An interesting thought occured to me today.
I have noticed that recently, there seems to be a interesting relationship between the system requirements of a game and its sales and success.
If we have a brief look at some of the more recent, highly successful games (Team Fortress 2, Call of Duty 4, World of Warcraft, Sins of a Solar Empire and World in Conflict - games I just happen to own) we can see that there is a possible relationship between their success and their system requirements.
All games mentioned above have very reasonable system requirements and efficient game engines. If you own a system that was built 4 years ago, chances are you could still play those games reasonably well (with some small, expected sacrifices to graphics).
Of course the games mentioned above could have just had excellent designs that appealled to alot of gamers out there, but its a big coincidence that they also have low system requirements.
I think this has been mentioned by Brad Wardell (CEO of Stardock) in the past someplace on the internet (probably in an interview I read), in that the lower system requirements your game has - the amount of potential buyers for your title will increase. This and the game's genre are the two main factors that affect who may potentially buy your game. Hence why there was a focus with Sins of a Solar Empire to keep the game as accessible as feasibly possible in regards to required hardware. I believe that approach has worked well for Stardock and Ironclad, as we can see Sins of a Solar Empire has been a #1 seller (if you take into account digital sales + retail).
This relationship is somewhat true when viewed in the opposite sense. Crysis while being a solid, good quality title with revolutionary graphics (I happen to enjoy it alot and am fortunate enough to be able to enjoy it at high settings) - has very demanding system requirements. This may have had the effect of limiting the amount of sales the title should've really deserved, because no-one is going to buy a game they can't run at a reasonable level.
Well, you could use atari logic, but well, we all know where that wound up.
Or maybe it's because some idiot in marketing wanted shiny screenshots and a checklist of fancy effects. See the Pentium 4 for other examples of letting marketing design a product.
I think steam caught on to that trend ages ago when they started running their on-installation hardware survey. and if you look at the results some of their customers use some pretty ancient relics.
http://www.steampowered.com/status/survey.html
Thanks for that; I always forget that survey is out there when these kind of conversations come up. Some generally useful tidbits:
- 25% of Steam's users have less than 1GB of RAM.- 59% of Steam's users are using a single-core CPU.- 12% of Steam's users have a CPU speed under 2GHz.- 73% of Steam's users are using 4:3 monitors.- 97% of Steam's users have only one monitor.- 78% of Steam's users have a desktop resolution of 1280x960 or less.- 32% of Steam's users are still using 1024x768.
Unless Stardock has some reason to believe the users of their digital download service are significantly more sophisticated than the users of Valve's digital download service (which seems intuitively unlikely), it seems apparent that even targeting the people that we forum posters typically label the "average user" is aiming too high. A quarter of potential game downloaders have 1GB of RAM or less and a third of them will be playing at 1024x768. You can't exclude percentages that large and expect the game to sell well.
Really puts our conversations about the wonders of 64-bit games (a 64-bit version can't take advantage of RAM that people don't have), multi-monitor support (only 3% of downloaders have more than one monitor) or the like in perspective, doesn't it?
- Ash
Not quite accurate. Since 64 bit OS's allow you to exceed the 2GB program size limit, you could make use of virtual memory -- but it'd run as slow as hell unless most of what's in virtual memory isn't accessed a lot. I add the emphasis because I've seen / heard of cases where games decide they need memory space, but never really use it.
*cough* AoC online *cough*
I remember when I use to play everquest 2. My old rig was not that good but they had made their game scalable. So we could play it well with less graphics. Now I'm sure with my new laptop I could probably play it at full with little trouble. The game as been out what 4 years now or something like that.
So you guys are right, if they did lower the game requirments a lot more poeple would buy it. Is it so hard to make it scalable?
I have forgone a few games knowing I could not play it. No way will I buy something new for a GAME. I buiy a computer usually every four years. Period.
Regarding MOO3, I recall that most of its graphics were quite servicable. The alien diplomats, in particular, were gorgeous. If it had graphical trouble, it was the space battles, which were sort of messy, with everything being too small, including the view area. There were plenty of other more pressing problems than the graphics though.
I think Baldur's Gate still looks decent enough to be playable today actually, because of the painted backgrounds. It's just a shame that the resolution is locked.
I think there is a small flaw in the Steam hardware survey: it completely misses the 1280*1024 resolution, which is very common for 19" monitors. Still, it does drive the point home doesn't it?
I'm sure part of the problem is that the developers are using a souped-up rig (as well they should). You need to make sure the playtesters are using older platforms at some point, and some companies may forget. One could argue that a souped-up box may enhance a QA tester's productivity as well, but the side-effect is that it relegates testing the older platforms to "compatibility testing"--which usually means, "not tested very well". It's a fact of life that software will never be more reliable than on the platform it was originally developed on.
Yes, let's have a look at those recent, highly successful games you mentioned:
Team Fortress 2's success is an unknown quantity - it was not sold on it's own for a long time after release. There's no hard data on how successful it would have been on it's own. Note also that it was a Valve game, as well as a sequel (thus being part of an already established brand). It's success did not depend on system Reqs.
CoD 4 is, again, a sequel to a highly successful brand, and has been subjected to a long and expensive ad campaign. It's success did not depend on System Reqs.
World of Warcraft, yet again, is part of an already established brand, by a developer that get's a free pass by 99% of the gaming media and their followers. It's success did not depend on System Reqs. Note also that WoW is a statistical outlier that is unrivaled, yet sadly taken as "the goal" for MMO development. 11 Million paying subscribers, 132 million dollars per month, is neither normal, nor necessary for a successful MMO, or any game for that matter.
Sins of a Solar Empire was a "success" due to several factors: The amount of sold units that constitutes "success" in it's genre is far lower than those of the previous examples (incidentally the same is true for WiC). It had heavy support from the enthusiast press, due to it's DRM-less nature, and it's moderate system reqs were bought by skimping on visual detail and restricting modders.
As for World in Conflict - I have yet to see evidence of it's success. It is by now all but forgotten by everyone but's its most hardcore playerbase, with the upcoming expansion being an example of "too little, too late". It's success (if it had that) was not a result of it's low system requirements, although this is one of the games where your claim of efficient game engines is true to an extreme. WiC looks far better than it's hardware need would suggest.
Crysis being considered a failure is largely the result of an enormous failure of their PR/Marketing. Someone, somewhere started the Urban Myth that Crysis needs a NASA supercomputer to run and all but a scant few press outlets simply parroted that claim without factchecking it. After they propagated it, the potential buyers believed it too, again without factchecking it.
The truth is Crysis is actually very well optimized. It runs with maxed out settings on a system that is cheaper or as expensive as a run of the mill "next-gen" console.
Actually, Stardock has reason to believe that their users of their service have more sophisticated rigs. There's several reasons for that, but the number one reason is one game. Counterstrike. Not the source version, the old one. After one of the latter patches it required Steam for online play, which will drive an enourmous number of "outdated" machines onto the steam service, to potentially answer such surveys. Don't forget, CS is/was one of the most successful multiplayer FPS games of all times, which means it has/had a huge installbase, all of which would be forced onto steam.
Edit: Sorry for the double post, but it was getting a little long, plus a different arguement/quote altogether.
Did you read that after you posted it?
You're stating that the customer base has more sophisticated rigs than reported, because the customer base has been inflated by all those users that wouldn't be customers without the older, highly popular game they're still playing on their outdated machines.
When you develop a game for the high end of the spectrum, even if it's only top of the last generation requirements, you do more than just cut out potential customers. The development costs of Doom 3 are far higher than they would be for an exactly replica gameplay and story wise that had a game engine a year older in tech. The cutting edge costs a lot of money, the bleeding edge costs a fortune. Those five million dollar game engines require a much larger customer base. Spending four million bucks pushing the boundaries on current capabilities to cut out 20% of your market is stupid. They're all acting like morons, designing games as if they have no talent and can only draw customers in through graphics. Gameplay is first, that is why the CS mod is the most popular online FPS of all time, turning a very successful game into a record breaker. More games would be successful if they'd just stop pissing away small fortunes on graphics in an attempt to get a few more customers than they're losing with the higher requirements.
Actually I'm stating that IMPULSE customer base would have more sophisticated rigs - the STEAM customer base has a host of underpowered machines due to Counterstrike. I didn't word it perfectly though, you're right. You should still be able to see that though if you think it through.
Here we go - you're claiming that CS was successful because it had low system reqs, and despite it's bad graphics. What you're forgetting is that neither of those claims were particularily true at the time of CS's rise to success.
Halflife was a very pretty game at it's time, and it's engine was very scaleable, especially in terms of resolution. CS was successful because it was the first of it's sub-genre that was effectively free (it was a mod) - it was built on an FPS that already had success.
The continued success of CS, even when it's graphics and gameplay started to look dated, was, again, not a result of it's low hardware need, but rather of the simple fact that the so called "pro gaming" scene is deathly afraid of any change whatsoever. That sad characteristic can be observed in all of the genre's but CS is one of the prime examples of it. (it took years and masses of sponsor money to get them to change to CS:Source, despite it being effectively the same game only prettier)
Valve's hardware survey is a fine example of how to do things the right way: Instead of spending endless days philosophizing about it, they went ahead and gathered statistics. They don't need to make guesses about their audience - the have the numbers. That's probably the most valuable tool any game developer can have.
Sure, people can philosophize about why the numbers are what they are - but at the end of the day, the numbers speak for themselves, and you'll pay attention to those numbers if you want to appeal to the largest possible audience.
That gives me an idea - why not a hardware survey with Impulse? I'd love to see a comparison, and I think it would greatly benefit Stardock to see what types of platforms they are appealing to.
I'm going to go out on a limb and say that Stardock may actually be appealing to more lower end systems than Steam!
Remember: Most of Stardock's income is from applications, not games.
I'm one of those 73%. Sadly, monitors still are one of the most expensive parts of a computer, and I have yet to be able to buy a new one. I'm guessing the same is true for most people who are still using 4:3 monitors.
"philosophize" about the results? It's called analyzing. It's what you do with Statistics, otherwise they're meaningless numbers (well, more meaningless than they are, any statistics professional will tell you that you can twist stats into anything).
Numbers never speak for themselves without context.
There are many great features available to you once you register, including:
Sign in or Create Account
