What We Have Lost

Planescape: Torment was released ten years ago yesterday. It is a masterwork and still unsurpassed in many ways. It remains one of my favourite games of all time and it is only by technicality that it's not dominating every "Best Of 00s" list. While I don't plan on doing a Best of 00s list of my own (too many choices, too busy with other endeavours), I have been thinking about how games have changed in the past decade. The transition from text to voice acted dialog in games, especially RPGs, is one interesting change to consider.

We're wrapping most of the recording for dialog in DeathSpank right now, which is probably why I've been thinking about this. DeathSpank has about 7000 recorded lines of dialog by Ron's count. That seems like a lot, especially considering the effort I've seen the folks closest to that aspect of the game putting in. But then one hears Dragon Age shipped with over 80,000 lines of recorded dialog. I am horrified thinking about the logistical scope that much dialog necessitates.

Planescape: Torment had about the same number of lines as Dragon Age, but very few of them were recorded (similar to Fallout). And from what I remember, most of the recorded lines are more like barks rather than conversational dialog. I can't help but wonder how the lack of that constraint affected the development of Planescape: Torment's writing.

In game development, dialog recording is almost always done last. There's too much in flux for dialog recording to take place early. But with 80,000 lines, that's a tremendous amount of content that needs to be cast, recorded, edited, implemented and tested. If the game is sim-shipping worldwide with foreign language recording, all of that work must be done in time to allow translation and testing as well. Relative to the development time of the entire project, the time writing can be flexible is shorter than it was a decade ago.

With text only dialog, written content can be changing right up to the point where it needs to go off for localization. And even then, sending a few pages with changes that require additional translation is much easier than rerecording lines in a half-dozen languages.

I can't help but think having that additional freedom to make changes and keep improving and polishing the script contributed to the excellence we saw in games of the late 90s/early 00s. Aside from Planescape: Torment, Baldur's Gate II, Fallout 1/2, Arcanum and others still stand out for their writing excellence.

As much as poor voice can pull one out of an otherwise excellent game, a good voice actor truly can bring a character to life in ways otherwise impossible. Once the recorded dialog started going into DeathSpank, I couldn't help but start listening to conversations I had read many times before, simply because of how much more engaging it is with a good performance (and hopefully you'll agree). I don't think we're going to see voice acting in games go away and as long as we can get good performances, I think this is a good thing. But I can't help but think we've lost something along the way.

While the delivery is different, how the dialog is presented remains the same for most games as it was a decade ago. Perhaps Mass Effect did take the right tack in creating a system for interacting with dialog that fits more naturally with voice acted dialog conversations. If we're never going to surpass the quality of Planescape: Torment's writing on its own terms, perhaps we should find ways to best exploit the advantages of voice acted dialog, rather that just having an actor read the same text the player would have done in the past.

While I appreciate what a good voice performance can bring to writing in games, it's maybe unfortunate all games opt for voice acting by default. While I can't see this trend reversing, I would love to see more experimentation in how players interact with dialog in games.

In an unrelated anniversary, I've also been blogging for just over one year, with 70 posts made starting December 9th, 2008. My first post was a reflection on something Steve wrote, and I think it really exemplifies what I've found so excellent about this endeavour. Being engaged in the conversation has challenged me to not only pay more attention to what other have to say, but to be more reflective about my own thoughts and opinions. So to everyone that has given me something to think about, or taken the time to consider something I've written, my sincere thanks.

Window Dressing for Free

I've been playing a bit of Arkham Asylum recently, behind the main swell as I opted for the PC version. To add to the echo, yes, it's bloody fantastic. But more importantly, it's quite possibly the purest example of how exactly to take a license and make a game that captures the license's spirit exactly.

Many deride licensed games, and fairly so. But the condemnation is often categorical, both on the part of developers/publishers who churn out uninspired rubbish for the lowest cost possible, and audiences who assume anything licensed is terrible until proven wrong. This leads to self-fulfilling prophecy and the licensed ghetto continues.

There are still plenty of excellent licensed games (even going back decades). Their excellence is largely due to two characteristics- they allow players to inhabit an interesting world and craft their own story. The latter is doubly important. So often, license adaptations are merely translations of some existing story (usually, a movie). This basically robs the player of almost everything interesting about games: agency, exploration of the new space, suspense, etc. Recreating the footsteps of a big screen protagonist is the last thing most players want.

I don't think it's any coincidence that some of the best licensed video games are licenses of other game properties. D&D, Vampire: The Masquerade, Warhammer 40K; these properties are built expressly for allowing players to craft their own stories. And this includes the best licensed game ever made, and one of the greatest games ever. Planescape: Torment worked in part because Chris Avellone had all the richness of the setting to draw upon. But Black Isle didn't just recreate an existing Planescape adventure (and there were plenty of excellent candidates), they created a new journey through Sigil and the planes.

For those familiar with the setting, we were getting to see familiar things through a whole new lens. The first cranium rat, the omnipresent razorvine, the rebus speech of the Dabus; all of these evoked both familiarity and appreciation. For those new to the setting, they were able to marvel at the depth and diversity of content.

This is where Arkham Asylum succeeds as well. For me at least, Batman has always been about the villains (and part of the triumph of Burton and Nolan's Batman films was finding ways to make Batman himself interesting). Rocksteady fully embraced this and found ways both overt and subtle to make the villains the primary backdrop of the game. Mark Hamil's Joker, with his constant boasts, taunts and straight madness, is fantastic. More subtly, small moments involving the other villains, even ones not featured in the game, give Arkham more substance and atmosphere. Seeing Catwoman's equipment in a glass case or seeing Hush's name on an X-ray are great nods for those more familiar with the Batman setting.

The Batman of Arkham Asylum is fantastically boring, but as a cipher for the player, this actually works rather well. The gameplay focus on the experience of Batman, rather than the character. Between the plethora of gadgets, relatively single but incredibly fluid combat, reasonable use of stealth and the extreme vulnerability to firearms, Arkham Asylum nails the feeling of Batman. And it's far more rewarding than watching some 3D model in a cutscene tell you what it's like to be Batman.

One of the hardest parts of creating a new IP where setting and character are central is simply getting people to care about your world at all. Creators very, very often fall dangerously in love with their own creations. It's the quintessential DM who has written tomes about his world and carries on about "the impacts of the Third War of Succession due to Prince Bakolar's betrayal of the Treaty of Darkfall," oblivious to his players being bored to tears.

Game designers working on new IP can easily fall into the same trap and avoiding this requires great discipline and restraint. Probably the most telling acknowledgment of this is when Ken Levine said "Nobody wants to reading your fucking design document." He wasn't condemning the work that should be put into background worldbuilding, but rather that what actually ends up in the game should be relevant to the player. The rest is a latticework the team can use in creating environments and interactions that are consistent and rich. Bioshock was commendable in its restraint. I didn't have to listen to three minutes of audio diary explaining how Andrew Ryan made his money, because that doesn't really matter. But a 15 second anecdote about how he'd rather burn down a forest he owns rather than grant it to the government to make public? Telling and pointed.

At the end of the day, a great deal of licensed games are still drek. The economics of releasing a game in coordination with a major movie more or less encourage cutting corners whenever possible. But there are lots of excellent licenses being wasted because developers are trying to hard to recreate an experience, rather than allowing the player to have their own. My sincere gratitude to Rocksteady for delivering the best use of a major license and demonstrating it doesn't have to be pandering garbage. A few more successes like this and maybe we'll see a few more minds opening toward some of the truly excellent licenses out there, just waiting for treatment like Arkham Asylum.