Making Games

[roadkill]

If you want to make Z-list movies, that's what you do.

It's well known practise!

Filmmakers do utterly sensible things like ignore anything outside of shot, apply makeup to furniture etc. When you actually look into film production you realise that all the clever stuff is done in post, be it effects or editing. Pixar are notorious for having films that test really badly, and need recutting into shape. David Lynch is notorious for effectively shooting randomly and turning it into a story later. This whole "post" stage is impossible with games, you're better off starting from scratch again, which is exactly how Nintendo are rumoured to work.

Any programmer will tell you that without users programming is dead easy. The moment you have some absolute idiot playing around your complexity shoots through the roof. In film they're dealing with finite data sets, still big, but ultimately finite.

These days the overlap between games and film is very real. I've met ex-Renderware staff that worked on the last Harry Potter for example.

[smac]

These days the overlap between games and film is very real. I've met ex-Renderware staff that worked on the last Harry Potter for example.

Sure, but did they work on interactivity -- "gameplay" -- or modelling and animation?

[tiedtiger]

Sure, and games are also much more "front-loaded" than film; if you don't make the right decisions up front (technical, design), then you're screwed later on as the tech. gets frozen.

As games reach the end of production, and control that the designers exercise over the final product disappears, and the technical constraints rule. You have to get your major creative decisions over with at a relatively early stage.

Contrast this with film -- I contend that a major creative act of the director takes place at the end; after all of the raw film is in, the director (usually) edits the film to make the final cut. And can change this edit if screen-tests warrant it.

Your contention is very very wrong.

In the main, you're locked with a film's script as soon as production begins. During production there are times when scripts do get changed, but this is mild changes, specific lines and that sort of thing. Substantially changing scripts are not unknown - in horror stories of film production. But that is also not unknown in some games productions.

Once a shoot is done, it is usually done. Editing takes a lot of time and energy (and is done by an editor, not a director. Directors get to oversee if they're lucky). As for re-shoots of endings, yes this happens. It is not unknown for games to change significantly late in the day based on publisher feedback, however. In both cases it is a horribly expensive decision to make, however, and not usually done lightly.

Unless you're running some sort of wildly iterative process, you don't get that in games design.

Nor in film.

So prototyping becomes much more important. You'd think publishers would realise it's probably better to fund a half-dozen prototypes costing $1m each than spend the money on funding a single monolithic development. Spread the risk.

Welcome to financial ruin. Games don't sell enough on average to recoup the cash of half a dozen prototypes in the hope that one will flower into a successful project. That 6 million dollars plus the development of the product that you pick. The industry needs some much cheaper method of vetting a much larger bunch of ideas. one-in-six is not very likely.

[tiedtiger]

My point is that a scriptwriter can write a script, and on the basis of that script a film can be given the go-ahead. He doesn't need to produce a storyboard, or shoot a scene himself.

Meanwhile a game designer needs proof that his ideas will work, and that means he either needs extra abilities as JPickford says, or he needs mates/co-workers who can do that for him.

But this what I'm saying. There has to be a way for the industry to successfully manage to do this at a sane level, either in written form or physically modelled form that negates having to dole out a lot of money. A scriptwriter can write a script all he likes, but he usually needs to be able to prove to whatever producer he's dealing with that it will work as a film too, btw.

The interactivity plays a huge part. It's no use designing a game without at least knowing if it's going to be fun to play, and the best way to do that is prototyping. And that requires alternate skills. Meanwhile a scriptwriter can easily visualise his film and know if it is working. It's a basic skill most people are born with.

God you really don't have a clue, do you?

Writing a film is NOT easy. There is no 'easily' about it whatsoever.

Anyway, my point regarding interactivity was that its presence does not suddenly render the whole game design process somehow inviolate from the area of adopting some cheaper standardised ways of being approached in either documentary form or otherwise. Interactivity, however, does not imply that games are somehow incredibly harder to make than another equally complex project like a film.

[tiedtiger]

Well, not strictly speaking "production" time -- the film industry has come up with the concept of pre-production, production and post-production phases.

A lot of the timescales your talking about is spent in pre-production, with fewer people involved and lower ongoing costs. One difference between film and games is that the film indstry seems geared to fund pre-production, where the games industry isn't. Except maybe for first-party devcos.

And sure, film also has loads of long days. Unlike software crunch times, 'though, there's a *lot* of sitting around waiting for the next take.

No, pre-production is not included in the time that I quoted. Production, as in shooting the fucker, can take anywhere from 8 weeks to half a year depending on the project. Pre-production, on the other hand, can take a further 6 to 12 months on top of that, and post production is usually 4 months afterward. And writing time is not included in pre-production, and can take anywhere from three months to a year.

Films take bloody ages to get together.

[biglime]

And writing time is not included in pre-production, and can take anywhere from three months to a year.

Films take bloody ages to get together.

Worse than this.

I had a feature film script in development for three and a half years. That's three and a half years of being paid to re-write, re-write and re-write, constantly incorporating the ideas of new people who thought they knew what would make the script better.

R.

[tiedtiger]

It's well known practise!

Not in the demeaning sense that you implied. There is *something* of an art and skill to cinematography, writing and direction you know. Post-production editing won't fix bad photography no matter how hard you try.

Filmmakers do utterly sensible things like ignore anything outside of shot, apply makeup to furniture etc. When you actually look into film production you realise that all the clever stuff is done in post, be it effects or editing. Pixar are notorious for having films that test really badly, and need recutting into shape. David Lynch is notorious for effectively shooting randomly and turning it into a story later. This whole "post" stage is impossible with games, you're better off starting from scratch again, which is exactly how Nintendo are rumoured to work.

David Lynch is also notorious for making seriously hit-and-miss affairs. Most other directors work from a well-developed script.

I've worked on one game that was radically changed three times over the course of its development, and two others that were heavily shifted around. I know of two very famous forthcoming titles that have had a similar experience. Hell, even Halo was an RTS when it started (or so the rumour went).

Any programmer will tell you that without users programming is dead easy. The moment you have some absolute idiot playing around your complexity shoots through the roof. In film they're dealing with finite data sets, still big, but ultimately finite.

Really?

You think dealing with idiot actors is some sort of finite data set? It certainly is not.

[tiedtiger]

Worse than this.

I had a feature film script in development for three and a half years. That's three and a half years of being paid to re-write, re-write and re-write, constantly incorporating the ideas of new people who thought they knew what would make the script better.

R.

Was it a sweet deal or did it make you weep for the insanity of it?

[biglime]

Was it a sweet deal or did it make you weep for the insanity of it?

Sweet insanity.

R.

[tiedtiger]

Sweet insanity.

R.

[Fry Crayola]

God you really don't have a clue, do you?

Writing a film is NOT easy. There is no 'easily' about it whatsoever.

Compare the two. Writing a script is just as hard as designing a game. This is true, I accept this, and though I'm fairly good at one and shite at the other, I'm not about to demean the one I know less about.

However, a scriptwriter can visualise his film. When he's writing it, he can easily see it in his head. If he can't, what the fuck is he doing?

A games designer cannot easily see his design, because it needs to be produced and become something tangible to be tested. He can visualise it in the same way a scriptwriter can, but that's useless because it's an interactive piece and thus should be tested in an interactive form.

Hence my point. There is room in the film industry for pure scriptwriters, because they can do that without needing additional skills. They are not required to create short films to illustrate their film. They know to a large degree during the writing process if the film is workable, because they can play it out in their head, and so can the people who are reading his script. In order to sell it, he requires that the people reading it have an active enough imagination to see what's going on in his film while it's just a document on paper.

However, a pure games designer has his work cut out, because he can't prove his concepts to people without a demo. He can't sell his idea without actually showing it in action to people. And because of this, he either needs more skills, or people who can do it for him. That's why there is less possibility of this becoming an option.

You could argue that a games design group be set up, consisting of the designers and the people who make these designs come to simple life in the form of a basic demonstration of the concept. This could then be shown to prospective developers and the design sold, contracts drawn up, etc. However, would the demo developers be happy with a job that robs them of the true satisfaction of a completed project? There are questions to answer.

[tiedtiger]

Compare the two. Writing a script is just as hard as designing a game. This is true, I accept this, and though I'm fairly good at one and shite at the other, I'm not about to demean the one I know less about.

However, a scriptwriter can visualise his film. When he's writing it, he can easily see it in his head. If he can't, what the fuck is he doing?

Quite often what he's doing is trying to get to the point where he can see what's going on. There's a world of difference between having an idea for a scene and being able to frame it so that it makes sense cinematorgraphically, plot-wise, story-wise, character-wise. It may sound like you're playing a film in your mind and recording what happens on paper, but this is not the case. This is why scenes have to be reworked, reworked and reworked again, because they don't make sense.

A games designer cannot easily see his design, because it needs to be produced and become something tangible to be tested. He can visualise it in the same way a scriptwriter can, but that's useless because it's an interactive piece and thus should be tested in an interactive form.

This I disagree with as well, in so far as the same sort of imaginative process applies. Games designers frequently can visualise the sorts of ideas that they have in mind for their game and draw them out. However, like the scriptwriter, he has to work at it and work at it to see if the idea is going to make sense. So, for example, he might work at it on paper, produce analog equivalents of it, and so on.

However, a pure games designer has his work cut out, because he can't prove his concepts to people without a demo. He can't sell his idea without actually showing it in action to people. And because of this, he either needs more skills, or people who can do it for him. That's why there is less possibility of this becoming an option.

But the reason that he can't sell the idea unless it is in demo form is because nobody has developed an interactive equivalent to the script to show what it is that they intend to make. What you're saying is that game designers need to make demos because that's the way it is. What I'm saying is that is the way it is, but for the industry to have any hope of evolving large quantities of new ideas, the hang-up with demos is a bad idea.

You could argue that a games design group be set up, consisting of the designers and the people who make these designs come to simple life in the form of a basic demonstration of the concept. This could then be shown to prospective developers and the design sold, contracts drawn up, etc. However, would the demo developers be happy with a job that robs them of the true satisfaction of a completed project? There are questions to answer.

Maybe they would, maybe they wouldn't.

In the idea that I had for a small design company, the mode of operation involved around focussing on intellectual property design first and then some means of producing the equivalent of a tv format proposal. I thought that a clever way to get around the problem of the need for a big game engine would be to produce a small video reel instead of what the game would look like and play like, as a means to propose to sell it. The advantage of doing this is that you can turn around a lot of ideas quite quickly. What the business would then be trying to sell the publisher on was the idea of getting the publisher to fund pre-production up to the point of a demo, and then moving from there.

[smac]

In the main, you're locked with a film's script as soon as production begins.

Yes, but production is not the start of the film's development; it's just one phase.

Pre-production is. You're not locked in to the script as soon as pre-production begings but you seem not to count that as part of the film development. The point of pre-production is to do what you say you want to do for games -- spend a *lot* of time thinking about the equivalent of "design" (i.e. scripting for film, design for games).

During pre-production you have the opportunity to write, rewrite, throw out and restart the script; set designers; storyboard or whatever.

I don't know if you could nail down games development to a fixed design developed during a pre-production phase; but you should be able to address high-risk areas of development.

Maybe once games dev. moves from the equivalent of building its own cameras and projection systems.

[smac]

Welcome to financial ruin. Games don't sell enough on average to recoup the cash of half a dozen prototypes in the hope that one will flower into a successful project. That 6 million dollars plus the development of the product that you pick. The industry needs some much cheaper method of vetting a much larger bunch of ideas. one-in-six is not very likely.

Well, pick your own figures from the air to make it work if my random examples are not to your liking :-).

I agree that there has to be a system for prototyping ideas more cheaply; but a game is interactive; you need to be able to demonstrate the mechanics to the extent that the backer is reasonably confident that it will sing. The more gameplay over content, the harder that is.

I guess if you can say "this is in genre X and will play like Y" your problem becomes less one of demonstrating mechanics and one of demonstrating content, which becomes a more film-like proposition.

And how do you demonstrate AI and game logic to a backer, if that's important to your game? I had considered using UML class or state diagrams to model game entities, but let's face it, that's a bit too technical for a typical publisher or producer, and even then is unlikely to demonstrate technical arcana, anyway.

If I were a publisher, I might prefer just to go off previous track records; and then only where a team is working on a genre they are comfortable with.

Oh, wait . . .

[Edit: I originally said "I might prefer just to go off previous tack records". I think "tack" is probably more appropriate for some publishers/devcos . . . my favourite typo today.]

[JPickford (retired mod)]

I agree, prototyping lots of concepts is the only practical solution. They need not be expensive so long as the prototype focused on gameplay over visuals.

The BIG problem with this approach is that it's marketing departments making the decisions nowadays and these people don't play games. They plug a game into a spreadsheet based on it's perceived genre (which pretty much precludes anything very original being approved) and, providing the numbers come out right, they make any further judgment based on how pretty it looks.

It's this kind of approach that got The Sims (temporarily) canned during production; there was no similar game to compare it to - so no sales forecast; and it wasn't especially pretty.

The movie industry is at least prepared to use a little imagination. Films is a visual medium but they are prepared to green light projects based on a script alone.

[roadkill]

The BIG problem with this approach is that it's marketing departments making the decisions nowadays and these people don't play games. They plug a game into a spreadsheet based on it's perceived genre (which pretty much precludes anything very original being approved) and, providing the numbers come out right, they make any further judgment based on how pretty it looks.

This is exactly my problem with many many things. Music and film industries also operate this way, but that initial hurdle, and cost of entry is lower. (ie you can make a single in your bedroom, or write a script for next to nothing, which can then be given to an A&R bod who goes "have 30 million and do it properly).

I did bits of work in the tv production side last year, and I had never realised the near contempt that most people deciding what gets made actually hold for their end consumers. This combined with an ever growing desire by the larger content operators to monitor everything we do, and generate revenue at every conceivable opportunity, games being one, and I can't help thinking we've entered a bit of a downward curve here.

[smac]

What the business would then be trying to sell the publisher on was the idea of getting the publisher to fund pre-production up to the point of a demo, and then moving from there.

Um, I thought you said that funding pre-production up to the point of a demo was a bad idea. Just after you said "Welcome to financial ruin".

When I suggested investing in (out of thin air, the figure of, say) six prototypes rather than one risky project.

[tiedtiger]

I don't know if you could nail down games development to a fixed design developed during a pre-production phase; but you should be able to address high-risk areas of development.

Maybe once games dev. moves from the equivalent of building its own cameras and projection systems.

Actually, many mid-to-large developers have moved (or are moving) into a phases style of development like this. They are looking at long technology and concept prototyping periods (pre-production) followed by large bursts of art and animation production, and then a period of fixing everything at the end (post). Still doesn't fix the initial design problems though.

[roadkill]

I'm coming round to the idea that the art and stuff is what should be done last, not first as we currently do. You should use placeholders until the end, and then make a style that fits with the style of play, and the demographic that's going to appeal to as opposed to the trying to shoehorn a game into an inappropriate style, as so often happens.

This also has the benefit of not having to constantly recreate assets as the technology improves. The downside is the marketers can't hype your game to stupidity whilst it's in production.

[tiedtiger]

Well, pick your own figures from the air to make it work if my random examples are not to your liking :-).

Try one in twenty.

Historically, about 5% of games sold on the street actually are successful, so it's probably a more accurate figure, even counting in those few games that are pushed so heavily that they are successes regardless of quality.

I agree that there has to be a system for prototyping ideas more cheaply; but a game is interactive; you need to be able to demonstrate the mechanics to the extent that the backer is reasonably confident that it will sing. The more gameplay over content, the harder that is.

I guess if you can say "this is in genre X and will play like Y" your problem becomes less one of demonstrating mechanics and one of demonstrating content, which becomes a more film-like proposition.

And how do you demonstrate AI and game logic to a backer, if that's important to your game? I had considered using UML class or state diagrams to model game entities, but let's face it, that's a bit too technical for a typical publisher or producer, and even then is unlikely to demonstrate technical arcana, anyway.

Well the question that has to be asked here is whether you need to show these things to a backer at all. More and more, the answer to that question is 'no'.

I know that sounds counter-intuitive, but bear with me for a second.

The industry's model of oepration has changed radically in recent years, coming around to the idea that it is in the business of franchise entertainment rather than gameplay-delivery. (The blockbuster film industry has done much the same thing). The theory goes that if the product can be shown to have enough bells and whistles, enough character, and preferably an iconic character, then those are the things that will give the brand legs and make it walk, not necessarily the gameplay.

In this new world order, the aesthetic appeal of the brand is as important if not more so than whether the gameplay is startlingly original. No company wants to make a game that is total shit, but they are happy with the idea that the game you're proposing has competent gameplay of a standard type. What matters to them much more now is the setting and the feel. Why?

Because most consumers who buy and play games are more interested in the setting and the feel than the latest avant garde gameplay. It is a mistake to interpret this as 'mainstreamers only want comfort food movie franchises'. Rather, they are just looking for entertainment on an aesthetic level more so than on a mechanical level. But if the mechanical level is great, so much the better.

This is why so many startlingly original ideas fall on the shoals of publishers. Those ideas are usually not backed up by a solid IP design. The IP is usually weak, wishy washy stuff, half-conceptualised cheese. And publishers aren't interested in that because it has no prospects.

So is there a need to show the technical prowess and gameplay intricacy of your game idea above all else? No, there isn't. What you need is the ability to show that you are going to produce a product that will sell.

[tiedtiger]

The movie industry is at least prepared to use a little imagination. Films is a visual medium but they are prepared to green light projects based on a script alone.

But that's only because they've managed to crack onto the idea of a script in the first place. Games need something similar.

[tiedtiger]

Um, I thought you said that funding pre-production up to the point of a demo was a bad idea. Just after you said "Welcome to financial ruin".

When I suggested investing in (out of thin air, the figure of, say) six prototypes rather than one risky project.

Ah, but you've missed the point a bit.

This cheap turn-around idea structure doesn't involve spending a million quid per concept.

What it is is a pre-demo filter, allowing you to vet a much larger number of ideas at first to help you settle on those few demos that you do have the budget to fund.

[roadkill]

But that's only because they've managed to crack onto the idea of a script in the first place. Games need something similar.

The key thing is you can tell what is going to work in a film from a script more easily.

Many of the best games can sound absolutely terrible however well you describe them. (Monkey Ball and Puzzle Bobble really spring to mind there). It's only if you make the mistake of confusing games with interactive stories that you can compare them.

And many crap games sound fantastic in concept . . .

[tiedtiger]

The key thing is you can tell what is going to work in a film from a script more easily.

Many of the best games can sound absolutely terrible however well you describe them. (Monkey Ball and Puzzle Bobble really spring to mind there). It's only if you make the mistake of confusing games with interactive stories that you can compare them.

And many crap games sound fantastic in concept . . .

You could probably convey the whole of Monkey ball with a 30 second animated clip. A lot cheaper and faster than a demo.

[JPickford (retired mod)]

You could probably convey the whole of Monkey ball with a 30 second animated clip. A lot cheaper and faster than a demo.

Nope.

You can't tell if a game plays well from a video. This is the whole problem. There is no substitute for a prototype. We need to find cheap ways to construct PLAYABLE prototypes.

[tiedtiger]

Nope.

You can't tell if a game plays well from a video. This is the whole problem. There is no substitute for a prototype. We need to find cheap ways to construct PLAYABLE prototypes.

I meant as a product pitching device.

[roadkill]

I meant as a product pitching device.

Welcome to mockup screenshots.

Which are literally made to look good with no consideration for how it plays at all. And we've come full circle to the mess the industry is now in anyway.

[tiedtiger]

Welcome to mockup screenshots.

Which are literally made to look good with no consideration for how it plays at all. And we've come full circle to the mess the industry is now in anyway.

Not quite.

Exeecs do like to see an idea in motion. It really sells them.

And a short reel is way cheaper than a demo for those purposes.

[roadkill]

Not quite.

Exeecs do like to see an idea in motion. It really sells them.

And a short reel is way cheaper than a demo for those purposes.

It still means precisely bugger all as to whether the game is fun to actually play.

[tiedtiger]

It still means precisely bugger all as to whether the game is fun to actually play.

Correct.

But that's not why you'd produce it.