diff --git a/events.xml b/events.xml
index 9c0e1af..71ad9f5 100755
--- a/events.xml
+++ b/events.xml
@@ -1,7 +1,30 @@
 <eventdefs>
 <!-- Spring 2004 -->
 
-  <eventitem date="2004-07-17" time="11:30 AM" room="RCH 308"
+<eventitem date="2004-07-27" time="4:30 PM" room="MC 2065"
+title="Game Complexity Theorists Ponder, by Jonathan Buss">
+<short>Attention AI buffs: Game Complexity presentation</short>
+<abstract>
+<p>
+Why are some games hard to play well?  The study of computational
+complexity gives one answer: the games encode long computations.</p>
+
+<p>Any computation can be interpreted as an abstract game.  Playing the
+game perfectly requires performing the computation.  Remarkably, some
+natural games can encode these abstract games and thus simulate
+general computations.  The more complex the game, the more complex the
+computations it can encode; games that can encode intractable problems
+are themselves intractable.</p>
+
+<p>
+I will describe how games can encode computations, and discuss some
+examples of both provably hard games (checkers, chess, go, etc.) and
+games that are believed to be hard (hex, jigsaw puzzles, etc.).
+</p>
+</abstract>
+</eventitem>  
+
+<eventitem date="2004-07-17" time="11:30 AM" room="RCH 308"
    title="Case Modding Workshop!">
    <short>Come and learn how to make your computer 1337!</short>
    <abstract>
@@ -80,6 +103,8 @@ welcome to attend.
     </p>
    </abstract>
   </eventitem>
+  
+
   <eventitem date="2004-05-26" time="5:30 PM"
    room="DC 1350" title="Computing's Next Great Empires: The True Future of Software">
    <short>A talk by Larry Smith</short>