Remote game services (OnLive, Gaikai, etc.)

[Cyan]

Do you stream videos?

Yes, I do. I don't mind videos because they are just that, videos, and they are streamed, yes, but also cached.

Videogames can't be cached, that's a Huge difference.

You wrote in the comments of an Eurogamer article that you can't stand poor IQ on consoles anymore (have you ever played a PS4, WiiU, XOne game on a well calibrated TV?), and yet you are okay with video artefacts.

The best use for this technology is imho as a complement for mobile gaming where artefacts aren't an issue and the video stream could help with battery life, I guess, which is usually pretty poor in today's smartphones.

I don't care any more than I care about having a box to play DVDs instead of steaming videos from the net. I just want the content/experience, and don't care for the hardware. The days of interesting hardware are long gone.

That doesn't sound like you at all, knowing how geeky you have been on hardware configurations ever since I've been in this forum.

I wonder what's so uninteresting about current hardware, when we are seeing giant strides, technology wise, in hardware quality and complexity.

If we ever have to play via a video stream and not actual hardware, I am leaving videogames. I like brands, I like things with a soul. I like to buy machines that I know are mine and can modify or love because of their design or capabilities.

What do you prefer, seeing a model flaunting her body in the beach and talking in person to her or seeing a broadcast featuring said model in the beach? That's how streaming videos on the internet look to me compared to the real thing.

 

[Cyan]

Every

networked game hides the latency.

If they're streaming, none whatsoever. If they're not, the game could certainly compensate on their end. If it doesn't, you might have a higher latency, but then again if your TV is lower latency and theirs is higher, or they've got an inferior connection to you, they can still be disadvantaged. Online gaming is something of an illusion. You'll never have player equality and all the interactions are prone to late instructions and best guesses and predicted outcomes and emergency corrections. Most of these are small enough and handled subtly enough that you don't notice, but it's quite possible that your shot's TCIP message is delayed 150 ms longer than your rivals and the system has to adapt to this.

Got your point, I think. In the previous era I remember playing with 150-200ms of lag, yet I never ever had more fun playing a game online like in the case of Call of Juarez multiplayer. I was crying with laughter at times, the game had some unique multiplayer moments that one can't easily forget. Because of the latency, I lost some battles, but still....

Aside from that, I think video streaming of a game is totally impractical with the future, and by that I mean VR and AR.

 

[Shifty Geezer]

You wrote in the comments of an Eurogamer article that you can't stand poor IQ on consoles anymore (have you ever played a PS4, WiiU, XOne game on a well calibrated TV?), and yet you are okay with video artefacts.

Compression artefacts aren't anything like as jarring as shimmer, shader aliasing, and jaggies. Borderlands streamed onto Shield Tablet was much better IQ and framerate than PS3 was.

That doesn't sound like you at all, knowing how geeky you have been on hardware configurations ever since I've been in this forum.

In the olden days, hardware was interesting as it took different philosophies. Now it's just a platform for supplying the content. And I'm fine with that.

I wonder what's so uninteresting about current hardware, when we are seeing giant strides, technology wise, in hardware quality and complexity.

There's one solution, the PC way. There's very little diversity in the hardware resulting in no great difference in what's produced.

I like brands, I like things with a soul. I like to buy machines that I know are mine and can modify or love because of their design or capabilities.

I understand that philosophy, but can't agree to it. There's no reason to view consoles as different from video players.

What do you prefer, seeing a model flaunting her body in the beach and talking in person to her or seeing a broadcast featuring said model in the beach? That's how streaming videos on the internet look to me compared to the real thing.

This is a personal, romaticised psychological interpretation of the hardware on your part. I don't see a console or game as a 'model' and want a personal interaction. It's a set of sights and sounds created in response to my input. The mechanics of that are immaterial - it's the end result I'm interested in. I want games on screen that look good and play well. How they get there is basically a matter of economy, not romantic notions of intrinsic hardware values.

 

[Shifty Geezer]

Aside from that, I think video streaming of a game is totally impractical with the future, and by that I mean VR and AR.

Probably for VR, but I doubt all gaming is going to go VR. There'll be plenty of consumption of 2D games which streaming can supply. In fact, developers may well prefer a single streamed game on every platform when not using the local hardware. Think how much more convenient it would be if COD et al were created in a single server-based executable that is streamed to every PC, console and tablet. Leave the VR game developers to worry about reworking their game for three or four different platforms, and everything else can move to the cloud for remarkable convenience. Of course, that's not possible until cloud gaming can reach every household suitably well. As long as the market is split, it's just more headaches for the poor devs.

 

[Prophecy2k]

It's been said before but serverside streaming tech for games is a great idea in theory, but in practice the business model just doesn't really work.

Last gen, sold 80+ million XB360s and PS3's each. So 160 million+ consoles. Lets say that only 60% of those are active, that's still 96 million boxes. For a streaming service to be capable of replacing these, you'd need datacentres and network infrastructure able to deliver 96 million console's worth of CPU and GPU performance at a minimum by the end of a traditional console generation. But with streaming tech no longer needing to be cyclic like consoles are, that number just increases as more and more people sign up to the service, putting more and more strain on the existing servers/network.

The problem is that no one or even two big players can afford to run the server backend required to service the performance needs of the service consumers. And the economics of the business mean that there is essentially an artificial cap placed on how large in terms of no.s of service subscribers any one service can grow to, which makes it unattractive as a business model to investors.

Streaming tech will forever be relegated to side-add-on services to traditional gaming HW, as long as there still exists a marlet for AAA console-level games. They can't ever really replace a meaty box under the telly. The economics just won't work, unless server chip providers like intel, IBM etc, go full on father christmas and start basically giving the HW away for free (and even then the energy costs for running the datacentres will prove increasingly prohibitive).

 

[Shifty Geezer]

But with streaming tech no longer needing to be cyclic like consoles are, that number just increases as more and more people sign up to the service, putting more and more strain on the existing servers/network.

That depends entirely on what the income is like from the users. If they cover the hardware costs suitably quickly, it'll be financially visible.

I've no idea how that'd cost. EG had a £300 PC article over the weekend. £300 of hardware is barely enough to cover a console's worth of juice, but that's now. The price to put a PS4 online will be reducing.

Streaming tech will forever be relegated to side-add-on services to traditional gaming HW, as long as there still exists a marlet for AAA console-level games. They can't ever really replace a meaty box under the telly. The economics just won't work, unless server chip providers like intel, IBM etc, go full on father christmas and start basically giving the HW away for free (and even then the energy costs for running the datacentres will prove increasingly prohibitive).

Energy costs for running consoles are already part of the cost. That'll be deferred to the costs of running the datacentres.

Or basically, no-one's actually come up with a working business strategy, but that doesn't mean one doesn't exist. Let's say £25 per month all you can eat games from the library. That'd be £300 per annum, covering hardware costs in the first year and providing lots of profit after that. It'll have positives and negatives, but I can see that become an option. Just as movie and audio discs are a dying breed but not gone yet, the industry will likely head towards streamed content of every sort. Although things like VR push the hardware limits so hard that consoles may be around quite a while yet. But never say, "forever"! 50 years from now? Impossible to predict.

 

[dobwal]

It's been said before but serverside streaming tech for games is a great idea in theory, but in practice the business model just doesn't really work.

Last gen, sold 80+ million XB360s and PS3's each. So 160 million+ consoles. Lets say that only 60% of those are active, that's still 96 million boxes. For a streaming service to be capable of replacing these, you'd need datacentres and network infrastructure able to deliver 96 million console's worth of CPU and GPU performance at a minimum by the end of a traditional console generation. But with streaming tech no longer needing to be cyclic like consoles are, that number just increases as more and more people sign up to the service, putting more and more strain on the existing servers/network.

The problem is that no one or even two big players can afford to run the server backend required to service the performance needs of the service consumers. And the economics of the business mean that there is essentially an artificial cap placed on how large in terms of no.s of service subscribers any one service can grow to, which makes it unattractive as a business model to investors.

Streaming tech will forever be relegated to side-add-on services to traditional gaming HW, as long as there still exists a marlet for AAA console-level games. They can't ever really replace a meaty box under the telly. The economics just won't work, unless server chip providers like intel, IBM etc, go full on father christmas and start basically giving the HW away for free (and even then the energy costs for running the datacentres will prove increasingly prohibitive).

Companies like Microsoft, Amazon and Google can more easily support a streaming service because they are already in the business of big data centers. Simply employing a common platform that configurable enough to support the workloads of different services can save on costs.

It isn't about setting up data centers to support a single service. You amortize the cost of the centers across a number of services.

 

[Prophecy2k]

Fixed price "all you can eat" (AYCE) annual subcriptions are precisely why this kind of business doesn't work economically. It's why Onlive struggled to make money, and it's why Sony initially proposedto launch PS Now with a per title rental model, as opposed to this. It's a sure fire way to sink a platform and see a mass exodus of game content providers.

If I'm a AAA game publisher, why would I pump money into a title to try to battle head-to-head with other publishers on a streaming service platform for a measly ever dwinling slice of a per-consumer fixed subscription fee, when II can simply sell consumers a new boxed product every year and reap the benefits.

With the AYCE model, the consumer pay for their sub and doesn't pay another dime, placing hard limits on the content providers' ability to generate revenue growth for their games, whilst also making them almost entirely dependent on the streaming platform owner to grow the subscriber base. No publisher is going to want to place themselves at the mercy of another company like that.

So streaming services entirely replacing console I just cannot see happening, because essentially miracles need to happen before it would ever become a reality.

 

[Prophecy2k]

Companies like Microsoft, Amazon and Google can more easily support a streaming service because they are already in the business of big data centers. Simply employing a common platform that configurable enough to support the workloads of different services can save on costs.

Only to an extent!

Plus those big datacentres are also giving up existing capacity to non-game services that they would have to share with.

Rendering games at > PS4 fidelity per user would need a significant amount more server computing resource than either individually have, if they were to for example transition the 30+ million current gen console owners onto a server based network tomorrow. The CAPEX alone for additional server capacity would more than likely not be covered by the first year's subs, and by the time you start carving away content provider royalties the whole thing starts to look increasingly less worth it.

 

[Prophecy2k]

Energy costs for running consoles are already part of the cost. That'll be deferred to the costs of running the datacentres.

From the streaming service provider's perspective, I'm not sure why thisis a good thing?

Currently a 180W (max) console is paid for by the consumer who buys the box, the games, any online fees and then their elec bill.

By deferring all that ancillary cost onto the service platform provider, how will said provider thinkt to be able to make money whilst still trying to lure consumers away from the status quo? The only viable answer is to increase prices (with whatever business model you like), but that's simply not gonna work for obvious reasons because services like Spotify and Netflix have groomed consumers to expect cheaper pricing of games and services with a streaming service, rather than the converse.

The problem is that centralised game rendering and thin clients isn't like centralised power generation for example. You don't get the same economies of scale. In fact you incur additional costs that would otherwise have been paid for by someone else.

 

[dobwal]

Only to an extent!

Plus those big datacentres are also giving up existing capacity to non-game services that they would have to share with.

Rendering games at > PS4 fidelity per user would need a significant amount more server computing resource than either individually have, if they were to for example transition the 30+ million current gen console owners onto a server based network tomorrow. The CAPEX alone for additional server capacity would more than likely not be covered by the first year's subs, and by the time you start carving away content provider royalties the whole thing starts to look increasingly less worth it.

Thats why they are not trying to do it by tomorrow. Its why MS's plan had the move to streaming happening in the middle of the gen. Generational transitions don't have to be as abrupt or obvious to users when it comes to streaming. The first responsibility of the new hardware can simply be to emulate older hardware until the number of units added gets to the point of ably supporting the release of new gen software.

Furthermore, the subscription model will probably mirror whats we currently have with cable subscription. Mostly likely tiered based with new releases requiring purchases and only older titles offered as free with sub. So at the start of the new gen of hardware, there only has to be enough hardware to support the number of users who want to buy something like the new Battlefield 8.

Since a streaming model isn't about selling the hardware anymore, managing transitions to newer hardware will be different for a streaming model.

 

[Shifty Geezer]

Fixed price "all you can eat" (AYCE) annual subcriptions are precisely why this kind of business doesn't work economically.

It depends on the price. I expect something like mobile plans - however much money for however many minutes. By 'all you can eat', I really meant full access to the libraries. With every game ever (at least from whenever it starts and content is added) at your disposal, an increased price to game is balanced by increased value (game anywhere, any game).

From the streaming service provider's perspective, I'm not sure why thisis a good thing?

As in, just move that cost to the consumers and factor it into the pricing. If console owners pay $20 a year on electricity to run their console, include $20 a year in their subscription to cover running costs.

 

[Silent_Buddha]

As in, just move that cost to the consumers and factor it into the pricing. If console owners pay $20 a year on electricity to run their console, include $20 a year in their subscription to cover running costs.

Except that if it costs 20 USD a year in electricity for the user then it'll cost a fair bit more than that to stream the game to the user. You still need that 20 USD just to run the game. Then you need additional electricity for the compressing and converting it to a streaming format. Preferably faster than realtime in order to lower latency as much as possible, so relatively beefy specs compared to just streaming video content to your TV which only needs to be in realtime. Add in electricity for the additional equipment needed for the far greater network data that will be generated per user compared to traditional cloud services. Then on top of that, whatever client the user is using to play the streamed content is also going to need electricity.

It's a relatively massive undertaking compared to currend cloud offerings which consists mostly of data storage with some relatively light computing (per user) for consumer based cloud services. Even for current consumer cloud applications, much of the work is still done on the client's machine.

For game streaming, pretty much everything is done server side. And on top of that you have the equipment required just to convert that to a gaming friendly, low latency stream. And on top of that, it's going to generate network data in volumes greater than or equal to video streaming services.

And then you have to have a revenue model that is palatable to developers and publishers. IE - whatever it is has to be worth the loss of direct game sales to the consumers.

It's a very difficult thing to make profitable. As you have to make it attractive to consumers, otherwise you have no revenue. You have to make it attractive to publishers and developers, otherwise you have no games. And then you still have to price it so you can make a profit on top of all the costs involved.

Regards,
SB

 

[dobwal]

Except that if it costs 20 USD a year in electricity for the user then it'll cost a fair bit more than that to stream the game to the user. You still need that 20 USD just to run the game. Then you need additional electricity for the compressing and converting it to a streaming format. Preferably faster than realtime in order to lower latency as much as possible, so relatively beefy specs compared to just streaming video content to your TV which only needs to be in realtime. Add in electricity for the additional equipment needed for the far greater network data that will be generated per user compared to traditional cloud services. Then on top of that, whatever client the user is using to play the streamed content is also going to need electricity.

It's a relatively massive undertaking compared to currend cloud offerings which consists mostly of data storage with some relatively light computing (per user) for consumer based cloud services. Even for current consumer cloud applications, much of the work is still done on the client's machine.

For game streaming, pretty much everything is done server side. And on top of that you have the equipment required just to convert that to a gaming friendly, low latency stream. And on top of that, it's going to generate network data in volumes greater than or equal to video streaming services.

And then you have to have a revenue model that is palatable to developers and publishers. IE - whatever it is has to be worth the loss of direct game sales to the consumers.

It's a very difficult thing to make profitable. As you have to make it attractive to consumers, otherwise you have no revenue. You have to make it attractive to publishers and developers, otherwise you have no games. And then you still have to price it so you can make a profit on top of all the costs involved.

Regards,
SB

Yes, its a massive undertaking, but so is the effort to try to get 60-120 million gamers to buy your home console. Only Sony has been able to consistently get those type of numbers every gen.

We all know that both Sony and MS are either already offering or have plans to offer streaming. They are not doing so with the long term plan of only servicing a very limited subset of gamers. Its not even a question of hardware costs being a prohibitive factor anymore. Its a question of adoption.

The average time gamers spend on gaming is actually small.

http://time.com/120476/nielsen-video-games/

6.3 hours per week for the average gamer. A million consoles are being used 6.3 million hours a week for games but offer 168 million hours of potential use per week. The most important figure would be the number of consoles being used during peak times for gaming, but thats an unknown. However, I hardly doubt it would be close to 100% utilization.

MS and Sony probably have a good ideal of the proportion of the userbases thats playing during peak hours. Its probably whats motivating those companies in this direction. Moving consumers to a subscription model with its steady revenue stream and the cost saving of housing a fraction of the hardware in data centers versus what necessary for a home console.

You can't be totally rooted in the realities of today when it comes to technology. Xbox's multiplayer requiring broadband was done at a time when broadband access was limited to a minuscule number of users. The 360 and PS3 were released when the majority of homes were still only SDTV capable.

 

[Shifty Geezer]

The average time gamers spend on gaming is actually small.

http://time.com/120476/nielsen-video-games/

6.3 hours per week for the average gamer. A million consoles are being used 6.3 million hours a week for games but offer 168 million hours of potential use per week. The most important figure would be the number of consoles being used during peak times for gaming, but thats an unknown. However, I hardly doubt it would be close to 100% utilization.

That's a good stat. Sadly it's not as flexible as we might hope, as within a region you'll have peak use times, and due to the need for low latency you'll need local servers for each territory. It won't be possible to allocate the US servers for Japan during the night. But you're right that you don't need to recreate the entire hardware. That's one of the things that occurs to me - to hardware cost of furnishing every users home with a console is significant and inefficient. If you could time-share those consoles, you'd need much less hardware. That's effectively what game streaming does.

It's just working out financial model that'll support it, which no-one's managed yet. But unlike some, I think it's doable. Importantly, the technical challenges seem mostly solved to me.

 

[Prophecy2k]

I agree with you Shifty that the technical challenges are basically solved. That's not really up for debate. The technolgy to do it exists, that's all well and good. The question however is can it be done profitably, and in my mind if there is no immediately obvious way(s) to do that then I'm sorry I just can't believe that anyone is even going to give it enough attention to explore radical business models, when the most obvious simply can't cut it.

I just don't think that the commercial challenges will be solveable for game streaming as the primary platform for delivering high end games to consumers, unless there is a massive and total shift in the entire gaming industry landscape from what we currently have now.

Silent_Budda really eloquently lays out the main challenges facing game streaming:

It's a very difficult thing to make profitable. As (a) you have to make it attractive to consumers, otherwise you have no revenue. (b) You have to make it attractive to publishers and developers, otherwise you have no games. And then (c) you still have to price it so you can make a profit on top of all the costs involved.

(b) and (c) can be easily done, but not without giving up the ability to achieve (a).

I don't think anyone will be able to find a financing model that will satisfy all three, because ultimately the costs being offset by moving HW to the cloud (i.e. the HW and energy consumption) will still need to be paid somewhere, and jacking up prices to pass that onto the consumer will just drive them away.

Trying to obfuscate prices by hiding everything away behind a single subscription payment might fool most consumers, but it will not prove an attractive proposition to content providers who would rather just sell games, DLC and season passes directly to consumers themselves.

 

[Shifty Geezer]

You might be right. However...

Trying to obfuscate prices by hiding everything away behind a single subscription payment might fool most consumers, but it will not prove an attractive proposition to content providers who would rather just sell games, DLC and season passes directly to consumers themselves.

Everyone is moving to subscription models. From game engines to office software to productivity apps to game services (EA's $4 per month service on XB1 I believe). And this has been happening long enough and with enough getting involved that it must be economically viable. I don't think content providers just want to sell games. I think they want to make money, and that might be possible with subscription services. I don't really know. I know Blizzard weren't too cut up about a subscription for WoW instead of selling it. I expect EA would love a $30 per month subscription to access their games. Maybe the future is channels, just like TV content? Subscribe to your favourite providers' channels, and buy monthly passes for one-off titles from other providers.

 

[Prophecy2k]

You might be right. However...Everyone is moving to subscription models. From game engines to office software to productivity apps to game services (EA's $4 per month service on XB1 I believe). And this has been happening long enough and with enough getting involved that it must be economically viable. I don't think content providers just want to sell games. I think they want to make money, and that might be possible with subscription services. I don't really know. I know Blizzard weren't too cut up about a subscription for WoW instead of selling it. I expect EA would love a $30 per month subscription to access their games. Maybe the future is channels, just like TV content? Subscribe to your favourite providers' channels, and buy monthly passes for one-off titles from other providers.

I don't necessarly agree with this assumption. Subscriptions can be economically viable in many industries and sectors, but it depends on the factor that drive each individual industry and the available means for each company to generate revenue.

I wouldn't agree at all that "everyone is moving to subscription models". I would however agree that in the games industry at least, subscriptions models are being utilised where possible to attempt to extract additional revenue from consumers of boxed products and services already provided. In the current gaming industry climate with increasing development budgets and fixed prices on boxed game products, publishers and developers need to do SOMETHING to further monetise their consumerbase, beyond simply trying to sell more games through ads and marketing. DLC and subs in this regard are a necessary addition to the boxed product videogame content sold to consumers and in no case are they intended to supplant them.

If you look at TV for example, content providers don't just rely on a slice of subscription revenues for their income. They can further expand their ability to generate revenue through merchandise, boxed sets and the most important, ad revenue. Game publsihers do not have this luxury, and so a streaming platform asking them to give up selling boxed products to simply live solely off a small slice of sub revenue isn't going to get you a positive response.

Sure, streaming platforms could be designed in a way to allow content providers to profit off ad revenues. But then by doing so you compromise the gaming experience of the consumer who doesn't have to deal with ads interrupting his gameplay sessions with the current status quo, but would with your new system.

No matter how you look at it, streaming just seems it will be a very dicey proposition for either consumers, platform owners, content providers or all of the above. And so the biggest challenge will be garnering investment for it from investor who cannot readily see how such a platform can really have a mainstream future. The relative failures of Onlive and Gaikai to try to carve out any more than a minute niche for themselves only goes to further reinforce this.

 

[Silent_Buddha]

You might be right. However...Everyone is moving to subscription models. From game engines to office software to productivity apps to game services (EA's $4 per month service on XB1 I believe). And this has been happening long enough and with enough getting involved that it must be economically viable. I don't think content providers just want to sell games. I think they want to make money, and that might be possible with subscription services. I don't really know. I know Blizzard weren't too cut up about a subscription for WoW instead of selling it. I expect EA would love a $30 per month subscription to access their games. Maybe the future is channels, just like TV content? Subscribe to your favourite providers' channels, and buy monthly passes for one-off titles from other providers.

EA's model doesn't apply to streaming of games hosted in the cloud, however. So, there's not hosting costs involved for them, other than hosting servers to provide downloads of the game. Which are needed for the sale of the games anyway. So there's no additional costs involved that must be amortized in some way.

It's also generally limited to older games with limited sales potential. And it's still not clear if Microsoft is getting a percentage of monthly revenues from their game "rental" service.

Going back to it, game streaming of older games (greater than 1 year old) somewhat makes sense. You have a limited audience that will limit the hardware investment. Gearing it towards a niche market also allows you to charge higher prices than you'd need if you wanted a larger audience. Older games means it isn't competing with retail games that still command a high profit margin.

In other words, there's a potential market for game streaming. It's still tough. But it's something that might be possible versus a game streaming service that has day and date availability with new games (which I don't think there is any possible way to make work).

Regards,
SB

 

[Prophecy2k]

Entirely agree with you Silent_Budda.