How to IC and TIC
There are two programs available. IC produces integer numbers.
TIC produces time values.
The interface is divided into three columns: input parameters,
The right column contains the action buttons. Take a moment and
think about your question before you toss.
The center column is a text area that will contain the
tossed answers. You can toss multiple times, and each time your
answers will be appended to the center text area.
TIP: You can also type anywhere in the text area, to label or
comment on the questions and answers you are producing - useful
for when you print or save your answers and refer to them later.
The left column is where you set the input parameters, that is,
where you shape your question.
Both programs share these input paramaters:
numqstns: The number of questions you are asking. This is the
quantity of answers you can expect.
bot: Defines the bottom of the range of answers.
top: Defines the top of the range of answers. Answers will fall
between these two values, inclusively.
bias: An advanced feature we will explain later.
layout: Your answers can be formatted as a list, one per row,
or a list, comma separated (these two are useful for pasting into
code or other programs).
Classic layout is how the very first versions of IC and TIC were
layed out for easy printing and transport.
TIC has an extra paramater called "time units", which affects
both input and output.
The numeric values of "bot" and "top" are multiplied by the
defined "time units".
An example: If you set the time units to 20 seconds, and the
range from 3 to 6, your answers will be one of 00:01:00,
00:01:40, 00:02:00 (the bottom is 3 times 20 seconds = 1 minute,
the top is 6 times 20 seconds = 2 minutes).
Another example: If you set the units to 5 seconds, and the
range from 0 to 120, your answers will fall between 0 and 10
minutes, because there are 12 5-second
intervals in a minute, and 120 of them in 10 minutes.
Not all values can go into the unit's numeric paramater. They
have to divide into 60. Acceptable values are:
To understand how bias affects your answers, you have to know a
little about how IC and TIC simulate the I Ching coin tossing
oracle. Tossing the I Ching using coins requires three of them.
One side is assigned a value of two, the other three. Once the
coins are back on terra firma, you add up the values (twos or
threes) to get a single value of 6, 7, 8, or 9. A 7 or 8 are
twice as likely as a 6 or 9.
For the sake of simplicity in this explanation, you can think of
6 as equivalent to 7, and 9 as equivalent to 8 -- meaning, each
toss is binary, having only two possible outcomes.
You toss the coins five more times, total six, then combine the
six binary numbers into one, each occupying one bit of a six bit
binary number. Two to the power of six = 64. There are 64
possible values for your outcome.
Following this procedure, IC and TIC first produce a set of
answers between 1 and 64. If you specify some other range, we
have to map 1-64 to it. The mapping is done using whole numbers
only. If your range divides without any remainder into 64 or 4096
(64x64), our mapping is easy and even. But if not, we divvy up
the range as equitably as possible -- meaning, there will be a
slight, unavoidable tendancy towards some results over others.
Realizing that in many cases there has to be a bias (as small as
can make it), we decided to play with the idea of bias itself.
The bias:none setting behaves as described above -- the bias is
a slight as possible.
The bias:bias setting internally tosses its own answers to
divide 1 - 64 into as many parts as needed to map your range.
The division sizes are not as equitable as possible, rather,
whatever chance determines, and their differences could be
slight, acute, or
anything in between. What you end up with is a bunch of answers
that, to a greater or lesser degree, favour some numbers, and
disfavour others. The effect can be very subtle, or very
Furthermore, these biased divisions do not
necessarily pertain for the entire set of required answers. We
toss again to find out how long the current bias pertains (up to
a maximim of 64 answers), then, if there are more answers
required, we toss again to get a different bias, which in turn
pertains for up to the next 64 answers. In summary, the mapping
is as biased as chance determines, and lasts as long as chance
determines, reconfiguring itself repeatedly after
The bias:immobias setting acts in the same manner as the bias
setting, except the first tossed bias is used for all answers:
there is no change in bias, hence, immobias = immobile bias.
The bias:nonrepeat setting is biased in a completely different
way: no answers can be the same as any previous answer. Note that
nonrepeat demands that numqstns be less than or equal to the