Premium Tutorials

Balancing a game is a tricky process that involves many different systems, including enemy NPCs. Getting it right for players of different skill levels even more so! The sniper mission “All Ghillied Up” is one of the most famous and memorable of in MODERN WARFARE , and does a great job of illustrating how to handle difficulty levels. The mission includes a combination of linear sections with specifically spawned enemies, as well as open sandbox areas with difficulty adjusted enemies.

How would you go about designing and architecting an AI to play MAGICKA? Specifically, what do you need to build a bot capable of replacing human co-op partners if necessary, accompanying the player through single player missions. The game presents many challenges, including a wide variety of spells that result from combinations of basic elements, not forgetting the challenges of finding good positions and balancing attack with defense.

Over the past decade, blend trees have incrementally established themselves as the de-facto technique for implementing character animation. Since these blend trees are becoming more complex, implementing them efficiently is becoming a priority, allowing games to offload the bulk of the computation to separate dedicated processors like the SPU.

The best way to learn is from real examples! In this review, you'll hear about four different work-in-progress indie games, and in each case the AI hasn't been developed (much) yet. You'll learn how to approach the design, what factors to consider, as well as how to architect the system and lay out the implementation plan. By going through this process for iOS games, web-games and PC games, you'll be able to apply this to your own games too.

Character animation remains one of the biggest challenges for creating believable NPC behavior, and it's arguably the fastest moving fields of game development. Staying in touch with recent games is one of the best ways to stay on top of trends! What's everyone doing and which patterns are most common? What techniques should you implement in your own games? Where are things going in the near future?

With single player campaigns of first-person shooters expected to last 10h, it's important for the enemy soldiers to exhibit diverse behaviors. You don't want your players getting bored! When there are many enemies on screen, and multiple waves descend on the player's position, this can particularly be a problem. How do you add more variety and diversity to the AI's combat behaviors?

As Jay Wilbur once said, "Level Design is where the rubber hits the road." No matter how great your programming or how well designed your NPC behaviors, a level designer can make your work look like the most terrible or amazing AI ever built!

An AI's decisions are only as good as the information it gets, so for things like tactical and strategic reasoning influence maps are a critical component. They help not only at the individual level for picking positions, but also deciding where to attack or build bases. In fact, it's safe to say the more you use influence maps at all levels of your game AI the better it will be overall!

Increasingly, emergence is not enough to build compelling gameplay experiences. You need some kind of monitoring of the player's behavior, allowing you to controlling the game's flow towards states that you know to be fun. That's where AI Directors come in, providing a layer of adaptive logic that improves upon near-random experiences.

While many games use procedural generation of geometry to reduce the workload of artists, more developers are experimenting with procedural levels for gameplay reasons. This is something that classic games on older platforms did regularly, however as Soren Johnson (Lead AI & Gameplay on CIVILIZATION 4) said recently: “I'm afraid it's becoming a bit of a lost art.”