For video game designers who want to derive from this feature, be careful. Throwing in a river of lava or two alone isn't going to cut it. The player needs meaningful ways to interact with space and move both himself and others around if the terrain is really going to matter.

If the player can't push orcs into the lava, the lava isn't going to add much fun. Even then, just adding in some physics motion is not enough -- it needs to have recognized and easily understandable gameplay effects.

If pushing follows actual physics, it's going to be hard for the player to know how it works -- instead, make pushing a gameplay system so the player always knows he can push small enemies a certain distance, and larger enemies a different distance. D&D powers are all universal, even when they don't make sense.

For example, a rules question on a forum asked if Gelatinous Cubes could be knocked prone, and what that would look like. The answer: just treat them the same as anyone else who is prone, even if the character can't see a difference -- gameplay system trumps reality.

Even with intuitive player inputs into the system, the system can become tedious if all the player can do is knock things into lava. Just like with monsters, there must be enough depth in the trap/hazard system to populate multiple encounters. Movement mechanics should be as varied and well paced as combat mechanics, giving players multiple different ways to use these new abilities.

Allow allies to knock each other out of the way of incoming boulders. Set up enemy abilities that buff each other within a limited range so the player is rewarded if he can separate them. Have the combat system provide an interesting response to prone enemies, and double reward knocking enemies into each other. Fit everything together into a clear systems design, then build enough content to support it from start to finish.

Team Builds

Synergy is not a new idea, but it's well executed in 4E. Rogues like to position themselves around enemies, and Warlords and other classes like to move allies around. Multiple classes (Warlord, Cleric, Fighter, Paladin) have multiple powers that directly aid and abet their allies.

Many class powers react to states that other classes are good at creating. Already, people are not only posting character builds on the forums, but also posting team builds, utilizing synergies from multiple characters working together.

One key concept that 4E uses to make this work is setting up most team abilities so that they do more than just help others. Players like to do things. Helping others works better as a side effect of an attack than its own action. I enjoy playing a healer in an MMO, and spending all my time concentrating on the micro-spreadsheet of allies' health, but I'm the exception rather than the rule.

Adapting this rule from a turn-based pen-and-paper game to a real-time electronic game requires some simplification -- picking an ally to target in the middle of an attack isn't exactly streamlined. But that's may have a simple fix -- target the nearest ally or build the game around a limited number of wingmen, for example.

And don't forget to carefully parcel out the abilities so that each class provides something that the other classes can't get by themselves. If the Rogue could position himself perfectly by himself, he wouldn't need the Warlord's help.

Hold off on some good abilities to make allies really matter. Single-player games get less benefit from this, but even single-player games often include support NPCs. And for MMOs or cooperative gameplay, anything that encourages players to support each other is critical.