If a game does well, its time to lay off half (or more) of the team. Same result happens if a game does poorly of course. But it seems the only way to 'win' is to be at the top, or simply not play.
I've seen this now with everything from Harmonix to Irrational Games. There seems to be a huge amount of money made with these blockbuster games, but vanishingly few companies seem to be able to manage their game development cycles efficiently as to always need a staff. It always comes off as terrible management/project management.
For example, Harmonix's Rock Band was huge. There was around $299 million of bonuses paid to people at the top. Yet, I had friends work there get laid off repeatedly (once right before Christmas), sometimes shortly after the people at the top got the bonuses. Why in the world didn't they think to diversify a bit, run a few concurrent development cycles, etc...
The most sane way to do game development seems to be to start your own indie studio and keep your expenses very low. Everything else seems... irrational.
Programming jobs in the games industry (and the entertainment industry in general) are not like "normal" programming jobs, despite the fact that they require an almost identical skill set. Put another way, they require the skills of a programmer but function on the employment schedule of a Hollywood lightning technician.
(It sounds like the company your friend worked at was also mismanaged on top of this cycle, which always makes things worse. Sometimes very much worse.)
It's also not uncommon for the internal tooling that gets created to allow the game development folks to build out the game to dwarf the actual game code is size/scope. There tends to be a large upfront effort from the software engineering side to get the engine ready and the supporting internal tooling ready for the game designers to go to town, after that it can be years in the game development process with only bugs, maintenance and smaller feature additions to the engine/tooling for the software dev side of the process.
With such a lopsided time investment schedule between disciplines you need to be either a massive company or only work on games with a small scope to achieve smooth employment for all involved.
Last game I worked it was around 100 dev people where 30% was software development, and from those 30% maybe 20% were tooling support.
All this works on the assumption that scripters are not coders which is debatable.
The movie industry should be the same, but it isn't because outside of a few key people working the project (the actors, directors, producers etc.), almost all of the rest of the people are hired in via contract to production companies. So an SFX house or a Foley studio can keep their staff working on several projects at once and not have to go through as much of the hire/work/fire cycle. Cameras are rented from equipment companies and camera operators are similarly contracted.
So when you go to make a movie, the studio hires the key staff, then contracts out pretty much all the rest of the work to companies like Maslow Media Group or whatever and you get vetted professional tradespeople who come in, set up, do their job, and then move on to the next gig their production company has lined up for them. So their layoffs really only happen when the entire industry takes a hit.
In the game industry, they're still learning this lesson. There's a few places where outside contracting exists, mostly in sound design and music and middleware developers (various engines or tools), but that's about it. They still hire on teams of developers and artists, build the thing then fire all of them.
You could imagine that remote work might help in this situation, but I've heard stories about the difficulty of communicating artistic vision across geographically disperse teams.
Instead of going through expensive and time consuming HR processes every year, the publishers could just go through half a dozen contract negotiations instead.
One well known example of this model is TOSE, https://en.wikipedia.org/wiki/Tose_Co.,_Ltd.
So the only way to keep all of Irrational as it is now (remember, the team grew substantially through the development of Infinite) together is to immediately turn around and work on another AAA game, and at that one that's already ready for immediate large-scale development. That's the model of someplace like Infinity Ward or DICE, where you need to churn out this year's iteration of Franchise X. It doesn't sound like Levine wanted to do that, though, and if the 15 or so people he's keeping with him are all key people at Irrational, you're left with a large staff but no leadership to shepherd them through getting Obligatory Bioshock Sequel developed.
When you develop Windows 8, you know what's next (depending on the scale you want to look at -- there's Windows 8.1, Windows 9). Same with Office -- you iterate on the existing product to improve it and add features. There's a handful of cash cow game franchises that feature a similar iterative development model. But much of video game development is a more artistic or creative process, for a lack of a better term (that isn't to say that regular computer programming isn't creative, I think it very much is, but it's a different sort of creativity). And expecting to get that kind of creativity through project management is how you get Call of Duty games. I'd rather Levine work on more original games like Bioshock Infinite, rather than working on more Bioshock games that are unoriginal.
Absolutely! And you know why ? Because they don't aspire to make good /games/. They are actively trying to become like Hollywood. They're making spectacles and stories. Replayablity be damned. It's a dirty word. Modern games are often sold on production values.
Trailers! Trailers are a thing straight from movie industry. Nowadays, even strategy games are commonly advertised with trailers. Even strategy games usually have promotional screenshots with GUI turned off. They don't want to show you how the game /plays/, just how it /looks/. Gameplay videos are increasingly rare, and the term is becoming meaningless anyway because of the coming of "gameplay" trailers, which dillute the meaning of the former.
Bullshit. Nobody's forcing anyone to go and work in the games industry if they don't want to.
The big game development studios have always been like this. Half the time, they do have concurrent dev cycles, multiple games in the works, concepts being fleshed out for the next game long before the current game ships.
Here's the problem: the publishers. They give your studio money at the beginning of a development process, and that entitles them to be utter ogres about schedules and ship dates. It's very common for a "next project" to be yanked out of your studio's hands and placed with another firm. It's also common for development studios to pull in outside firms to help them on projects, as when Gearbox pushed the Aliens Colonial Marines game to Timegate Studios.
The end result is that there is this humongous push to get whatever is next to ship out the door, and to hire and freelance help around that project. Once that project is done, shifting to something new would require repurposing everyone, which means a few months of unproductive work, or worse yet: a few months of no work... Check out some of the details in http://www.gameinformer.com/b/features/archive/2014/02/12/fa...
Going from the period of game development where you're experimenting, to the time where you're full-bore pushing on a final vision is not something you do overnight. it requires a lot of prep work from game designers, engine developers, and art directors. If they've just been doodling around and trying to "find the fun" as is the primary goal in most early game dev processes, then they're probably not ready to gear up for a huge push on something as soon as the previous game ships.
Thus, you get this scenario where AAA require millions of dollars to make, but developers don't have that kind of cash (save a very few), so they get it from a publisher, who then handcuffs them to ship dates. Anything you can do to get that ship date met will result in a return of vastly more money to your studio, due to how contracts tend to work, so you throw money at the problem through staffing up.
Then, when the game's done, you go around to publishers with the next idea and take on another million dollar contract. Why? Because the funds you receive from publishing that last game don't come into your account (if you actually get any money at all: many agreements are only for the up-front $) immediately. These are effectively royalties, so you will basically be waiting a full quarter to get your next check from the publisher. At this point, you can't afford to float 200 AAA developers and artists for a full quarter.
This is why online, in-game-transaction games are utterly decimating the publishers in terms of growth. Riot Games is immense. Valve makes oodles. Why? There are no publishers between them and the player. All the money comes back to them I don't think Riot's had a layoff yet. Valve's layoffs were entirely project-based, not funds-based.
Publishing yourself online is the future, in every way. It encourages agile development, and lets people go home at night to sleep (except the server admins, but they can't sleep in any industry).
> There’s no great way to lay people off, and our first concern is to make sure that the people who are leaving have as much support as we can give them during this transition.
I've been through that recently (individually, though: specific skill-set, not right for the stage of development of the game) and recommendations go a very long way. Gaming, especially AAA, is actually a lot like the film industry on the HR point of view, except they don't really admit it: one-to-two-year-long projects, with a release date; little need for the same workers after that; many very specific, creative, expertise; a lot of rule of thumb, best practice to master through apprenticeship; recommendation being the tell-all; producers with far too often a horrible reputation; etc. Having a structure that handles that (union, special temporary worker status) would help facing the reality, which is frequently changing employers. Studios, movies and games alike, like to work with the same people for the next projects if things went well -- but half the time, no re-employment in the film industry is not a bad review, just a different creative direction. I'm not saying Hollywood Unions are the best example (or that French’ high direct subsidies and Status d'intermittent are cheap) but thinking about the differences between the two industry would help.
(Keep in mind this is a semantic distinction, and not a value judgement of the game itself.)
As others have noted, AAA games are hit-based propositions. You pretty much have a few (3-5) titles that make all of the money each year, another 10 or so that hold the line, and everyone else loses. It's the last group that are flops, the second group are the par result. I think we can safely say that Bioshock Infinite was in the second group...
Valve gets a 30% slice for distribution on Steam.
I do not know what the breakdown of costs are on consoles, but I assume both Sony and MS still get a cut of the sales.
Consider that the bulk of a game's sales will be realized during the first week, and the rapid time to discount, it's no wonder T2 decided to pull the cord on Irrational Games.
Source: I bought it on Steam for 8€
Many of the games produced are fine games, and some even sell well, but may not match the (possibly extremely overoptimistic) financial forecasts. So, "flop" is probably pejorative. It does fit what the business reaction seems to be to some of these games, but that may be a problem originating with something other than the game.
To me, "flop" means it failed to meet the rather ambitious expectations, and instead performed only moderately well.
By comparison, Modern Warfare 3 sold 10m copies on day one.
Rough outline of the business plan behind most AAA titles:
1. Assume all 10 of us have a >50% chance at being among the year's top 3 sellers.
2. Scale budget according to assumptions from step 1.
Instead, I read a story of a guy who wants to do something else, and decided that the best way to do that is to fire most of his employees. Quite frankly, I'm surprised this is even legal. Handing over the group seems like a much more sensible plan.
there's nothing stopping these fired employees from banding together and forming their own company.
* SWAT 4: How cool is a multiplayer shooter where you actually have to breach a room from multiple sides to pressure the enemy into _not shooting_? And hold your guns until you saw any indication they would? We played that game for nights in one room for better communication.
* System Shock 2: Deeply flawed in some regards, but also the first game that creeped me out in a _perfectly well lit and bright environment_. Shodan, as always, was a great enemy.
* Freedom Force Series: A comic strategy game. It wasn't that hard (it wasn't easy, either), but had "comic" written all over the place. The description if you hovered the cursor over a mere building was "A proud participant of the Patriot City skyline." Someone put an ironic joke on the patriotic theme of the game in the description of a boring apartment block... How fun is that?
BioShock was a culmination of all that. Would you kindly pay them your respect?
No, simply because Bioshock, while being wildly entertaining to many people, was so at the cost of many of the good design choices made in the System Shock games. Just see for example your note about the bright environments and contrast with Bioshock.
(It also irks me a little that your post is missing just enough detail to make it seem as if they made SS2, instead of being co-creators under the lead of Looking Glass.)
I am split whether Bioshock threw away too much or just enough from System Shock 2. At my first playthrough, I was on the edge about this. The "would you kindly" scene cought me as off-guard as the Shodan reveal in SS2, so I forgive Bioshock a lot in the story department. It was definitely the more polished game in many regards.
SS2 is hard to play nowadays and some things (like the ghosts) just really didn't work in hindsight. Also, the crafting and inventory-tetris didn't really add to the game. There is a lot in it that could be removed for good. On the other hand, SS2 stood at a time where those things hadn't been properly tried yet and is a good example of a game that has to be commended for trying out new things. And damn, was it great to actually explore the story.
I heard a (german) podcast about old games a while ago where one of the podcasters said that he thinks that gamers in the 90s had more tolerance for games trying something and failing, because it was such a rapidly evolving medium.
Perhaps, Bioshock was the System Shock for the 2000s and us old-school gamers wanted it to be something that relentlessly tries out things just as in the old times.
Still, Bioshock is light-years ahead in story-telling and world-building than most games of that era. "Would you kindly" is a great example: It shows that the game was built around a narrative and not around a series of set pieces.
Oh man I miss this game so badly. I just loved every piece of it. I tried to buy it now, but unfortunately it's out of stock everywhere. Apparently because the publisher has closed. And there's not much on eBay :(
I recently played through SS2 again, largely for nostalgic reasons, and here's a list of the mods I used, in the order in which I installed them. (Install order matters because, SS2 not being designed for mods, they overwrite core game files; System Shock 2 Mod Manager is highly recommended for keeping everything straight.)
I won't link them because they should all be easily searchable, at most by adding "ss2 mod" to the mod name in your search query; if you can't find one, comment and I'll update with a link.
* Anomalies, Discrepancies, and Outright Bugs (ADaOB) v0.3.0 - bugfixes, &c.
* four hundred v?? - hi-res textures
* SHTUP Beta 6 - more texture improvements
* DeepFriedBeer's sound upgrade v3b - much improved game sound
* PSI-Amp V2 - High-poly/high-res texture replacement for the psionic amplifier weapon.
* Rebirth 02 - High-poly-count model replacements; goes very well with high-res texture packs.
* The following "Tacticool" weapon replacements:
- Assault rifle v1.0
- Grenade launcher v1.0
- Laser pistol v1.1
- Shotgun v1.0
- Wrench v1.0
- Pistol v.11b (ADaOB v0.2.8 compatible)
* Vurt's Hi-Res Space v1.1 - High-resolution skybox texture
I won't pretend that, even with these mods, the result is anything like a modern game in terms of looks. But it's a lot better than stock SS2, and if you're having a lot of trouble getting past the game's looks -- and that's worth doing, if you like your games to make you think and keep you on your toes -- then a setup like this one may help.
The amazing thing is that the environment looks so good despite the graphical limitations of the time. A stellar job was done folding the starship interiors into the boxy limitations of the day's graphical tech.
Still, if you can look behind the graphics and can somehow relate to the time back then, it was also so different from Half-Life in many regards.
Possible explanations include: 1) there is not as much success going on at Irrational as implied; 2) Ken Levine is just really attached to the name, and so wouldn't let it continue in present form while he leaves to do something else under a new name; 3) ...?
He made it sound like (and maybe this is true) "Congratulations team, you've done well, we're successful, now we're going to lay you all off. Not even find you somewhere else in Take Two." The Game industry sounds a little harsh and even self-destructive right now.
But I'm sure something was left out...
Besides financial support, the staff will have access to the studio for a period of time to say their goodbyes and put together their portfolios. Other Take-Two studios will be on hand to discuss opportunities within the company, and we’ll be hosting a recruiting day where we’ll be giving 3rd party studios and publishers a chance to hold interviews with departing Irrational staff.
It definitely reads like the employees have to re-apply and re-interview for any of those jobs. That's different from what you'd expect at companies in virtually any other industry or category of tech.
What? That's fairly routine in many industries...
I feel bad about linking to about.com, but it emphasizes how common it is.
Depends a lot on who is going to be left after they leave.
They may just want to kill the name while it is known for positive things rather than letting it languish.
Best of luck to the new team, looking forward to awesome stuff...
My guess, though, is that he didn't actually make that choice and was told by Take-Two to downsize, and this is him pretending it was a voluntary "creative" decision, like politicians who want to spend more time with their family.
That would probably be the nicest thing anyone has ever said about him. He is not well liked by past co-workers or employees.
A game that would be considered incomplete ~5 years ago is launched. You then ship up to a dozen DLC packages that add incremental content so that the complete purchase price of the game and all DLC is 2 to 3x the original retail price.
That's how they justify the huge teams. I know it's very different market, but just look at how Turbine went from 500-600 employees to under 200. Look at how Zynga had 2,000 employees. Now wonder what, exactly, all those people were doing?
It's compelling, especially to veterans like Levine, to return a team that's no more than 3 or 4 dozen people and create / ship a complete product that they have a high amount of creative control over. Reminiscent of Double Fine / Psychonauts.
At least I hope that is what they are doing.
Irrational Games will enter the waters of baptism, and a new studio will be born. An infinite number of Irrational Games studios are opening and closing at this moment, like lighthouses on an ever-expanding ocean. The only difference between past and present is semantics.
If what I'm saying sounds crazy, you owe it to yourself to play Bioshock Infinite. It's without a doubt one of the most beautiful and surreal games ever created.
Let's call those two games, "pi" and "e" while we are at it.
Infinite finishes with an Everstory, with infinite branches... yet in all of those branches, there is only one of two possible outcomes for the main character in each of them.
I really need to proofread my comments more.
> we will focus exclusively on content delivered digitally.
And bioshock infinite was threading on too safe ground. I really hope that his new studio will have bursts of creativity and success and the left out employees find better jobs soon.
I dunno about that -- what always seemed to me to be holding the Bioshocks back from greatness was that they were designed around a gameplay mechanism (FPS running and gunning) that clashed pretty severely with the type of game the designers seemed to want badly to make, namely an interactive story. The result was a sort of schizophrenia: Irrational would build these incredible environments and characters, and then stick them in the exact type of game where players couldn't linger over and savor them. They were just a blur that would flash past your gunsights as you killed people.
This problem was so evident in the Bioshock games that they led people smarter than I am to eventually coin a term for it: "ludonarrative dissonance" (see http://clicknothing.typepad.com/click_nothing/2007/10/ludona...). Which basically means a game whose story is trying to do one thing while its mechanics are trying to do something completely different, pulling the player in incompatible directions. That -- and Irrational's seeming lack of interest in reconciling the two elements in their work -- always seemed like a bigger factor holding them back from greatness than their relative lack of moddability, at least to me.
One of my favorite articles about Infinite is this one where it's argued that it's the worst game of the year: http://tevisthompson.com/on-videogame-reviews/
And the dissonance was maybe due to publisher interference the same way the main story line and every single sentece connected to said line in Dishonored are terrible and naive and overly simplistic. And the moment you get to listen, watch and read something that you are not under obligation to do - the quality jumps up next to 0.5 Torments.
Probably there were moment when it was - add more action, too much horror more action, too much talking more auto regen health.
Also, the crunch time was horrendous for much of the project. The project churned through people who left bitter and burned out.
I'm not sayin' BI is a bad game, but this isn't the story of a well-managed studio.
> To make narrative-driven games for the core gamer that are highly replayable.
During the last years there have been many indie games with a strong narrative focus. His mission statement together with his actions (small team, digital distribution) fits more with that than free to play mobile only games.
It seems to me as though Ken Levine saw that he was making very narrative focused games (Bioshock Infinite in particular) with a side dish of shooting for that AAA appeal. Maybe he also saw all the indie developers making games with narrative focus who were not afraid of getting rid of such AAA trappings altogether because with digital distribution everyone can reach their niche and make sustainable amounts of money.
Ken Levine is obviously not going all the way, but he is going pretty far with the downscaling (and it’s always worth to talk about whether he has gone the right way). AAA development gets you some nice things (God Only Knows sung by a barbershop quartet, for example) but comes with its own set of restrictions (you have to go for mass appeal) and when you want to make something smaller and more focused that doesn’t appeal to everyone then there’s lots to be said for getting smaller.
(During the last few years we have seen the middle fall out from game development, with mid-sized developers and publishers going under. It seems that this gap is now filled from the bottom and the top, with former AAA developers scaling down and successful indie developers – Jonathan Blow with The Witness, Mike Bithell with Volume, … – scaling up.)
There is whole diverse and crazy world below AAA titles. Some of those games are awful, but some are awesome. There is lots that can be done. And lots of cool things that are done.
Why would you think of Candy Crush first? Especially the mention of the narrative focus makes me think of many excellent (and successful!) recent indie titles that also had a strong narrative focus: Stanley Parable, Gone Home, Kentucky Route Zero, …
It seems to me that Ken Levine has something like that in mind plus his own game mechanics twist (the highly repayable part, whatever that means), certainly not some free to play mobile only bullshit.
Less cynically, you mean taking the Dwarf Fortress adventure-mode route? Good luck... But even an RPG with multiple interesting endings (where choices you make actually affect things later in non-obvious ways) would fit the distinction of highly replayable.
I think every game does that these days. There usually are fairly official numbers for sales of video games as it is a highly mature market. What are the real sales numbers?
Here they are:
I suspect they budget was insanely high for the game though, so they needed not just a blockbuster, but a massive one.
Bioshock dropped pretty quick, IIRC. I remember picking it up on special during the Steam Summer Sale for $19.99. It was less than six months after release.
However, I got Battlefield 4 for $20 in less than a month after release, so it's possible that's the new norm.
On the PS3, it wasn't worth the bandwidth it took to download. The shoddy textures they used for the console version made it look half a step removed from a late gen PS 2 game. The gameplay and the recycled story didn't help either. It felt like they just took extra character models and levels from the previous BioShock games development, reskinned them and sold it to us it back to us with a plot that consisted of cross out the names of characters and places and writing in new ones.
I hope they do better in the future, but playing that game has dispositioned me to thinking it may not be a bad time to shut down and start over: they need some new ideas.
Half Life 1, which was a huge blockbuster in 1998 took a year to sell 1/4 the number of coppies BioShock Infinite sold in 5 months.
I'm just impressed that premier games have kept prices so high, as the price of most data products becomes so low. 'Social' games are comodity, but blockbusters are not.
Developers can make a 'social' game with some 2D art, but blockbusters actually require the input of hundreds of people.
It's sad to see the name being retired, but it's better than seeing the name ruined by a flop or diluted by endless sequels.
The announcement is pretty opaque---I expect the rumor mill to churn for awhile.
I just finished Bioshock Infinite a couple weeks ago, and I was left feeling weirdly angry. It was a great story, shoehorned into a mediocre shooter. All of the problems with Bioshock were amplified. Which was necessary; in order to justify that kind of budget, they needed to make a game that would sell to the lowest-common-denominator.
Hating on Bioshock Infinite is a bit like criticizing a Hollywood blockbuster for being dumbed-down. What did dummies like me expect?
“To make narrative-driven games for the core gamer that are highly replayable.” This is what fans of System Shock and Bioshock have been clamoring for. Set free of Take Two's blockbuster expectations, Levin will be free to deliver it.
First of all Irrational Games is (wholly) owned by Take Two. So only Take Two decides what happens with the studio. From the statement i guess they decided that is "less bad" if Levine steps up and says they "wind down" instead of Take 2 announcing they close the studio. Basically PR damage.
Regarding the reason....looking at their output it seems they had it coming. After the original Bioshock (2007) they took 6 (!!) years to release Infinite (and even that was codeveloped with 2K Australia). The financial losses must have been pretty big.
as soon as I get to a PC I will post sources
I just read that as; we are gonna make an even more narative based System Shock 2 equivalent.
Kickstarter success is reputation/prototype based. You need either a great reputation or a great prototype. It's essentially a form of grant. Many of the games which succeed will ultimately fail in the "fun" sense. Yet I'm convinced that, as broken as AAA publishers are, throwing dice gives me a better chance.
Check out board games. There's big innovation and it's entirely game mechanic based.
I wonder what happened. It sounds like Bioshock Infinite didn't bring in the cash they thought it would? Reminds me of Ensemble Studios closing after AoE:III and Halo Wars.
Outside of computer/console 'game' industry, games are rules. You distinguish two games by their rules. "How do you play it ?" is the question you need to ask. In video 'game' world, "game" has become an umbrella term for:
- actual multiplayer games,
- playgrounds/toys (Minecraft, MMO...)
Basically any interactive software that is used for entertainment is called a game these days. I guess vlc also meets the criteria, after all you can use it to watch porn.
There's a parallel between Test Driven Development and board games. Today, computer games have become so complex and have so many moving parts that they have more in common with simulations than board games they largely came from. This is because you no longer understand all or even most of its RULES. Computer is kind enough to calculate everything for you. You don't know why a fireball deals 23 damage or why your city suffers an epidemic. It could be because it's scripted that way, because something gives it a +20% bonus (added before or after X ? Is it actually +20% or * 1.2 ? Wording is ofter ambiguous). The player is only expected to move around in a world and bump into things. Often the quickest way to learn a game is to try it. Learn the way children learn languages - not by memorizing grammar rules, but by practicing.
Why TDD ? It is often claimed that in Test Driven Development, you write tests before you write code. THEN you write minimal code to pass those tests. By definition, you have practically 100% test coverage, which can be nice.
By definition, you know all the rules of a board game. The rules tell you how and when move pieces, cards, tokens around.
Modding and house rules is rampant in board game world. It happens even accidentally - when you fail to learn rules correctly, and later decide it's more fun that way. It's absolutel fine to play with "wrong" rules if all players agree to do it beforehand. Board game players often make their expansions, variants, prettier game art.
Fun fact: "Sacrifice" spell in Heroes of Might and Magic 3 is widely considered one of worst spells in the game. Sacrifice a unit to resurrect another one ? Load game, I wasted so many resources on level5 guild!!
Unless you know the formula, which is NOT specified in the game. Then you can produce a table like this:
With a modest investment (3 skill points and growth bonus building), you can sacrifice 1 week worth of imps(level1) to resurrect 2 weeks worth of Archdevils (level7). Assuming you get the guild, it all comes early enough to be relevant in a cutthroat multiplayer game. And it gets better and better as the game progresses, because unlike with most spells the effect doesn't increase in a logarithmic fashion (experience levels), but with troop counts.
This is not good. This is devastating. The huge discrepancy between the perceived value of the spell and its actual value is due to bad documentation and hidden rules. It is extremely common for computer games to contain hidden, unexplained rules.
> By definition, you know all the rules of a board game. The rules tell you how and when move pieces, cards, tokens around.
Your comment doesn't mention table-top RPGs, and it seems to me that it's them, rather than simulations, that modern computer games are closer to. In a table-top RPG the players are often not expected to know or understand all the rules and "moving parts" of the game. The GM is also free to decide to bend the rules if he or she thinks it would enhance the experience for the players.
A few are still doing games now, in the indie scene. The most interesting of those, imo, is Richard Evans (formerly of Black & White and The Sims 3), who's building an interactive-fiction platform on top of a novel logic-based social-simulation engine (http://www.versu.com). But a lot are doing entirely different things: some moved to "regular" programming jobs, ranging from simulation programming for big engineering companies or defense contractors (somewhat related) to almost anything at all that provides a regular job. A number have gone back to grad school. A few have become freelance consultants.
Imo it really stunts the growth of game AI as well, because there are remarkably few experienced senior AI devs who actually stay in the field, long enough to develop a sense of how to architect systems, how to interface with designers, and actually improve things over time.
Each step away from games has lead to more sane company environments and higher proportional incentives. The reality of games is that programmers are underpaid in comparison to other industries. If a web programmer with X years of experience can command a salary of Y you can expect a game programmer with X years of experience to command approximately 2/3 * Y. Exact same geographical region.
In return for being underpaid you will encounter far more 'crunch time'. I was fortunate in prioritizing this concern in the companies I chose to work for, but many of my friends in the AAA games industry have become completely used to working 10 hour days, 6 days a week, for 3-6 months straight at a time. Every. Single. Fucking. Year. Around July is when it begins; holiday launch schedules being the driving factor.
On top of this, as illustrated in TFA, even if your game is a huge success you have extremely little job security. Saying the industry is entirely hit driven is actually missing the point; the issue is most AAA game studios are essentially huge, expensive, single-focus (and usually single-project) consultancies. Even if you have a success the lions share of the returns is going to go to the publisher and you are immediately back to pitching to the publisher to keep the lights turned on.
So why is it this way? IMHO there are two main reasons.
The first is because AAA titles are ridiculously expensive now. Internally financing one is completely out of the capabilities of any studio at this point. In the past it was possible to save from a good success and self-finance the next: no way now. In the past it was possible for a rag-tag group of passionate developers to have a AAA breakaway success. Impossible now.
The second is because it's something lots of people dream of doing. The only sane reason to work in games is the exact reason most people who work in games continue to work in games: they love it. They love games. They love the tech in games (which, compared to the CRUD factories the vast majority of programming gigs outside of games, is really fucking awesome). They are following the advice of 'do what you love' and paying for it. Because there are TONS of other people that also love it.
It's stunning the amount of turnover in the games industry. 5 years of experience is senior. 10 years is old school veteran. Because people burn out and leave. Because there is an endless supply of less experienced cheaper developers to fill in the gaps.
The answer 'go indie' is just naive. For every indie success there are ungodly numbers of complete failures and you can no longer twist the knobs of a publisher's titan AAA marketing machine to stack the deck. Going indie is doing a startup which is impossible to fund and which has no hope or aspirations of Facebook level success. It is a giant success in the indie game world is to be successful enough to just support yourselves! It is something to be done by those with the love and passion for doing so for the reason of having the love and passion of doing so. It's never a rational decision from a purely financial point of view. I have the utmost respect and admiration for those who follow this path precisely for this reason.
The games industry is an entertainment industry. It's more apt to compare game developers to musicians or actors or writers in the way the economics operate. It's not a business; it never will be. And honestly, I do miss it.
How do you think is it possible that one day Web Development (generally speaking, industry of Internet startups) may fall into the same gap? There are a lot of guys around who enjoy Web programming very much. Fortunately, we have a lot of job opportunities, and work conditions look good. But I fear how long will it be.