Hacker News Reader

Comment by boardwaalk | Hacker News Reader

The Venn diagram for target audience and situational use cases of these services has always seemed very small to me.

At least the subscription model and game selection seems better than Stadia.

edit: Also, it seems to me like the target audience is for the same sort (and I’m not trying to denigrate) that watch movies on their TV with internal speakers or can’t fathom why you’d pay more than 30 bucks for headphones or earbuds. Some people just don’t care about the latency, hiccups, or compression artifacts. And that’s fine, but I’m not part of that group.

View on HN

Bear in mind: While very few consumers benefit from cloud-based gaming, the outcome of dragging consumers over to it for businesses is huge.

Why sell software and hardware once when you can sell software, the hardware it's hosted on, and the support and services to manage that software and hardware every single month. It's why companies like Adobe and Microsoft have tried to push their main, relatively static[1] software packages like Office and Creative Suite over to a rent-seeking model. And it's why selling cloud services is all the rage.

It's not a good way to do things, it's not good for any consumer or user of those things, but it prints money on a stable and continuous basis in perpetuity, because people lose everything they paid for when they stop paying.

That's why cloud gaming is going to keep coming up year after year, no matter how impractical or stupid it is.

[1] For the vast majority of users, the difference between any given version of Office or Creative Suite and one released five or six years prior is nearly zero. Almost nobody needs to continually buy these.

> very few consumers benefit from cloud-based gaming

> It's not a good way to do things, it's not good for any consumer or user of those things

> no matter how impractical or stupid it is

Hard disagree on this take.

I would personally love a stable, smooth gaming-as-a-service experience - where I can pay based on my usage level, access any games I want to, pick up my saved game from any device, and never have to worry about sitting around for hours waiting for updates or downloads.

I expect there's a _massive_ market for that. You may not be in it, and that's fine, but I wouldn't be so quick to dismiss it.

> and never have to worry about sitting around for hours waiting for updates or downloads

Here's the thing though. Those downloads and updates have a massive efficiency factor. Once you have downloaded your assets, they don't have do be downloaded again to be used.

With these game streaming platforms, it doesn't matter that you have downloaded 100GB in screen frames. They can only be used once and then are discarded. In some cases this is a benefit - on Flight Simulator 2020 the servers only have to send you a fraction of their terrain data. But that's a small part of it.

> stable, smooth gaming-as-a-service experience

Netflix and the like work because you can buffer. Is your connection a bit choppy today? Hit pause and let it buffer a bit, hopefully it will stabilize. But games do not tolerate ANY latency issues. Even things like Steam Link - using only your local network - can struggle if there's too much traffic or interference.

Then there's the business model. I don't want a 'subscription'. If I buy a game today I may be spending $60 upfront(if the game has been out for a while, often much less), but I can play as much as I want.

Do you have fantastic internet connectivity, an under-powered device, and you want to quickly shuffle through different games from an online selection? Ok, this could make sense.

Still, this sounds like a far better deal for corporations, as opposed to consumers.

The 100GB of game data I have to wait around to download.

The 100GB of streamed frames happen while I'm playing so I don't care.

This benefit is for those who value their time highly and want something to 'just work' without having to bother with installing stuff, updating, buying and setting up new hardware every year, etc.

Imagine you only have 30 mins a month to do gaming... Like many busy adults... With a PS4 or PC, you'd spend most of your 30 mins waiting for updates to complete.

If you only have 30 minutes a month, you’re not the target audience.

It’s also wildly expensive at that point. How much are you paying to game per hour? Much more than going to a movie.

30 minutes is a bit of a trite example but if you only have say less than 10 hours a month the dollars/hour for owning your own your own hardware looks a lot worse than the subscription model.

Even then, it would still be massively more efficient for the game to just stream in the assets in the background while playing, while still rendering everything locally. WoW, for one, has done that for years.

Or you could just start Steam a few minutes before your monthly 30 minutes I guess. If you have the internet connection required for stable cloud gaming, you won't be spending long waiting for the patches anyway anyway.

You’re missing the point. Most people don’t want to download a video instead of streaming directly, which is a far lower barrier than procuring equipment, dealing with compatibility issues, doing system updates etc. I doubt competitive gamers will ever move across but it hugely reduces the barriers for casual or time poor gamers.

You may speak for yourself but you most certainly do not speak for every consumer, regardless of how many superlatives you use.

I, for one, prefer the subscription model for some things. I like that I can dip my toes in and out of the Photoshop ecosystem without having to shell out hundreds up front. I prefer to rent Netflix for a month rather than purchasing DVDs or Blurays I may watch once or twice. And when cloud gaming is baked (it will be a long time away for me as I live in a small island nation) I will love being able to play any game from a huge range with no outlay other than my monthly subscription - no hardware requirements, no upgrading, just a few bucks a month, and only on the months when I choose to keep my subscription.

My only caveat with cloud gaming is that mouse + KB is non-negotiable, I'm not sure how realistic that will be. I will not play shooter or strategy games with a controller.

The nationwide fibre in NZ means that latency from Auckland to anywhere with fibre is very low. Cloud gaming could arrive in NZ just as soon as Azure does!

I think the problem with Cloud gaming is the latency. And the network requirements push it to solving a problem for a very small subset of users (games which where latency doesn't matter can usually dont need cloud gaming).

The primary issue is the games catalog, something neither Amazon or Google can crack a dent in

Yes. All I want is a latency-optimized EC2 (I'm no twitch gamer, <= 25ms is fine) instance with a decent streaming client on pc/mobile. Let me decide to install steam or epic or origin or whatever and use my own stuff.

This is where NVidia went wrong. They're basically providing that sort of thing, but wanted to present it as their own game service as though all the other storefronts were just backend plumbing. Not a bad consumer experience, but also not what publishers are licensing when they stick their game on Steam.

25ms is a huge ask

Even locally, on a console, you’ll not get latency (input to visible response on screen) down to 25ms these days.

Some LCD TVs alone can add 40-60ms of latency, and software can add another couple of frames. Often 30hz frames, too...

Which kind of sucks if you grew up on CRTs and arcade games that often managed much lower latency.

But I guess it means that game streaming really could offer a console-like experience if as much non-network latency as possible was eliminated...

I believe that's the joke. I wish latency was getting better but every new thing that comes along moves the acceptance further away. Many don't even notice 500 ms delay from click to something actually happening. On my Atari I often had the feeling something started happening before the mouse button was all the way down.

I can't believe that people wouldn't notice 500ms of latency when playing a game. Even 150-200ms is very noticeable.

Sure, in 3D games it's a different topic but I have seen people playing with wireless mouse that lags behind quite significantly without noticing.

It's no better than my normal broadband, which is 10 to 20ms, but I suppose you're right. I've streamed games with up to 100ms without much issue with input lag so I suppose that would be fine, although 100ms would probably be too high for some types of games & multiplayer.

Under 25ms is less than current game consoles deliver. Most deliver around 100ms latencies.

I think you're missing the point: Yes, multiple pieces of equipment in the process adds some ms in between input, computer processing it, and screen outputting it.

The issue here is how much will ping time add to that when streaming a game. Consoles hover in the 100ms are depending on game. PC's it's highly dependent on video hardware, but I good gaming system, display, and keyboard/mouse will be around 70ms.

The issue of streaming game ping ms in part comes down to human reaction time, which averages around 200ms-250ms. Given this, you want a ping time that doesn't add enough latency to the setup to put you above that 200-250ms reaction time, or you start to perceive more noticeable input lag. Given 70-100ms baseline, you want a bandwidth ping absolutely no more than about 100ms on top of that.

The math squares with my personal experience as well. I've played a few streaming services, and when my ping gets to the 100ms area I start to notice a little difference. It's often sill playable, but if it gets higher it can be very distracting.

fwiw the Adobe Photography plan requires a 1-year commitment (paid monthly) and has an early termination fee if you do try to cancel, so it's not as straightforward.

I don't have a controller. Mouse + Keyboard works well in Stadia (I played Doom Eternal). Stable low latency internet is a must though.

I found Adobe opaque and difficult to work with. Subscriptions are presented with an $x/month but you have to pay for a year. Cancelling, at least at the time I last dealt with them, was also very difficult: I paid for a year, but then wanted to cancel auto-renewal: Nope! I'd have to wait until the last 30 days of my subscription. Since then I've decided to suck it up and deal with Gimp and Inkscape even if they lack some features and feel a bit clunky at times.

If you only need the basic stuff , lookup application portables.

> I like that I can dip my toes in and out of the Photoshop ecosystem without having to shell out hundreds up front.

I agree, that part is nice. However, once I've dipped in a toe and know I want to OWN it, I can't. The non-saas option is gone.

So now I have a perverse incentive to argue against you, for my own sake. Because the more people like you enter a market, the less likely that market is to provide a single purchase product.

> I will not play shooter or strategy games with a controller.

I was like you, years ago, but what pulled me to the other side was "couch convenience". Might be similar to the other convenience factors you list. Unless you're playing competitively (where %s matter), there's something great about playing a game like Control while lazing around on the couch

I haven't used office products unless obligated too in ~6 years. cloud based options, even terrible cloud based options like quip are an order of magnitude better than word. It's faster to open the documents I care about, easier to share, easier to collaborate, and just plain faster.

The word docs I still use end up having 3 dozen copies in my downloads folder from different versions sent via email. It's a nightmare.

If it requires a $6/mo subscription for the cloud based equivalent of a $100 consumer software package, I'd gladly pay it.

I think its interesting how almost all content has gone to streaming and subscriptions but games have not. The way most people use games seems to be fundamentally different though. I listen to new music every day so subscriptions work best. With games I always play the same stuff, I buy a new game once every few months so a subscription would cost me more.

If games go subscription there will be less money for the big developers. Netflix is a net looser and selling dvds makes more money for publishers.

I think this is a long tail type issue, assuming publishers can find a way to get a profit sharing agreement from the service providers based on time played. In the short of things, publishers make some money on each sale and then monetize on micro transaction, season passes, etc. The clear push towards “games as a service” from all the big publishers reflects this line of thought, i.e. we want to keep making money from this forever (or as long as the company can be counted on to actually run the service). But, and I admit this part assumes some form of payment to the publisher based on continued play, if the game is played via streaming subscription (which would probably have tiers or some such for access to individual publisher’s libraries) then the publishers can in theory keep making money on a finished product over the long-term.

However, I agree that this system is very likely to NOT benefit developers, it will be the publishers that make the cash and it will become a haggling point in dev/pub contracts as to the amount, of any, the devs see from these continuing fee arrangements.

All said, I cannot see a time when I’m willing to go in for full game streaming as a stand alone product. I can imangine paying for Microsoft’s game network or something if I wanted a console this generation. But I have 7000+ DOS games, and a few external drives full of Windows games, and then emulators and ROMs, so I think I’m good for gaming and don’t fit the target for a streaming service.

At this point, even after Stadia, if Amazon is getting into, and i trust Amazon biz and market analytics more than Google or Xbox, the only conclusion I can come to is that they know the venn diagram is small but they must have data showing that it's getting bigger. They must be targeting people growing up just playing on their phones already paying subscription who want to play those cool AAA games with the smallest investment possible. That's me backing into this conclusion. I'll look forward to see what the next Deconstructor of Fun podcast says about this. They focus on the biz/money part of games and have been appropriately bearish on streaming (Stadia) in the past.

The latency is the only “situational use case” to care about. This has been mitigated more and more. The rest is just gatekeeping gamers imagining what a larger group of wallets would care about and being strong. Chronic problem in that community. Isnt the biggest protest amongst enthusiast gamers based around at what point they will pay the company anyway? Hard to take any stance there seriously when there is zero power over what happens, companies dont get cancelled, 2% of their audience just doesnt preorder the thing they were going to buy anyway, amazing.

Hiccups and artifacts are largely solved and where they aren't, the market tolerates it just fine.

There has been a whole decade of this stuff. The arguments are literally from last decade and they weren’t issues even with the inferior infrastructure then.

I dont think any of these services have proven themselves as a profitable endeavor on their own yet, but as far as advertising or a competitive ecosystem for a larger brand, theyve done really well.

It's not just people gatekeeping, developers are usually going to target whatever gives them the biggest return. PC games have been gimped for a long time because the developers make sure the game runs on consoles as well, and if cloud gaming takes off, it'll be even less likely that people develop games with PC players in mind. I can't blame people for wanting cloud gaming to fail; if it's succesful it could cause major changes to their hobby as games are developed for the lowest common denominator.

Not to mention it makes modding dead on arrival. I get it, most gamers do not mod. But being able to write mods and work with Microsoft Flight Simulator in the 90s is what got me hooked on computing in the first place.

First, you have people who are "real" gamers. Whatever that means, but we'll define it as owning some piece of hardware that was bought for the purpose of gaming (PS4, Xbox, Switch, PC). This number is between 250-500 million people.

Yes. Its that high. PS4 sold ~115M units. XB1 ~50M. Switch ~70M. Steam has ~100M MAUs. There's overlap in those, and those numbers also really under-represent China and the huge PC gaming scene there, but I think we're in the right magnitude.

By and large, this group's interest in cloud gaming would only significantly extend to being able to access the games they already play on the hardware they have, when out and about. Xbox, Nvidia, PlayStation, and whatever Valve releases will own this group. They're not re-buying games on Stadia or Luna and locking themselves into that platform, nor dealing with locked cross-saves, cross-achievements, cross-social, its just not happening.

Second group: Lets say iOS users. Representing 1.5 billion people worldwide, significant overlap with the previous group. These users probably wont get too heavily invested into any cloud streaming service, because they are all inaccessible on their mobile device. Yeah, there's a Luna web app: I'm sure it'll be horrible.

Third group: People with bad or no internet. 160 million Americans lack broadband. Some of them still buy game consoles and suffer through weeks of downloading updates. Most don't. Online play sucks, but 150 ping is manageable. You learn to play with the ping. These people will never use any cloud streaming service.

We could go on, but the point I'm reaching is: I, too, have no clue who will be signing up for these services. It really feels like these cloud companies just have graphics cards laying around and are supremely uncreative with how to get value from them. Or, they want to own the world and its taken them thirty fucking years to recognize that, hey, gaming is a thing and boy are they late to the party.

Gaming hardware is not expensive. A Switch costs $200; a brand spanking new Xbox Series S, $300. You don't need a two thousand dollar PC to play all the great games out there. Stadia and Luna want you to buy their controllers and their dongles, the upfront cost is not zero, and consumers are not blind to the ongoing cost and expensive games (Game Pass anyone?! Its a fucking steal! And it practically comes with xCloud! Why do people still think Amazon and Google can compete with this?)

If there was some investment vehicle which allowed me to specifically short Stadia and Luna, I would put every cent in my savings and retirement into it, down to the last penny. Cloud streaming has a future, but it won't be via these two products.

> Why do people still think Amazon and Google can compete with this?

Maybe their target demographics are different, they want attract a new set of gamers who were not serious about gaming earlier.

I used to care about those things and then I had kids and my priorities changed and I just want real quick access to 45 mins of media or games.

I bet that will change yet again when they get older.

I really like your perspective: there’s millions of consumers out there with different priorities. You don’t have to “get” it you just have to appreciate that they exist.

Hawe you actually tried Stadia or Geforce Now?

I don't have fiber, so I'm at the mercy of how much demand there is for Comcast at any given time. Sometimes it works great, other times not so much. I'd rather just install games locally like with Game Pass, Steam, or EPIC. My local computer can handle these games just fine. I'm definitely not the target audience since I can afford the hardware.

The only place I can think of in the US with fiber are university dorms.

Outside of providing a cheap substitute for hardware, the only other use case I can see this for is for MMO's that have drastically changing environments like Rec Room, VR Chat or anything like 2nd Life

You don't need fiber. As for use cases, i just now played an hour of The Division 2 on my laptop (with intel integrated gpu) using 4g connection from my phone.

You do if you don't want lag or weird graphical artifacts. You need fiber if you want a local install experience.

The experience playing on 4g was about the same or even better than local install with Geforce 1050.

My connection degrades depending on peak demand, because cable, and because internet duopolies. It's great that yours doesn't, but that's not reality for everyone.

Also for most people in the US, 4G has significant usage caps.

Also - not everyone lives in the US :)

That’s why I brought it up. Our internet services are worse than most industrialized countries. It’s both slower and more expensive. To my knowledge only Canada’s internet services are worse

Currently, Ookla ranks the US's average fixed broadband speeds as 11th globally. This outranks Sweeden, Denmark, Norway, Luxembourg, New Zealand, Japan, Taiwan, Portugal, Israel, Poland, Germany, Finland, Belgium, Russia, Estonia, the Netherlands, and so many others. It might be more expensive than many other countries, but its definitely not the slowest average internet out there.

https://www.speedtest.net/global-index

I used to live in both Taiwan's and Japan's major metros. I have doubts that they are slower and more expensive than the US's metros unless their fiber network was somehow destroyed.

Depending on the time frames in question, it easily could have been that Japan and Taiwan had massively faster internet on average. Ten years ago, a 10Mbit connection here in the US seemed to be pretty dang fast.[0] However, speeds in many metro areas in the US have massively caught up to the rest of the world. Rural areas are still often left in the cold, but in many cities its not uncommon to find 200Mbit+ residential internet connections available. And like I've mentioned elsewhere, gigabit is available in literally hundreds of towns in the US. The average AT&T internet user these days has ~99Mbit, the 90th percentile having 323Mbit.[1] I doubt that was the case even five years ago. So its not a case of Japan or Taiwan's major fiber metro areas getting destroyed, but more of the US finally getting good metro fiber deployed to more customers. Maybe the internet hasn't changed in the last decade where you're at in the US, but its very different in many other areas.

Of course this speaks nothing to the average prices these consumers pay. However, its probably a lot cheaper to wire up an area less than the size of California (Japan + Taiwan) with a much more massive population density. If everything else was equal (its not) it would have surprised me for internet to have been more expensive in Japan or Taiwan than in the US.

[0] https://transition.fcc.gov/national-broadband-plan/broadband... [1] https://broadbandnow.com/ATT-speed-test

1. You'll never have a local install experience

2. Anything above ~25mbps connection will give you the optimal experience you can have

So no, you do not need Fiber, you just need a decent broadband connection.

> The only place I ca think of in the US with fiber are university dorms

While definitely nowhere near the majority of the US, there are many metro areas with large residential fiber to the home (FTTH) deployments. Suggesting it is only available at university dorms is massively misrepresenting reality.

In pretty much all the areas I've lived for the last decade, FTTH has been extremely common. I'm certain lots of other people have had very different experiences from me. Lots of other people have had experiences vastly different from yours.

https://broadbandnow.com/report/fiber-optic-availability-map...

https://www.att.com/shop/internet/gigapower/coverage-map.htm...

https://www.verizon.com/home/fiosavailability/

Just tried Geforce now as I spotted they had a free offering that links to my Steam library.

Booted up "Just Cause 3" (first one I spotted that was supported and I had). The first bit of the game has you on top of a plane shooting at ground targets.

I'm on a 40Mb/s connection with 22ms of ping to 8.8.8.8. All wired ethernet, no wifi.

The latency was awful, felt like ~1s from mouse move to action. I could not aim at the targets at all.

0/10 unplayable ;)

I disagree, I think pay once for a game probably matches better with the VEnn diagram of the target audience, which probably are people who don't game enough to buy a whole console/PC, but want to try the latest NBA or big game like Cyberpunk. Paying a one time 60$ for one game seems better than subscription if you play a few hours a week.

being able to game while also being cheap laptop only is something huge numbers of teens with disposable pocket money have. These services (fortnite, csgo, etc) want the billions all those pocket money slices add up to.

Reply on news.ycombinator.com