Another Rockstar VR Game Could Be in The Works, According to Actor’s Resume

That’s the reason for the “sells to (much) less than 50%/would sell less than 2mn”. I agree, these numbers are gross overestimations. The 2mn is basically the ceiling, a very unlikely best-case scenario that I use to not immediately get shot down for “being too pessimistic”. And even with this overly optimistic scenario, the resulting numbers aren’t anywhere they’d need to be for a AAA to consider releasing one of their usual games for VR.

Without something changing dramatically, the most likely options will remain indie titles, the occasional AA venture, VR modes for existing games and very rare AAA flagship projects like HL:A, HCotM or Quest RE4 with a significant strategic investment from the platform owner. And even the flagship projects tend to be mostly slices of the worlds from established franchises with reduced scope like Assassin’s Creed Nexus, with the 2021 Quest RE4 being the exception due to being the port of a 2011 HD remaster of a 2005 GameCube title.

The typical sales ratio between Quest and PCVR isn’t as bad though. There are several games that sold 10x as many on Quest as on PCVR, for example Ultrawings 2, but these are usually games featuring simpler graphics that PCVR users aren’t fond of. Ultrawings 2 was also the only “realistic” flight-sim on Quest, while PCVR had X-Plane, MSFS etc. And in early 2021 Skydance didn’t say that the Quest version of Saints & Sinners sold 10x as many as the PCVR version, instead it sold 10x as many as the (not cross-buy) Rift/Rift S version on the Oculus store. That ratio probably has gotten a lot worse with now ~20K ratings for Quest vs. ~1K ratings for Rift/Rift S.

On Steam it has ~6.5K ratings, so it is more like ~2.5-3:1 Quest:PCVR. Typical sales ratios for games that actually make use of a VR PC like Green Hell VR are more 4:1, which makes sense given that we see ~2mn monthly Steam VR users, and monthly Quest use so far peaked at ~6.5mn, with some of these Quest users actually also making up 50% of the Steam VR users.

How many users are there:
– 1.87% (Steam May hardware survey) of 132mn monthly active users (Steam Year in Review 2021) ≈ 2,47mn monthly Steam VR users
– 6.37mn monthly Quest players for 2022-10 according to Wall Street Journal, based on internal Meta documents

Sales (requirements)/Costs:
– GTA V: USD 800mn revenue on launch day, USD 1bn revenue in the first three days, USD 6bn lifetime revenue incl. GTA Online with 90m copies sold reported by 2018. By now estimates are USD 7.7bn+ in revenue, 160mn copies of GTA V for ~230mn PS3-5, ~160mn Xbox 360-Series X/S and PC, with the PC market at less than 75% of the revenue of consoles and including free versions like from the Epic Store plus lots of promotions to get more people onto GTA Online. So ~670mn devices/160mn copies ≈ 24% market share, actually lower due to ignoring people getting multiple copies for multiple platforms/generations.
– GTA V is estimated to have cost USD 140mn to develop, 40% more than GTA IV, with the total budget for GTA V development and publishing (meaning including a gigantic marketing campaign) being USD 265mn.
– The RE4 release on Quest 2 hasn’t led to any significant jump in headset sales in late 2021, while the release of HL:A basically doubled the number of monthly Steam VR users within a few weeks without them later falling back to previous levels. Quest 2 holiday sales lead to a yearly jump up in Steam VR numbers, followed by a drop down almost to the level from the previous fall around March/April, which has been a repeating pattern. If RE4 had moved lots of new users to invest into a Quest 2, we would have seen a jump in new reviews for other VR games on the Quest store too in the following months, and a bigger jump in the percentage of Steam VR users with Quest 2. Neither of which happened, so “[my] assumption […] that there wouldn’t be any growth in headset sales because of the game” (that I didn’t actually make) would have some precedent.

Also I’m not the one mixing up mods and ports and hybrid design and native VR games. The starting point was the VR-only Assassin’s Creed Nexus, RoadToVR’s “pared down (re: not open-world)” statement and the misinterpretation of this being due to technical limitations instead of development costs. GTA V was my (extreme) example for the cost of creating a large and rich open world game and the unit sales numbers that pay for the development. Of course AAA will rerelease their successful games over and over again on multiple platforms/console generations to redistribute cost, and Skyrim for Vison Pro and Apple Watch is probably only a question of time. But they cannot count on creating the next 20 year franchise, each game has to make back its development and marketing costs itself, as success and therefore further cost reducing ports/re-releases/sequels aren’t guaranteed.

As mentioned in my other comments, the currently best way out of this dilemma for VR is hybrid design that allows spreading the development costs over flat and VR and multiple platforms/generations, with mods of existing games a workaround that doesn’t require support, but at least being tolerated by the game authors.

All this is well known. The numbers are all well known too. Which is why “someone has to take the plunge” is what VR enthusiasts want, but rarely happens, because instead “someone does the math and consequently decides against taking the plunge”. The VR community just declaring that the industry needs to jumpstart VR with AAA titles and will later benefit from the grown user numbers, doesn’t make it true. In fact sales for the few VR AAA ventures were both low and not very impactful, except for HL:A. Which was also the first release of one of the most desired franchises in gaming since 2007, getting people to buy it DESPITE VR and having to pay up to USD 1000 for an Index. And even though HL:A was an extremely successful VR game, the money it made absolutely pales to what Valve would have made with investing the same amount of time and effort into HL2 E3 or even HL3.

I think we know exactly why there is so little AAA engagement in VR, and all the ideas by VR enthusiasts why Microsoft or Sony or whoever should throw eight or nine figure sums at it aren’t based on assumptions at least considering the actual numbers, but mostly on wishful thinking.

Mods are something very different. Your reaction (that I responded to) was to a quote about Assassin’s Creed Nexus, which isn’t a mod of an existing game, but a new game entirely created for VR, meaning it has been created from scratch, with all the resulting development costs I was referring too. That’s like building a new house.

A mod only takes an existing game and adds/replaces camera and input to different degrees (and in some cases a lot more), which of course also incurs development costs, but not for building the world. That’s more like changing the locks in an existing house to give access to other people. Obviously a lot cheaper than building it from scratch.

VR mods have been around longer than VR games. People started hacking stereoscopic cameras into VR the moment the DK1 dropped, with early attempts for Skyrim, Dishonored or Mirror’s Edge keeping all gamepad controls and only adding 360° stereoscopy, with sometimes jarring effects like very tiny mountains and the sky hanging at ceiling level. And there are still a number of (very interesting!) YouTube videos of John Carmack at E3 2012, where he demonstrated a prototype of the DK1 before the Kickstarter launch with a Doom 3 BFG mod he had created.

The problem with mods has always been the different interaction style in VR. Flat FPS games have the character run at impossible speeds, take away the camera control and leave out all the small details that make a world feel alive, as players wouldn’t notice them while rushing by. So “designed for VR” games have been seen as the holy grail for a long time, with e.g. HL:A tuning down the speed a lot, but adding tons of details you can interact with and making interactions that would be a simple button press in a flat game, like reloading or upgrading a weapon, getting health and collecting items/loot, intricate physical activities.

That’s very hard to integrate into existing games, though everybody’s preferred base game for excessive modding, Skyrim, can be turned into something that has a better native feel than most VR native games thanks to a metric ton of mods. With “mods = cheap, but lacking integrated design” and “native = well integrated design, but require lots of sales to pay for development”, the compromise that Sony has been pushing for is VR modes from hybrid design, where VR isn’t the main target, but the integration is already planed for during design, resulting in a game that can work well both in flat and VR. The result is for example RE8 cut scenes playing in 3D, so you can still look around, while RE4 on Quest shows them on a virtual 16:9 screen.

So:
– GTA V modded for VR: quite cheap since Rock Star added a 1st person view, lacking native controls and only for those with a strong stomach. For some unknown reason Rock Star doesn’t like/stopped the 3rd party mods, significantly reducing the chance we’ll ever get to play GTA V or RDR 2 in VR.

– GTA X VR only: not going to happen, because GTA is one of the most expensive to produce franchises, with the matching required revenue. Meta’s Quest store generated USD 1bn in revenue with several hundreds of games over the course of 2.5 years, that’s what GTA V made in the first three days after launch. 80% of that was on launch day alone, with more than USD 6bn in lifetime revenue so far (incl. GTA online)

– GTA VI with hybrid design and well integrated VR mode: might happen, as the added development cost would be much closer to that of a well done mod, and it largely depends on whether the designers will see it a good fit. But the timeline for the release will to 99% be determined by the flat version, because a similar percentage of the development costs will be payed for by the flat releases on PlayStation and Xbox. VR may desperately need versions of GTA or RDR, but Rock Star doesn’t need VR.

RE8 managed to avoid a lot of the things that would make it an uncomfortable VR experience, but the whacky physics are a staple of GTA (and very stomach turning in VR), so it wouldn’t be trivial to come up with a mass-compatible hybrid design. Gorilla Games didn’t manage to find such a solution, so instead of extending the existing open-world Horizon Zero Dawn/Forbidden West, we got the (VR only) spin-off HCotM that takes place in the same world, but removes most of the open world aspect, replaces the combat with a more on-rails version and turns the travel-talk-puzzle-combat game loop into mostly a variant of “The Climb”, keeping almost only the very pretty look from the original game (and probably reusing assets from Forbidden West to reduce development costs), and not much else.