Three ways of model-forming that are explicitly listed and discussed are: Generalization, Deletion, and Distortion (SM 14–16). For each of them, Bandler and Grinder are careful to show that they can be helpful and useful techniques for people when coping with the world, but can also result in limiting choices and thus impoverishing their lives; much depends on the context and the manner in which they are used. Generalization is defined as "the process by which elements or pieces of a person's model become detached from their original experience and come to represent the entire category of which the experience is an example." (SM 14) Next is deletion, "by which we selectively pay attention to certain dimensions of our experience and exclude others." (SM 15) And finally, distortion "allows us to make shifts in our experience of sensory data." (SM 16) Examples of this latter form include fantasy, allowing us to "prepare for experiences which we may have before they occur", such as when we rehearse a speech in private before giving it in front of the actual audience. (This list is not exhaustive, as Bandler and Grinder point out, and of course the borders of these categories can be fuzzy; also, these forms of modeling influence and amplify each other via various sorts of feedback loops.)
Now, this sort of account of what goes on in people's minds in terms of representation (forming models that represent reality) has been severely criticized in recent decades by philosophers of mind and cognitive scientists; but for some practical purposes it may still by a helpful picture, and although I largely agree with the criticism, I'm not going further into it here. Instead, I'd like to draw attention to a certain parallel with Nelson Goodman's list of 'ways of worldmaking'; I think there's something to be learned from both accounts if we're interested in the ways of forming unreality (both out of reality and out of antecedent instances unreality — preexisting 'versions', as Goodman would call them).
In Goodman's view, there are many different worlds, or world-versions, which come about by our uses of symbol systems. Whenever we describe or depict what's going on around us, we produce another version of the world, potentially different from any other version that previously existed. Since versions are created in various ways and in different contexts (a version may well be unique to a particular person's viewpoint at a given place and time), there is a multiplicity here comparable to the one pointed out by Bandler and Grinder in The Structure of Magic. And since every version came into being by a creative process that included a symbol system (such as a scientific description language or a conventional style of artistic expression), there is not really a point in speaking of a world-in-itself, a world that isn't a version — versions, in Goodman's sense, are all there is to the world. Thus talk of multiple versions and talk of many worlds comes down to practically the same thing. (Goodman makes it clear from the outset hat this doesn't imply an arbitrary anything-goes relativism, see WWM 17–21.)
Goodman enumerates various ways of how world-versions are produced: Composition and Decomposition, Weighting, Ordering, Deletion and Supplementation, and Deformation (WWM 7–17). He makes pretty much the same observation about deletion as Bandler and Grinder, noting that "what we find what we are prepared to find (what we look for or what forcefully affronts our expectations), and that we are likely to be blind to what neither helps nor hinders our pursuits are commonplaces of everyday life [...] And even within what we do perceive and remember, we dismiss as illusory or negligible what cannot be fitted into the architecture of the world we are building." (WWM 14–15) He includes under the heading of deformation a somewhat different sort of phenomenon than distortion in Bandler and Grinder's sense: for him, this includes changes such as a physicist's smoothing out of a diagram curve to emphasize the underlying simple kind of shape, even though the data are actually off that shape a little, or a related distortion in caricatures that have the job of emphasizing a detail in the caricatured person (WWM 16). It's not quite clear to me why Goodman thought this a sufficiently different category from his second one, namely weighting. In any case, while his discussion of deletion and supplementation of details is richer than that in The Structure of Magic, and his discussion of weighting and ordering has no counterpart there, he doesn't explicitly include the process that Bandler and Grinder call distortion. The reason is probably to be found in a difference of perspective which he mentions before giving his list: "Actually, I am more concerned with certain relationships among worlds than with how or whether particular worlds are made from others" (WWM 7). The focus is rather on formal properties and relationships of world-versions than on the psychological or creative processes that bring them about.
This difference in focus is even clearer with respect to the first category in The Structure of Magic: generalization. The term is used there to refer to the formation of a rule of behavior, caused by some experience (possibly repeatedly made). Rules of behavior, although they can be formulated in a linguistic description (in the words and sentences of a natural language) or even in an artificial symbolic form (such as in a logical calculus or a computer programming language), are basically habits of perceiving and acting. Thus, they're not in the focus of Goodman's discussion. Still, the labeling and classification processes he refers to in the section about composition and decomposition cover the process of generalization that Bandler and Grinder describe (see esp. WWM 8–9), and much more.
 Richard Bandler and John Grinder, The Structure of Magic. A book about language and therapy, Palo Alto: Science and Behavior Books 1975, 5–20. References to The Structure of Magic are given with SM and page numbers.
 Nelson Goodman, Ways of Worldmaking, Indianapolis: Hackett 1978. Quoted as WWM with page numbers.