Where the principles come from
Almost every animation workshop opens with the same reference: The Illusion of Life: Disney Animation, published in 1981 by Frank Thomas and Ollie Johnston. Its twelve principles, including squash and stretch, anticipation, staging, follow-through, slow in and slow out, arcs and timing, were written for hand-drawn film. They became the common vocabulary of the craft, and you will hear them in critique sessions whether you work in Blender, Maya or directly inside a game engine.
Richard Williams' The Animator's Survival Kit sits next to it on most desks. Williams is especially useful for walks and runs, and for the idea that timing (how many frames) and spacing (where the drawings sit within those frames) are separate decisions. In 3D, spacing lives in the graph editor, and learning to read curves is one of the first concrete skills a workshop should give you.
The big change: responsiveness
In film, a character can take eight frames to crouch before a jump, and the audience enjoys the build-up. In a game, eight frames of anticipation at 60 frames per second is a delay of more than a tenth of a second between the button press and the take-off. Players feel that as sluggish controls. So game animators learn to move anticipation around rather than delete it.
Common workshop exercises include:
- Front-loading the action. The jump starts on frame one with a strong push-off pose; the crouch is implied by a compressed first key rather than shown in full.
- Putting anticipation in the idle. A fighting stance already reads as loaded, so the attack can fire instantly.
- Spending frames on recovery. Follow-through and settle happen after the gameplay-relevant moment, where the player can often cancel them with a new input.
Workshop drill: animate the same sword swing twice, once at film pacing and once with a two-frame wind-up and a long, cancellable recovery. Play both in an engine with a button mapped to each. You will feel the difference in seconds.
Squash and stretch on a rig
Stylised games such as Celeste or Hollow Knight show how much life a few squashed frames add to a landing. In 3D, squash and stretch depends on the rig: many production rigs support non-uniform scale on the spine or limbs, while realistic characters fake it with exaggerated poses, overlapping secondary motion and camera shake. Good workshops are honest about this. You learn when real deformation is appropriate and when to sell impact through timing and contrast instead.
Timing for loops and cycles
Cycles are the bread and butter of game animation: idles, walks, runs, swims. A cycle must loop without a visible pop, which means the first and last poses match and the curves have matching tangents. Workshop leads usually ask you to check curves with cycle infinity turned on in Maya's graph editor, or the Cycles modifier in Blender, so you can see the seam.
Timing also connects to gameplay numbers. A run cycle's stride length must match the character's movement speed, or the feet will slide. That is your first lesson in collaboration: the animator and the gameplay programmer agree on speeds, and the animation either drives movement (root motion) or follows it (in-place animation). Both approaches are valid; workshops should let you try each.
Arcs, weight and the camera
Game cameras orbit, zoom and shake. A pose that reads beautifully from the side can collapse from behind the shoulder. Practise checking key poses from the gameplay camera, not only from a perfect side view. Silhouette tests, where you fill the character in solid black, remain one of the fastest ways to judge readability.
Weight comes from contrast: slow, heavy holds against fast transitions. David Rosen's GDC talk on procedural animation for Overgrowth is a favourite in workshops because it shows how few keyframes you need when the timing and interpolation carry the weight.
How to practise this week
Pick one small action, a jump, a dodge or a light attack, and build three versions: film-paced, game-responsive and exaggerated. Bring them into Unreal or Unity, hook them to input, and ask a friend to play. Write down which one they choose without being told. That single playtest teaches more about game feel than an hour of theory, and it is exactly the kind of loop a strong workshop builds into every session.
