GDC '08: Epic ultrakills UT3

SAN FRANCISCO--The tail end of 2007 offered gamers a veritable ocean of first-person shooters to frolic in, including Valve's The Orange Box, Crytek's Crysis, and 2K Boston/Australia's BioShock. While it would have in all likelihood dominated the competition in most any other year, one title to barely surface among the flotsam and jetsam was Gears of War developer Epic Games' Unreal Tournament 3.

In a Game Developers Conference session earlier today titled "ULTRAKILL! An Unreal Tournament III Postmortem," Epic Games president Mike Capps, alongside UT3 senior producer Jeff Morris, addressed the game's fortunes in what they called the "FPS Christmas of All Time."

Kicking the presentation off, Capps first laid out what goals Epic had for the shooter, which began its life as a proof-of-concept tech demonstration to gain licensees for the company's highly popular middleware service Unreal Engine 3. For new features, the team was most excited about the new game type warfare, which had a heavy focus on vehicles and open environments. Vehicles, which were first introduced to the series with Unreal Tournament 2004, were also expanded with the project, with Epic upping the included number from six to nine. Epic also had a mind to create a fully fleshed-out single-player campaign for the first time, as well as place a heavier emphasis on character customization. Lastly, the publisher wanted to maintain its mod-friendly position, as well as extend the ability to make high-quality mods to consoles.

So how did they fare? As is the case for most games, there were both marked successes and disappointing failures. Addressing the successes, Capps said that because the team knows the first-person shooter genre so well, Epic was able to deliver UT's best fast-paced, signature gameplay ever. Several factors aided in this, including an extensive focus on playtesting, easy access to a wealth of assets in their content pipeline, and a focus on first getting the gameplay right before fleshing out the visuals through a process Capps called gameplay previsualization.

Capps was also pleased by the unified look and feel of the game, saying that for the first time, the game didn't have a hodgepodge aesthetic of disparate elements and visual designs. Mod support also turned out well, according to Capps, who said that Sony--and not so much Microsoft--has been extremely receptive to the idea of letting users create content and easily share it with others. Capps beamed that Epic was the first to offer this kind of functionality to users and said it turned out to be a great promotional tool.

As for what didn't turn out so well, the jovial Capps turned the presentation over to a not entirely morose Morris. One of the biggest problems with UT3, said Morris, was the game's user interface. Originally the UI design was driven by an artistic aesthetic. Though the UI had an appealing look and would have worked well on consoles, the team decided that it would have been nightmarish to navigate the menus on a PC. Consequently, the team did a last-minute reboot to implement a system that worked better on PCs, which didn't allow for enough time for iteration and polish.

The second issue with UT3 was the character customization, or lack thereof, that made its way to the final product. Here, the problem was that despite an extremely complicated and involved process to get the different textures for characters in the game, the varied models didn't add the diversity Epic was looking for and paled in comparison to Realtime Worlds APB, referencing Dave Jones' session from yesterday. The gameplay having the fast-paced nature that it does, Epic also found that players really didn't care what the customization looked like since the game moved so quickly.

The game's single-player campaign also didn't quite pan out to expectations, said Morris. As with character customization, Epic felt its campaign was coarse in comparison, with Morris saying, "It was good, but it wasn't BioShock." Morris also noted that though the game had high-quality cinematics--"certainly better than Gears," Capps piped in--they took a lot of time to get right, and their spacing was too far apart to have an indelible impact on the player.

Lastly, Morris called out problems on the production side of things, namely scheduling and promotion. Temporary console exclusivity on the PlayStation 3, as well as stringing out the releases across regions, turned out to be a major headache for Epic. Not only was this because it poorly maximized marketing efforts, but it was also because it wasn't conducive to attaining the critical mass of players that is essential for a multiplayer-oriented game. It also didn't help, Capps noted, that the game walked into the "FPS Christmas of All Time" and that there are only so many ways to slice the FPS pie.

Prefacing by saying there were no qualms with Midway's public relations, Morris also said it was extremely difficult to maintain player interest having announced the game in 2005. Announcing the game so early was a product of being in the engine licensing business, as the game was already well known because it had been sent to various developers as a way to sell the Unreal Engine 3. Such a long period of promotion also led to several lacking gameplay features that were heavily promoted but were ultimately cut from the final product. Lastly, Morris said that it wasn't exactly the best of business moves to change the name of the product well after the game was officially announced, and the early announcement also required a premature lockdown of the game's overall look.

Wrapping up, Capps said that despite the challenges, the game both sold and reviewed well. He primarily attributed this to the fact that Epic knows the genre extremely well and that it has such a well-established content pipeline. Adding a cautionary tale, Capps said that developers need to not be lured by opportunities that would drive marketing before the product is ready to be shown.