Human-like Bots in Unreal Tournament
Active from 2008 - 2012
Since 2008 the Botprize competition has posed a challenge to bot designers: Can you make a bot that will fool a human into thinking it is human? The competition is essentially a Turing Test for bots: Given matches containing both human players and judges, will the judges think the bots are humans?

Unreal Tournament 2004 is a popular commercial first-person shooter video game played online by humans. However, the game also features computer controlled bots for humans to fight. Players participate in deathmatches, in which they collect weapons and attempt to kill each other to maximize their scores. The game involves exploration of complex 3D levels to find items and enemies, chaotic combat against multiple opponents, and reasoning about the best strategy at any given point in the game. It is hard enough for bots to perform well in this environment, but even more difficult for them to look like they are controlled by humans. The most recent version of the competition is actually a judging game in which all players, including bots, can use an in-game gun to judge opponents as bots or humans.

In the 2012 competition, our bot UT^2 passed the 50% humanness threshold to win the grand prize! Earlier that same year, we also won the Human-like Bot Competition, which is essentially the same competition, but under a different name.

The following are links to several news articles and other media regarding BotPrize:
Risto Miikkulainen Faculty risto [at] cs utexas edu
Jacob Schrum Ph.D. Alumni schrum2 [at] southwestern edu
Igor V. Karpov Ph.D. Student ikarpov [at] gmail com
Believable Bot Navigation via Playback of Human Traces 2012
Igor V. Karpov, Jacob Schrum, Risto Miikkulainen, In Believable Bots, Philip F. Hingston (Eds.), pp. 151--170 2012. Springer Berlin Heidelberg.
Humanlike Combat Behavior via Multiobjective Neuroevolution 2012
Jacob Schrum, Igor V. Karpov and Risto Miikkulainen, In Believable Bots, Philip F. Hingston (Eds.), pp. 119--150 2012. Springer Berlin Heidelberg.
UT^2: Human-like Behavior via Neuroevolution of Combat Behavior and Replay of Human Traces 2011
Jacob Schrum, Igor V. Karpov and Risto Miikkulainen, In Proceedings of the IEEE Conference on Computational Intelligence and Games (CIG 2011), pp. 329--336, Seoul, South Korea, September 2011. IEEE.
UT^2: Winning Botprize 2012 Entry The Botprize Competition is an annual competition to program bots that appear human-l... 2012