Each reader's time per page is compared against an average time per page to take into account varying speeds of reading. The resulting number gives us a relative measure of how much longer or shorter the reader spent on that page versus each of the other page's. This can tell us which sections of a novel could use improvement. Similarly, by counting the number of readers who hit an individual page, we can see if there is any point in the novel where readers generally become disinterested. This information can then be used to improve the weak links so that the reader's continue to read the novel all the way through.

This is a typical example of a combination of hits per page and time per page being used to acquire a Confidence Rating for each section of a novel. The author then proceeds to revise the novel using a Reader Initiated, Author Executed adaptation method based on the rises and falls of the Confidence Rating throughout the novel.

Example II - Multi-Authored Continuing Story

By permitting sections of a story to be challenged, we allow the group to not only author the story as it grows, but also edit out parts of the story the group feels are inappropriate. Multiple challenges can be posted for a specific section of the story. Votes are taken through some type of voting scheme and the section that has the highest Confidence Rating is declared the winner and chosen to become permanent. Losing sections are then deleted from the story and the story continues.

A time limit can be placed on voting to speed up the process. After voting has occurred, the winning section (or original section if no challenges were made) is then frozen so that the plotline can continue. Different methods can be developed so that the plotline can continue even while previous sections are still in the voting process. One such method would be to not start the time limit for voting on a section until all prior sections have been frozen. The extra time limit allows for challenges to sections that may have been affected because they relied on sections that were still in the voting process when they were written.

Ties can be dealt with in one of two ways. The voting process can be extended until a section wins, for an "overtime" cycle, or for another full voting cycle. Alternatively, both sections can win and the story can branch. Stories can then branch and branch and branch and the authors are not only contributing to a linear story, but rather to a story more in the form of a tree.

Let's assume that each section resides on a single page, but that initially there are no relations between each section and any of the others. There is however, the ability to create a relation from a section to some other random section. A reader first entering this novel will have to choose to create a new relation on each page to move around. But as more readers read this novel, relations will already exist between pages from previous reader's treks.

Readers now have the option of continuing along the already existant relations, or creating a new random one. A method is developed whereby whenever a reader follows a relation between two sections, that relation's Confidence Rating is increased, whereas the other relations from the original page's Confidence Rating are decreased by a fractional amount. Once a relation's Confidence Rating falls below a certain threshold, that relation is erased. An environment is then created where links are now competing to be followed. Those that are followed most often will survive, while those that are never followed die.

The result is a novel that becomes like a forest, with certain well-travelled paths and other less-travelled paths. And paths that are ignored eventually become overgrown and cease to be paths anymore.

There are many methods that can be used to have sections adapt themselves. One example could be choosing a random word in the line and replacing that word with a suitable synonym retrieved from an on-line theasaurus. Another could be replacing the last word of a stanza with a word taken from a rhyming dictionary whose key is the last word of the previous line. Or words could decide to rearrange themselves.

An extension of this is to place the various sections of the poem in an artifical digital life type simulation, where each section derives 'energy' in the same manner a Confidence Rating is derived. Sections then compete in the boundaries of their world (the poem), by moving, changing and perhaps even eating and destroy other sections. Sections which are getting a high Confidence Rating will have less reason to adapt and change whereas sections with a low Confidence Rating will have to adapt or be destroyed by more powerful sections.