So it’s been about two weeks since the Steam Deck launched and reached the hands of the first gamers, besides official reviewers. Remember the Steam Machines? Well that did not go too well. Is this time different? How smooth was the whole thing? Did Valve do well? Let’s look at a couple of metrics together.
While some had expected that the Steam Deck could be delayed further, it was not. It made it on time as communicated by Valve. Now we have no precise idea how many units were actually part of the first batch, but that’s probably not a lot. Our own Podiki pre-ordered the Deck within 7 minutes of the pre-orders going live at the time and did not yet receive the confirmation email during the first batch - which may indicate that there were only a few thousands units ready to ship then. He has received on the 7th of March his confirmation so he should not be getting his unit soon.
Lawrence Yang from Valve confirmed in an interview with IGN that they would expect the following in terms of volume:
- In the first month in the tens of thousands
- In the second month in the hundreds of thousands
Quick napkin calculation reveals that if they planned to ship let’s say 50 000 units the first month and the production is not going at a linear pace but ramping up, the first week could be something like 5000 units and the last week 20 000 units: this would be consistent with about 50 000 units in the first 4 weeks.
It’s also important for Valve not to ship millions of units on Day 1, since they have many minor issues to iron out as the device reaches customers. The more time passes, the more the Steam Deck experience in new hands will improve for the large majority of the audience that’s in the future.
As a side note, the purchasing process is much better than trying to shop for a GPU card these days: at least you can secure your Deck at the announced price and there’s a way to pre-order it in advance without hunting for stock every 10 minutes.
Valve featured the Steam Deck front and center on the Store on the day of the launch. They also had some PR campaign going on with the OnDeck Twitter account, and they hired GabeN to deliver units by himself in the Seattle area. Not sure if that worked as expected (turns out most people were either not surprised to see GabeN or did not recognize him) but at least they tried.
Valve has been way more proactive in promoting the hardware and what they were doing about it in terms of design choices than what happened during the Steam Machines launch back in the days. Overall, a much more dedicated job this time around, for the better.
They went for a strategy to drive demand by having a great product and not spending millions in Marketing (a la Tesla) and they might just be able to pull it off. It was also a great move for GabeN to publicly announce that this is the first of numerous Steam Deck iterations and that they are creating a new category in the PC gaming market, not just a one-time project. This provides further confidence that investing in it (as a developer) will prove a good move.
Since most games are not made for the Steam Deck in mind, the Deck needs to have some level of developer buy-in to ensure that most games work well with it, whether they are native or using Proton. Valve was very proactive in send out devkits out there to encourage developers to try the Deck and test their games with it. There has been quite a few games where specific changes were made ahead of the Steam Deck’s launch, such as:
- Albion Online
- The Long Dark
- Slay the Spire
- Oxygen Not Included
- One Deck Dungeon
- Apex Legends
There was even proper Elden Ring support in a few hours after its release, with specific improvements to make gameplay smoother on the Deck with shader management.
Most surprising was the very recent love letter from the Xbox division to the Steam Deck, where they showcased front and center their games that work well on the platform:
They however had to declare that the following titles did not work because of anti-cheat:
- Gears 5
- Halo MCC
- Halo Infinite
- Microsoft Flight Simulator X
However there were some clear “nay, we will NOT support the Deck” people out there. Bungie, developer of Destiny 2, for example, threatens to ban anyone who tries to bypass BattlEye. Tim Sweeney has also clarified Fornite will not come on the Deck:
There’s also a legion of developers that still are on the fence - even Feral Interactive was apparently unwilling to do any additional work for their Linux ports for the Deck.
Line-up at launch
We mentioned it here before, there were between 800 and 900 games playable on the Deck by the time it launched. Now, it’s over 1,200!. And it’s increasing at a very steady pace of 25 games added on a daily basis. There’s still a long time to go to cover the whole Steam catalog, but at the end of the year there should be already a sizable part of the library playable.
There were however some minor issues: a few titles (a minority) categorized as Verified in the first place, then falling back to Playable status out of the blue: it’s still early days and the exact criteria they are applying may or may not be entirely consistent yet.
User Experience (Reviews)
I would qualify most reviews as positive, while clearly mentionning some potential flaws and annoyances. The hardware and performance is quite impressive all things considered, most people agree on that. But there are certain issues.
- The Steam store is not optimized for the Deck, and the filters don’t show where they should.
- The UI is not entirely consistent across the different menus.
- Battery life is OK-ish but could be better.
- The screen may not be very high quality (but at the same time you probably won’t notice it unless you compare side by side with another device).
While I am not a fan of LTT, their review of the Deck was actually very decent and balanced. If you have some time you might as well listen to what they had to say - Don’t pay attention to the clickbait title:
The fact that most of the issues are software related can give us good confidence that they are fixable, as long as Valve decides to put some manpower behind it.
The main hardware issue is around repairability. Removing the battery proves to be a major PITA and it’s not clear why Valve ended up with such a poor design on that end for an element that is bound to be in dire need of replacement a few years down the road.
Customer Service: Reacting on the early issues and feedback
Pretty good reactivity from Valve so far, following some important hardware issues mentioned by the first users:
- Dead pixels - Valve has been on record replacing units for end users experiencing it, seen a couple of times already on Reddit, if that counts for anything:
- The dreaded controller drift - addressed by firmware days after the report
Valve is apparently on top with very quick follow-up for those affected.
So, how was the launch so far?
- AVAILABILITY: Limited, but improving. At least the launch was not delayed further, and the production is ramping up, with a proper queue system in place.
- COMMUNICATION: Great so far if you follow Valve and Steam, fairly invisible if you are outside of that ecosystem, because there’s no campaign in the mainstream.
- DEVELOPER BUY-IN: Better than expected for the launch time-frame, now the question is whether this will improve over time as more units ship.
- LINE-UP: Probably the best line-up ever for a newly released portable handheld device. Not just large numbers of games playable, but very big titles too, including the very high profile Elden Ring from Day 1 of its release.
- USER EXPERIENCE: Pretty positive so far, while the software has rough edges (especially when it comes to purchasing games on the Deck). Also, it’s not comparable to a console where everything is optimized for it: here, you are expected to tweak more with your device. It’s a PC.
- COSTUMER SERVICE: hardware related issues have so far been handled well by Valve, with either replacements or fixes to the firmware.
All things considered, it was a great launch, It could have been much worse, and for the first entry in this market they did very well. Now, I hope they will however admit that they knew for a long time that the December timeframe was completely irrealistic, and not just for supply chain issues: they were very far from being that ready 3 months earlier. This delay proved very benefitial to give them just enough time to iron things out in every area.