For Impatient Users
You need to guess the cause(s) out of the 6 possible causes listed in the column headers in top and middle tables. To make a guess, you click on a column header. Before you make a guess, you may want to learn more information. If you see that there is a question mark associated with a person in a column and you would like to know whether the person was exposed to the cause at the head of the column, then click on that question mark ('y' means yes/exposed and 'n' means not exposed).
For Curious Users
Goal
You are trying to find the cause or causes (e.g. Blood Supply, Mosquitos,...) of the pandemic in as few game days as possible. A day will pass when you ask five questions or when you make a guess that some cause is (partly or wholly) responsible for the pandemic. You can ask up to four questions and make a guess in a single day.Parameter Setting
The default parameter settings are for an easy level. Feel free to make the game more challenging, either by increasing the level number of by manipulating the parameters. Here is what each parameter setting does:- Number of Causes for Pandemic: It is set as 1 by default. If you increase it to 2 or more, then any one of several causes is sufficient to make symptoms appear (perhaps with a delay and a probability, see below). That is the OR of these many causes is the symptoms column. Increasing this setting makes the game harder.
- Days to Show Symptoms: It is set to 0 by default. If 0 then symptoms appear right away (perhaps with some probability). If d days then symptoms may appear at any time up to d game days from when the game starts. That is, 'n' in symptoms column 'might' change to 'y' within 'd' days. For example, if you set this parameter to 3, then a 'n' may change to a 'y' any time in the first three days. Note that a 'y' will never change to an 'n'.
- Probability of Showing Symptom: It is set to 100% by default. If 100% then whenever the correct causes are present, the symptom will appear (perhaps with a delay), but if less than 100% then the symptom will appear only with that percentage. For example, if you set this to 70%, then 70% of the time when a symptom should appear, it will appear.
- Number of People Surveyed: The number of people who have been surveyed about their exposure to causes and whether they show symptoms.
- Percentage unanswered questions: If every person surveyed has answered all questions about his/her exposure to potential causes, then this percentage is 0. If greater than 0, then some of those questions are unanswered (and will be represented initially as question marks).
Play
If one cause, then you need to identify just one cause for the symptoms. It two or more causes, find the causes whose logical OR exactly matches the symptoms. For example if two causes are responsible then if the person surveyed has been exposed to either cause, he or she will acquire symptoms within the delay period (Days to Show Symptoms) with the probability specified (Probability to Show Symptom).
To pose a question, click on a question mark. You will see the answer. Remember you can ask five questions per day. Hence, best time to take a guess is after asking 4 questions.
To guess a cause, click on the column header corresponding to a cause. Remember that guessing a cause causes the game day to end (even if correct).
Information Displays
- Top:Correlation of a single column measures how well the answered questions of the column match the symptoms as currently visible. Correlation of 1 means exact match and correlation of -1 means exact NON-match. Each entry shows the range of correlations that define the "confidence interval" with the measured correlation in the middle. Intuitively, the confidence interval defines the range of most likely real correlation values between this column and the current values of the symptoms (if all questions in the column were answered).
- Bottom: Pairwise Corr box gives the correlation between the symptom column and both the row cause and column cause together (assuming an OR relationship between them). This measure reflects only answered questions and the shown symptom values.