Monday, August 11, 2008

My random question study

If you really want a random question that has a point such as to develop a game or a collaborative book it should start with questions that give have similar points but generate different answers. If you are interested in game development this could be used to create a game that is based purely on random strings that are generated from the story.
The following question can be regenerated dynamically in many different ways. We will look only at the year for now.

The year is 2089 and you have been given nearly unlimited control of a secret government program on one of the largest unused military bases on the planet. On the surface it is just another over funded research project. In truth it is...
Now lets look at what can be randomized to produce different responses.

The year is ("$") and you have been given nearly unlimited control of a secret government program on one of the largest unused military bases on the planet. On the surface it is just another over funded research project. In truth it is...
As the year changes the so will the reality of the responses. The closer to the actual year the more likely the response is to be based in what is known or perceived as truth. The further away from the current year the more likely that it is to be based in fiction. If the year is left to the respondent to fill in the control of the question would be harder to maintain. If the year chosen is done randomly the control of the answer is then harder to maintain. A possible solution for this is that the year is set by response to other random questions. When the year reaches a preselected level of response it can be locked and a new string can be developed.
Another possible solution could be that the string is generated dynamically by the users input to the question. The way this would work is that words and word groups can be given a rating that could be translated into years. This is to say that the more futuristic the wording the further into the future the year becomes.
This could be used as a direct challenge to the respondent to get the year to fall into a range the respondent wants. Each time the respondent clicks a button or a refresh interval is reached the year would update.
Once the respondent has finished the session they could choose to save as a draft or publish. The published response would then be sent to the story for rating. The rating process would use keywords rather than a numerical type rating system. Any time a comment is made on a published response the keywords are chosen and given a weight.

No comments: