The Great Video Game Business and Financial (In)Stability Thread

Help me walk through this bit from the TechCrunch article; "“The PS4/XB1 is the first generation to have technology that is worse than what is already out there. There are 2+ year old GPUs that outperform these boxes, and even budget GPUs releasing now in the $150 range outclass these machines. If you buy the highest end GPU on the market now, you have almost 3x the performance of these machines, and we are at the start of the generation. This is unprecedented."

Wouldn't this be a good thing for development studios in general? There's no real ramp up on the learning curve. Lower tech specs would make development costs remain grounded for duration of this console generation, correct?

Shouldn't PC ports essentially be mid-range engines with higher resolution textures?

Typing on the phone so the post will be abbreviated.

I think the challenge for the industry is the diversification of audience driving everything towards the unification of entertainment systems and ultimately homogenisation of products. Industry players are trying to dominate the field just as Apple has in the mobile/music/computing world with their closed systems. Of course, the problem is that this approach marginalises individual consumers who have different tastes and preferences.

I think there will be popular pockets of activity which target one area and do it well. The risk is trying to merge genres and models in an attempt to be everything to everyone.

For example, EA's push to monetize the F2P space. Activision Blizzard's RMT AH and their financial transactions fee. Sony and Microsoft's service for monthly subscription model. Mobile platform gaming with aggressive pricing and addictive psychological systems. COD vs BF. Cashing in nostalgia capital (wasteland2, satellite reign, pillars of eternity, insert kickstarter project based on old IP).

My thinking is that there is plenty of space for each pricing/development model. But trying to condense everything into a single offering won't work.

cube wrote:

Uh, doubling the QA requirements is stupid. If you want to make sure that nobody makes a game for your system, sure. That's the reason why no major AAA games actually required the 360 hard drive, even late in the cycle.

And "I'm just sayin" isn't an argument.

Wait, so now you think Nintendo IS doing this?

...

(-:

Hobbes2099 wrote:

Help me walk through this bit from the TechCrunch article; "“The PS4/XB1 is the first generation to have technology that is worse than what is already out there. There are 2+ year old GPUs that outperform these boxes, and even budget GPUs releasing now in the $150 range outclass these machines. If you buy the highest end GPU on the market now, you have almost 3x the performance of these machines, and we are at the start of the generation. This is unprecedented."

Wouldn't this be a good thing for development studios in general? There's no real ramp up on the learning curve. Lower tech specs would make development costs remain grounded for duration of this console generation, correct?

Shouldn't PC ports essentially be mid-range engines with higher resolution textures?

That is generally a good thing for developers yes in the same way that consoles always are because it keeps things consisten. But that statement is also nonsense and not backed up by any kind of real world testing or measurement, just the author's agenda-driven opinion. I would love to see a game like Killzone Shadow Fall actually running at a decent frame rate in 1080p with a $150 GPU. Not to say the PC doesn't quickly outclass consoles (as it always has) but you can't really make a direct comparison between a single component (the GPU) of an open platform where there isn't as much bare metal access to the hardware and a console which has 8GB of GDDR5 RAM in it, RAM which blows away anything you can currently put in a desktop PC in terms of performance. It's also the start of the generation when developers haven't yet figured out the hardware and haven't yet optimised engines to take full advantage of it, another thing that happens every single generation. Not to say there isn't concern to be discussed over the console market but this article's just fishing for clicks through controversy.

I follow a lot of game web sites and not a single one I watch has so much as mentioned the article's existence, much less any validity it might have. Given how much the gaming press loves to quote each other (another tactic for easy clicks), that they've all ignored this one speaks volumes. TechCrunch has never been a reliable journalistic source.

Fair enough; but here's my (poorly thought-out) thesis;

consoles are the line in the sand that developers enjoy now more than ever.
backtrack one generation, when the PS3 promised horsepower out the ears, but it was neigh impossible to code for it; I remember Sony having to burn through even more cash to train developers to properly code for cell processor.

Back to the PS4 and the XB1; instead of learning how to code for new architecture or even feeling the pressure to up the polygon count; Sony and Microsoft have clearly stated what games will look and play like. X amount of polygons on screen, Y of players per session, etc. etc, etc. Studios now have a target then can optimize costs against.

Parallax Abstraction wrote:

It's also the start of the generation when developers haven't yet figured out the hardware and haven't yet optimised engines to take full advantage of it, another thing that happens every single generation. Not to say there isn't concern to be discussed over the console market but this article's just fishing for clicks through controversy.

So even if the article is biased, skewed in favor of $150.00 video cards; it doesn't matter. What I read from this, is that AAA games will still cost $60.00 the first 6 weeks; but now Squenix won't have to sell 19bajillion copies or shut down 70% of their work force; games could be cheaper to produce as early as 2015.

I don't think the article's biased towards anything in particular. I think the author just wanted a contentious point to make to drive traffic.

I agree with you that it does appear these new platforms are much more about making things easier for developers. A lot of people are going "well the new consoles are just PCs" and while that's a bit reductionist, it's also very beneficial as it gives them architectures they easily understand, rather than the nightmare that apparently was the PS3. Of course, that also requires good APIs and such and apparently those still have some ways to go for both systems, likely why they haven't had all the power wrung out yet. In the end, this will also hopefully help the quality of PC ports too.

Came across this response to the TechCrunch article. I kinda wish he dug more into the numbers outside of this three month thing that TechCrunch put forth. Regardless ...

Now, let’s bring Nintendo back into it: The first Wii sold about 1.5 million consoles in its first three months in the U.S., while the Wii U sold 425,000 consoles in November 2012, 460,000 in December and 57,000 in January, for a total of 942,000. For those of you playing along at home, that’s a contraction.

Looking at the latest NPD numbers, the Xbox One is likely to overtake the PS4 in NA in March because sales were about even in February with the PS4 barely edging out the One. So, I'm already going to call Titanfall a hit as the bundled Ones with it and/or Forza 5 seem to have bumped the One's sales in February. That $100 jump between the PS4 and One is partially eliminated by those bundles, coupled with the relative lack of big titles being released for the PS4.

It's getting more interesting, folks.

We'll see. Obviously that's due to Titanfall and due to MS trying to compete more by addressing the price difference. But there is very little on the horizon for both consoles in terms of exclusives.

Europe is where I would worry the most about the Xbox One.. the 360 sold reasonable well in the EU this time around I think we might see a larger split between Sony and Microsoft in the EU. Japan of course is a lost cause and always will be for Microsoft. We will see if anything comes out of the partnership in China for Microsoft.. I still think its a stretch to sell a $500 Console in China.

How "easy" can a Sony, Nintendo or Microsoft plan a new console launch?

In terms of a bi-weekly or monthly release schedule? I think Titanfall coming in in March-April feels like the draught was spread a bit too much for Microsoft, but at least it's getting a good multiplayer shooter.

What's Sony's Q1 and Q2 Sytem Seller (TM)?

Hobbes2099 wrote:

How "easy" can a Sony, Nintendo or Microsoft plan a new console launch?

In terms of a bi-weekly or monthly release schedule? I think Titanfall coming in in March-April feels like the draught was spread a bit too much for Microsoft, but at least it's getting a good multiplayer shooter.

What's Sony's Q1 and Q2 Sytem Seller (TM)?

Probably Infamous this Friday

Tanglebones wrote:
Hobbes2099 wrote:

How "easy" can a Sony, Nintendo or Microsoft plan a new console launch?

In terms of a bi-weekly or monthly release schedule? I think Titanfall coming in in March-April feels like the draught was spread a bit too much for Microsoft, but at least it's getting a good multiplayer shooter.

What's Sony's Q1 and Q2 Sytem Seller (TM)?

Probably Infamous this Friday nothing

I think Sony and Microsoft both lost out when Watch Dogs delayed.

garion333 wrote:
Tanglebones wrote:
Hobbes2099 wrote:

How "easy" can a Sony, Nintendo or Microsoft plan a new console launch?

In terms of a bi-weekly or monthly release schedule? I think Titanfall coming in in March-April feels like the draught was spread a bit too much for Microsoft, but at least it's getting a good multiplayer shooter.

What's Sony's Q1 and Q2 Sytem Seller (TM)?

Probably Infamous this Friday nothing

I think Sony and Microsoft both lost out when Watch Dogs delayed.

That was what killed off my preorder of a PS4. On the other hand, they both seem to be doing just fine

Hobbes2099 wrote:

How "easy" can a Sony, Nintendo or Microsoft plan a new console launch?

In terms of a bi-weekly or monthly release schedule?

How do you mean? There weren't any releases in December for either new console, but there usually aren't. January saw the release of Tomb Raider: Definite Edition (1/28), and February saw The LEGO Movie (2/7), Rayman Legends (2/18), and Thief (2/25). Titanfall came out last week. Tomorrow gets Metal Gear Solid V-ish for both consoles and then Second Son on Friday for the PS4. April and May look pretty busy.

That seems like a steady stream of releases with smaller, downloadable games in between. Unless you're talking about big, system-selling exclusives, in which case most consoles get about one of those a year.

InFamous is the reason I'm planning to get a PS4 whenever they get them in stock up here. There's a lot of people who really dig that series (me among them.) That said, if the PS4 didn't also have a bunch of good downloadable stuff and games like Killzone Shadow Fall that I can rent, I probably would still be waiting. Even Watch_Dogs alone wouldn't sell me a new console.

Has there been any documentation of this series lifetime sales? I think of it as a solid IP but I never thought of it as a system seller (InFamous that is). In any case with 6M+ PS4's out there and a relatively lack of anything to play I know I'm buying a copy.... if it was released with a more "crowded" library I probably would have passed.. I'm sure there are a good amount of people like me so this title (like Titanfall) should sell well.

I'm up to 4 people I know that just got the Titanfall bundle XBox One so it certainly seems to be working for MS.

While I am a fan, I don't think the Infamous series has ever been a huge seller. It also probably hasn't been helped by releasing in June for the past iterations. The only real number I can find is that infamous 2 did 369,000 in its launch month in the NPDs which was reported as about twice what the original game moved in its first month. I haven't seen any cumulative numbers for either title, but even if they were out there they would likely be heavily skewed by giveaways. Infamous 1 was one of the free "we're sorry for losing your data" PSN outage titles, and Infamous 2 was a PS+ giveaway in both Europe and NA for multiple months.

The only positive number going for them is that they have shown growth. Infamous 2 doubled up on the original's launch sales and the later downloadable standalone title, Infamous: Festival of Blood, was supposedly PSN's highest selling title until Journey released and took that honor.

Infamous is no Uncharted, that's for sure.

but you can't really make a direct comparison between a single component (the GPU) of an open platform where there isn't as much bare metal access to the hardware and a console which has 8GB of GDDR5 RAM in it, RAM which blows away anything you can currently put in a desktop PC in terms of performance.

Actually, it doesn't, at least not for everything. GDDR5 is heavily optimized for bandwidth, which is what GPUs need for full screen effects; they need to read in the whole screen and process it, possibly multiple times, and that takes oodles of bandwidth.

But GDDR5 sucks for latency; it's much worse than DDR3. And general computation is much more bounded by latency than by bandwidth. There are bandwidth-choked algorithms for sure, but latency hurts everyone, all the time; nearly all algorithms are impaired by poor latency, where only some are impaired by poor bandwidth. This can be partially covered up with CPU caches and such, but it's a bandaid over a gaping wound.

The real big dirty secret of the industry, the thing that most people don't understand? Memory itself has barely changed in the last fifteen years! The total physical speedup of the actual cells on the actual chips has increased maybe 25 or 30 percent in a decade and a half. Memory is getting smaller, but not faster. So, all these different memory technologies (the various levels of DDR and GDDR) have been about stacking more and more chips next to each other, increasing the amount of data that can be delivered per tick. But the amount of real physical latency, the amount of time the CPU has to wait from when it asks for data from a random place in memory and when that data actually shows up, has barely budged in fifteen years.

GDDR5 is not magic. In fact, for many uses, it's worse. But it's much better for graphics, which is why PS4 games run at higher resolution and with more effects than Xbone ones do.

The PC's split-memory approach may actually end up being better, putting memory that's specialized for different functions in different address spaces. It's hard to tell.

But the amount of real physical latency, the amount of time the CPU has to wait from when it asks for data from a random place in memory and when that data actually shows up, has barely budged in fifteen years.

That's because memory latency is pretty much bounded by the speed of the electrons though the copper and silicon.

Man... do you people ever attribute quotes to authors? It's really annoying to follow a thread without attribution.

Man... do you people ever attribute quotes to authors? It's really annoying to follow a thread without attribution.

IMAGE(http://media.tumblr.com/dbf93f503633364c208a8ba0f597f7b5/tumblr_inline_mtfgcdBhPi1qb5ivq.gif)

That's because memory latency is pretty much bounded by the speed of the electrons though the copper and silicon.

If that were true, then why have CPUs gotten so much faster in the same time period?

There are places where the speed of light does hamper chip design, but reaction time of DRAM is not (yet) one where it's significant, AFAIK.

It's all still running at about 200Mhz, and it's been running at 200MHz since about Y2K.

Tanglebones wrote:
Man... do you people ever attribute quotes to authors? It's really annoying to follow a thread without attribution.

IMAGE(http://media.tumblr.com/dbf93f503633364c208a8ba0f597f7b5/tumblr_inline_mtfgcdBhPi1qb5ivq.gif)

This is where that leads to:
IMAGE(http://i.imgur.com/fRZBY20.png)

Malor wrote:
That's because memory latency is pretty much bounded by the speed of the electrons though the copper and silicon.

If that were true, then why have CPUs gotten so much faster in the same time period?

Lasers!

Malor wrote:
That's because memory latency is pretty much bounded by the speed of the electrons though the copper and silicon.

If that were true, then why have CPUs gotten so much faster in the same time period?

Because, in terms of pure clock speed, they haven't? CPUs have gotten more efficient with prefetching and caching, which increases the cycles per instruction. But in terms of clock, they're pretty much identical.

He who does not name wrote:

If that were true, then why have CPUs gotten so much faster in the same time period?

Latency and throughput aren't the same thing.

Kyle Orland at ars did some major, major number crunching on the publicly available steam info, and came up with this:

http://arstechnica.com/gaming/2014/0...

There are a few issues, especially with Steam under-reporting or flat out failing to report playtimes, but the methodology seems fairly sound, especially because their estimates seem to match what was announced. Also, the fact that the stat tracking is only from 2009 onwards hurts a lot of early releases.

The interesting here is that a lot of things devs have said that people question, like more people playing CoD single player instead of multiplayer, is shown in the numbers.

Also, Football Manager is the favorite game for the people who play it.

cube wrote:

Kyle Orland at ars did some major, major number crunching on the publicly available steam info, and came up with this:

http://arstechnica.com/gaming/2014/0...

There are a few issues, especially with Steam under-reporting or flat out failing to report playtimes, but the methodology seems fairly sound, especially because their estimates seem to match what was announced. Also, the fact that the stat tracking is only from 2009 onwards hurts a lot of early releases.

The interesting here is that a lot of things devs have said that people question, like more people playing CoD single player instead of multiplayer, is shown in the numbers.

Also, Football Manager is the favorite game for the people who play it.

That is incredibly interesting. Thanks for sharing!

cube wrote:

Also, Football Manager is the favorite game for the people who play it.

Not surprising at all.