These are some thoughts on commitment, making a move feel like there’s some weight or momentum behind it, and how cancels change how we think about a move and how it feels to us. It’s unfinished, but it can be an interesting insight.
Some games have a really soft feeling of commitment to actions and others have a really hard feeling of commitment to actions. Lets have a look at how games with low commitment can feel a bit too mushy and how ones with high commitment can feel a bit too rigid.
Obvious contrast is between bayonetta and DMC, DMC has much higher commitment, because fewer readily available actions can cancel commit actions. You can’t cancel attacks with dodges, you can’t cancel attacks with movement options. Instead in DMC3 and 4 to cancel options, you need to use royal guard, which functions as a universal cancel, to transition into a neutral state, and then use the move you want to cancel into. This retains a big feeling of commitment while giving a means of getting around it. By contrast, the original DMC only offers roll dodges as a means of evading attacks (and the invincibility frames on jumps). These roll dodges can only be initiated while not attacking, so attack or defending must be considered more carefully as there is a greater commitment to attacks overall.
Being able to cancel into movement options in general reduces the feeling of commitment to an action, obvious example being cancels into dodges that move you. The more flexible the dodge, the further the feeling of commitment is reduced. God Hand has dodge cancels, but you can only do them forward (practically royal guard), to the sides, and back. The back dodge itself has a long usage time, and cannot be canceled into anything else. Between that and the side dodges, in god hand you still have a strong feeling of commitment to actions. By comparison, in Bayonetta you are allowed to cancel any attack into a dodge at any time. Dodges in Bayonetta can go in any direction. This gives a much lower feeling of commitment, because combos fluidly flow into dodges in any direction at any time without restrictions.
Being able to move as an action is performed, like a jumping attack, reduces the feeling of commitment. Attacks like Stinger however still have a strong feeling of commitment.
IASA frames are a term from Smash Bros for frames in an animation at which point the animation allows itself to be canceled into any other. The acronym stands for “Interruptable As Soon As”. Many games feature these at the tail end of attack animations and it helps the transition out of the animation feel smoother by reducing commitment at the very end and giving the sense that the action flows into the next. Examples of games that do this are obviously smash bros, DMC, Bayonetta, God of War, and Dark Souls. An exception to this would be Kingdoms of Amaleur, and Psychonauts (just the examples that come most readily to mind).
In general, cancels from one animation to another give a sense that the two states are connected and flow into each other. This should give a bit of context for the differences in cancels illustrated above. Cancels like the royal guard cancel or god hand’s weave dodge cancel feel like one is making an effort to halt the prior action, where cancels like bayonetta’s dodge out of any action make it feel like at any time, one can just do anything.
On a further level, the inaccessibility of the royal guard action in DMC4 and the weave dodge in god hand further contribute to that feeling. Most DMC4 players probably do not play in royal guard mode all the time, so in order to royal guard, they have to first style switch, which involves removing your thumb from the control stick to actually reach up and press on the Dpad. In God Hand, the right stick, used for dodges, is removed from the face buttons and has to be reached for to dodge. There is more effort in the physical motion, so it feels like more of an effort to halt the action. By comparison, God of War has the block button readily accessible at any time, so its block cancels have less impact.
Combo chains give the sense that each move flows into each move as a sequence, where links feel more like each move is its own move and they are performed one after the other.
As players play, they build up a model in their head, a simulation of how the whole system works, on the basis of the many repeated times they use actions and the slight differences in outcome each time they use it. Players start games by essentially guessing at the function of things based on the visual symbols they are given, and over the course of play and exploration of “corner cases” (rare circumstances that demonstrate more fully the )
IASA frames are a great example of this type of model building. People rarely if ever see the full followthrough of animations in games, because they get so frequently interrupted once the IASA frames start. However players still get this smoother sense from their presence, because at one time or another they saw the full followthrough animation, or they can subtly perceive where it’s supposed to be and know it’s getting interrupted. In the Animator’s Survival kit and other animation guides, animators frequently mention how there are frames drawn in that convey motions that pass by far too quickly for the audience to really see and realize they saw, but which they will feel and give a greater impact to their perception of the motion. A classic example being squash and stretch used in inbetween frames. Games are filled with these types of things and the most obvious example is cancels themselves.