Demos...when are they justified

[TCO]

(I was surprised that the earlier demo thread was closed. I thought there was some interesting discussion and I've long disagreed with the idea that a thread starter "owns" the discussion and can get it closed...but that is another matter.)

I'd like to talk about when a game demo is justified. By justified, I mean when is it in the game-maker's economic interests to have a demo? I'd like to talk about the general issue and perhaps refer to Moo3 and SMAC as particular examples.

1. What are the costs associated with creating (and distributing!) the demo?

2. To what extent does a demo drive additional sales?

3. To what extent does a demo ****** sales (because the product is bad or because the demo is so good that the user is satisfied with the demo).

4. Are there certain genres or types of games (new release versus expansion, good game versus not so good, prestigious develeoper versus newby team) that argue for use of a demo?

[Harry Seldon]

I think demos are a double edged sword, which is why I don't usually play them. I don't like playing a crippled game, for one. I also think that by crippling the game into a demo, you are inherently harming its image because unless the playable parts are inclusive the parts reserved for the paying masses will look like flaws. I would also hate to play a demo that was substandard and have it turn me away from a really great game. To directly answer one of your questions, I'm not sure an expansion should get a demo. It seems that many expansion packs these days are just bug fixes and basic enhancements to the original game so what could you really showcase?

I guess I'm either blessed or cursed with the habit of identifying games I will like with just a description to work with. If the game looks like it's not my cup of tea, playing the demo probably won't change my mind.

[TCO]

Makes sense. I guess I could see a demo working well in an RPG or a Myst or story type game. Where they can highlight lots of art, get people hooked and than leave them with a teaser. Like having the first chapter of a book.

[Zed-F]

There is definately a real cost in terms of programming time required to create a demo and ensure that exactly what is desired to be shown is there, even starting from a complete or nearly so code base. Even seemingly simple changes required to "block off" certain aspects of the code have to be thought about carefully to ensure there are no hidden loopholes that will get around those blocks, or cause the code to crash looking for stuff that's not there, and the knock-on effects of removing/crippling some portion of the code are not always obvious. Then there's also the matter of adding in QA time to ensure that the programmers did the job right, and didn't break anything that's supposed to still be there in the process.

With a post-release demo the development team can usually afford to pay that cost unless the product is completely bug-ridden (in which case they probably would have preferred not to release in that condition in the first place.) With a pre-release demo, however, the development team has to plan this into the schedule in order to make sure that the opportunity to create a demo is there. Asking for a dev team to put out a pre-release demo when they have not planned to do one is pretty ludicrous, generally speaking. In that instance the costs for creating a demo go up dramatically as the game was not designed with a demo in mind, and there may be technical obstacles preventing the creation of a reasonably-sized demo in the first place.

In general, a post-release demo will be of better quality and more representative of the game (obviously) so I tend to prefer those. In some specific genres, however, it's easier to build a good pre-release demo. In the FPS world, usually the engine comes first, and content comes second. It's much easier to create a pre-release demo with most of the engine but very limited content for an FPS, as compared with creating a pre-release demo for a TBS like MOO3, where there's no natural breakpoint where a pre-release demo can be inserted.

I'm not sure a post-release demo has a significant impact on game sales, as probably most people that are truly interested in the game will have already got it by the time the demo comes out, and people have other sources of information about the quality of the game available (i.e. reviews, word-of-mouth.) How much a pre-release demo will help or ****** sales is difficult to judge. I would tend to think that a poorly done demo or a demo that doesn't show enough of the game will not help sales, in which case it is a waste of time and resources that could have been better spent on the final game. In that sense the cost-benefit analysis would tend to favour games where the cost of producing a quality pre-release demo is relatively small (such as an FPS,) assuming the game is at a reasonable quality level to begin with. You also don't want to release a demo too much in advance of the final product or the hype machine will get out of control too quickly and peak before the game is ready for release.

[kalbear]

Some times when a demo should be done:

When your game has natural stopping points. FPS games and RTS games are natural for this sort of thing, because they have levels. TBS games generally are not for the same reason.

When the engine does not take up a large percentage of the code size relative to content, or the content can be removed easily to reduce size. Again, FPS and RTS games are easy in this way, TBS games are not.

When there are cycles of programmers left to do so.

When it does not cripple the game to do so.

When the demo can actually be released on a CD or (ideally) be reasonably downloadable.

I suspect strongly that demos don't do so much in the way of selling a product anymore, not like they used to. But I'm no marketer, so I honestly don't know.

[vee4473]

The only reason I even brought up the idea of a demo is because, in my opinion, Firaxis did it nearly flawlessly with SMAC.

Proving that it can be done wonderfully for a TBS game.

SMAC is turn based with no natural stopping points or levels, but TBS games have an edge that RTS games don't that I think can be heavily exploited in a demo.

And the edge they have is the thing, that ironically, most people arguing against a demo would say is the best part of TBS games... The "one more turn syndrome"

If you are playing a TBS game and you have the "one more turn" thing going and the game won't let you take that next turn, you are effectively HOOKED.

I'm not gonna talk code and all that, but i'm just saying that SMAC did it well, so it's possible.

And as to this "cripple" comment...I really don't understand what you guys mean.

[kalbear]

By crippling a game, I mean by removing the components of a game that make it fun.

Removing most tech, opponents, size, customizability, weapons, interaction...the list goes on and on. Many demos limit what kind of interaction you can have with the game. Much of that is done because the interaction that is prohibited isn't quite ready for primetime yet, and they don't want to showcase it. Other because they want to hold things back.

Removing features from builds is often hard to do, and can result in more bugs than it is worth.

I too got totally addicted to the SMAC demo. It was quite good, though it had it's own problems. It is tough to see Moo3 being done as that kind of demo, due to the amount of graphics involved - the demo likely would be huge.

After the thing ships, they should have a demo. Beforehand? Nah.

[vee4473]

Well, you have a point that removing customizing options and techs or other things that make the full game great is a setback. I understand what you mean now.

But, that is what a demo is all about isn't it? To give players a taste of what they will get with the full version.

In effect saying "Here, try this , and if you think this is good, wait until you buy the full retail version and get to design your own units!"



And by the way, holding something out of a demo for the reason of "it's not ready yet" leads once again back to my earlier thread that perhaps there are deeper problems with the game than just some elusive MP bug.

And they aren't talking about it.


What do you guys think of the "one more turn syndrome" being a HUGE hook for TBS demos.

I mean, I think it should be used. If I play Starcraft demo and finish the demo level, I say, "cool game, maybe i'll see if the full game has more to offer"

but if i play a TBS game demo and i'm about to discover a tech that will let me design this new unit so I can take over that region, but the game ends , i say "damn this game has some depth, i can't wait to buy it"

Even if it takes more time and manpower, i say the "one more turn syndrome" has incredible power to hook new gamers.

[kalbear]

By removing things, I was talking more about not just features, but actual code behavior. That's a common thing about doing demos, because you're trying to artificially limit a product that depends on certain things that you don't want to reveal.

An example of what I'm not talking about: not including weapons in a FPS demo. Sure, those weapons would otherwise be in the game, but you can easily make it so that a weapon or two doesn't show up, because they're controllable. Heck, you can simply make it so they don't show up at all in the level (another modular object) and the player will never see 'em. Problem solved.

Something similar could be done with, say, races. Make it so that only two races are playable. You could possibly limit tech, though I doubt it - the trees are too interrelated, and there's a lot of 'preview' code that requires tech that you might not want to show. Turn limits are an artificial way of limiting, and they are good, but they require actually adding a whole nother bit of code to the game engine itself in order to fix that - and that requires work by devs and testers to make sure that it functions. Add to it the whole packaging it up in a palatable, executable form, making sure it works on X machine with Y graphics, and you've got the potential for some serious amounts of work.

I should know - we're working on hammering out a demo of a cell phone to CES as we speak, and it's a pain. We have to do it to meet goals, and it was known, but there's lots of things and corners that we're having to work on - and while we're doing that, the real product sits idly by.

[vee4473]

Kal

I'm not gonna parse coding language with you. Because i'm not a software programmer. I'm a mechanical designer. I know all about making demo units for show from a mechanical standpoint, not software.

All i'm trying to get across is a logical conclusion that anyone should make.

And that is if many game companies, inclusing TBS companies, can make a demo, then I don't believe it is unreasonable to expect the makers of MOO3 to be able to do the same.

Unless they can't because they don't know how, which again points to either IG being in over their heads from a publishing standpoint or the developer being not as good as they should be.

Demos are part of the gaming world I think.

But, hey, we are at a standstill here with this discussion.

I say, i'm really with you Kal and everyone else, when I say let's hope MOO3 is the best it can be!!!

I hope no one takes my posts in the wrong way, I just am very spirited in my love for these games.

We're all together in the end for wanting a great game!

[kalbear]

Well, my main argument is not that they can't do a demo, or that it can't be a good demo - both are demonstratably false. I'm arguing that unless they've budgeted time for it and planned in advance, a demo is often quite a bit more work than what most people think, and that work delays the main product.

My main goal was to educate folks that believe that demos should be easy and quick, since they're just a simplified version of the game. Simple doesn't make it easier to put out, always, and not all games are designed for that sort of thing.

[TCO]

If there were better code control, would that make it easier to do demos? (Are game codes full of kldged together stuff...goto lines and the like? Or are they well constructed?)

[kalbear]

I personally can't speak about QS's coding practices, but I seriously doubt that any game put out these days has that much kludge code. Games are simply too complex for that kind of thing to work, and require significant amounts of engineering to function properly. Usually ,that is.

Look at the source for Quake or Quake 2 some time to get a better idea about this sort of thing. It's quite good, really, even considering The Carmack was behind it.

If you plan a game with a demo in mind, I bet you could do one reasonably easy; some games, however, aren't particularly modular in nature, and this model doesn't fit so well.

[Zed-F]

I'd have to agree with Kalbear, there's no reason to think QS could not have done a demo had they chosen to do so while planning the product. However, it's a non-trivial operation, requires a significant amount of effort, and must be justified.

I wonder if the SMAC developers consider the SMAC demo a success, from a financial standpoint (which is why they did it.) The demo may have been good quality-wise, but that does not mean it had a good return on investment -- did it help sales enough to justify the time spent, or would that time have been better spent on a patch or on getting a jump start on the next product?

I also wonder if the SMAC demo was released before or after the game. That one at least should be easy to answer.

[Mongoose]

SMAC demo was released two months before the game.

SMAX demo was also released before the x-pack, but I cannot recall the interval.

Civ3 had no demo...maybe it didn't pay off in SMAC. Maybe it's a publisher decision and EA said do it and IG said don't.