target more capable hardware platforms. This means that a platform that can run AAA games from 2019 targeting Xbox One / PS4 but only “ok” and only at a limited resolution is not a good bet for ongoing support in 2022 (when it effectively launches) and beyond. Developers are going to keep moving minimum specs up, and this is already very much on the bottom end of spec targets. It bodes ill for the future.

4. The Price

This thing is expensive for what it is. The $400 variant ships with small and cheap eMMC storage, so while it is technically an entry level item the realistic purchases for decent hardware start at $529. For comparison you can buy a Nintendo Switch and an Xbox Series S for about the same cost (less if buying the un-dockable Switch). The Series S will run absolute rings around the Deck, while the Switch will cover any portability desires and has a tremendous indie library. Both also give you access to exclusive content not available on the Deck, are purpose-built gaming devices for their form factors, and have no looming software concerns in the way the Deck does.

So basically…

The Steam Deck is launching with a host of concerns.

  1. It is built on underpowered hardware with limited future runway
  2. It runs a gaming OS attempting to do API emulation for a more popular platform
  3. It comes with a vast gaming library expressly not built for its form factor or hardware and often ill-suited or unusable for it
  4. Ships at a high price point
  5. Comes from a vendor who tends to give up on failed projects quickly and who has a questionable commitment to actually making and supporting their own games, along with a spotty track record in hardware

Good luck to those who preordered, though 🙃