iPILS

The Re-verse Favorite (to be - or not to be)
The topic of re-verse favorites is related to the bias of an economic
marketplace (pari mutual). It is significant because likely randomness
should exist, in all practical events of chance. The latter presents us with
a series of points on a line, each indicating a possibility (the randomness
line). The re-verse is true of the re-verse favorite, which in the market,
represents a single point, indicating a clear probability (high or low),
which is much more rare than a good chance (in randomness). And the fewer
points we use to estimate (highly abstract), the less information involved
in the process (fewer ways), which inhibits the diminsions of prediction.

The moving average method's observations are lost at the beginning and at
the end of the time-series.

Data___MovingAverage_____Percentage
3.0_______
5.0_ ______ 6.0 _____________83.33333
10.0______ 7.0______________142.85714
6.0_______ etc.______________etc.
etc._______ etc.______________etc.

Above we have the solution to a moving average problem. The author doesnt
mention how he got these numbers. He says, "write a program to do a three
period moving average. Furthermore we want a program to compute each
original value as a percentage of the moving average opposite it, when the
qverage is properly centered. For example the series 3,5,10,6,... the output
should be as given above.

We are researching this. A four period moving average, involves finding the
arithmetic mean (average) of the first four items, then the mean of the
second through the fifth items, then the mean of the third through sixth
items, etc.

The arithmetic mean is a measure of the central tendency. Some other
computatuons involving the central tendency are the geometric mean (the nth
roo, of the product of n numbers), the quadratic mean (the square root, of
the average square of a set of numbers), and the harmonic mean (the
reciprocal, of the arithmetic mean of the reciprocals, of a set of numbers).

Returning to the problem, a moving average is used when all of the data have
not been collected at the same time, or if the prediction is spread over
time (i can do it on monday, maybe tuesday), or when the masses of data
involved require a moving average for sections of the massive data set.

Period Moving Average Formula
LOAD A,B,C,D (sets A=3, B=5, C=10, and D=6)
AVG=(A+B+C+D)/4 allocates formula memory
AVG=(3+5+10+6)/4 (calculate the average)
PRINT AVG (that average will display a 6)
A=B
B=C
C=D
READ D (this is a way to say GET next data item at its place, the data item
on disk)
GOTO 3 (line #3 is AVG formula)

What these steps did was calculate and output the average, and move to the
next data item, read that and do it again. Statements like A=B copies the
secong item to the first slot in the formula memory, so that D the last slot
get copied to C and then we load D with the next data item on the disk.

What im working toward is programming predictions of frequency (as we recall
that is based on the # of recurrences), from observations. But there is a
ways to go ..

========

_as for classes of information we can use the properties of data to look at
the results that can be gathered with use. any calculation that involves a
prediction, will yield a result that is a prediction, for example. and
otherwise, we should adjust properties so that we will have natural numbers
above zero to work with, which can be done by scaling (multiply by 10 for a
decimal place)
_after scaling, two ratios (fractions) are equal if a/b and c/d are equal
when cross multiplied, saying that (a,b) is the same as (c,d) if ad = bc. a
solution where these two progress lines will converge

===========

the above link leads to a chapter which discusses controlled spontaneity,
which is what i was looking for information about .. its a reflaction on
newer information. Yet older ways of being/awareness .. also control the
questions too . There isa reference to the extreme theoretical ..
=========

my own minor attempt to find an answer, points me toward a recognition that
the understanding of a spontaneous event occurs when a theory is present,
and seen. that would imply that there are theories which no one sees.. but
when the theory has been seen, some new drama begins in society
======
P.F. Kellermann
1. What was Moreno's contribution to group psychotherapy?
2. What is the focus of sociodrama in comparison to psychodrama?
3. How would you define psychodrama?
4. What is the function of the auxiliary ego in psychodrama?
5. What is the function of the auxiliary ego in human development?
6. Which are the four professional roles of the psychodramatist?
7. Describe some of the different kinds of warm-ups in psychodrama.
8. Describe shortly the eight phases of classical psychodrama within the
Hollander curve.
9. Which are the four main psychodramatic techniques?
10. Put these techniques in correct order within the process of child
development.
11. Give two definitions (wide and narrow) of the concept "sociometry".
12. What is a "sociogram"?
13. Give two definitions of the concept "social atom."
14. Define the concept of spontaneity according to Moreno and common usage.
15. Which two main components are included in the concept "role"?
16. Mescribe the six main therapeutic aspects of psychodrama with their
appropriate concepts.
17. Which two strategies are used to work with resistances?
18. According to which criteria is it possible to build psychodrama groups?
19. What is the meaning of the concept "tele"?
20. What is the difference between the meaning of "acting out" in
psychodrama and in psychoanalysis?
21. What is the meaning with the emphasis on "here-and-now" in psychodrama?
22. What is the meaning of the word "action insight" within psychodrama?
23. How do we call psychodrama without a group?
24. Who plays the role of the other in such a setting?
25. What is the connection between social psychology and the theory of
Moreno?
26. Which are the main differences between psychodrama and Gestal therapy?
27. Which are the main differences between psychodrama and drama therapy?
28. What did Moreno mean with his saying that the roles develop before the
ego?
29. Describe the meaning of the concept "surplus reality".
30. What did Moreno hope to achieve with "sociatry"?
31. Why did the chicken cross the road ?

=========

Apple Computer ran a Superbowl Commercial spot, in early 1984, where the
commercial proudly explained the great and varied abilities which the
computer provides to many of us. The computer allows us to discount the
possibility that society could be controlled by Newspeak propaganda, and
mass indoctrination, terror. The Apple approach was that given a means to
advance using the computer as a common tool, and given the computer programs
available, that the whole set of possibilites would shift toward a tendency
for higher function.

An evident change would be apparent within a data set, which was designed to
consider these questions. It could be a set of readings taken before, and
after, the Apple computer. But we would expect to see some greater total
variation, since the lower functioning types would not be able to use the
computer well, and the higher would be very able to greatly exert an effect
on society while using their computers.

By way of introduction, we have the "capture" of data for our purpose, and
now we would like to be descriptive and evaluate the superiority being
introduced along with the computer, how consistent was that effect, and why
(or what alternatives can be analyzed) like dumbing down the losers to
create a totally superior group of master race Apple computer users. These
issues demonstrate some significance of variation, whiich we can decribe in
terms of a shift (before and after), or rate of the effect, or what
alternatives are eliminated because of some new development (variation)..
It was said that none of the test scores have high value for the purposes
of our problem. It seems most apparent that individuality competes with
problems of society where explanations do not relate to individuals, while
modernization or technology causes more individuals to suffer due to the
system and those who exploit the "use" for information which is corruped in
value, or the nature of gathering, or false correspondences and assumptions,
or reprtitions of the false to distort the nature of truth .. Lecture
begins here .. >next paragraph

======
Betting the Big Spread(s)
_this is about using the speed figures that they give you in the racing
form. The first thing to consider, is that for a race having 9 or so, horses
you can have about 50 -90 speed figures from the form (10 lines or less).
First you have to put them on one line, in order, and identify each figure
to the horse it belongs to. Of course, some can be discarded if the horse
fell down or something ..
_second thing is to look at which of the figures is most typical, and how
close that figure is to the center of the line you made with the figures on
it. If 6 out of 10 of the figures, for example was a 66, you can look at the
range from there. The range being the high score minus the low score, and
see where that is next to the typical 66 you have as the typical score.
_you're looking for the "mean" or the target value. Call the target the
"expected".
_now go back the each horse and compare each of his speed figures to the
target. We are making the spread for the race at this point. Take the
difference, between the target and the speed figure, and multiply that
number by itself. Add all of those numbers up, and divide the sum by the
number of those numbers. Call that the variance (squared average
deviation)..
_that gives you a number which can be created by multiplying some two equal
numbers. If the variance is 16 (4 X 4) then we are looking for the number 4.
_we can look at the spread using this number (4) which is known as the
standard deviation. Best thing for a spread is to look at multiplying the
standard deviation by a whole number.
_lets say we take each speed figure and compare it to the standard deviation
X 3. Now instead of having the speed figures i.e., 25, 28, 30 .. we multiply
them each by the standard deviation X 3 (12), and also the "expected" value
by the same 12.
_now we can put all those values on one line, like we have done earlier. We
see the greater spread between the speed figures now .. and ve see which of
the horses has been doing much better than the others.. also which are doing
much worse .. and we can tell how close the race will be, because these
figures are already being exagerated, so if they look close, that means
there is a tight spread there..
next paragraph

_from the figures, look at the great horse possibility .. is there a truly
great horse that is not like the others. If that is the case, just calculate
the variance without the figures for the great horse .. _Otherwise the issue
of how much better he is than average will throw off the figures that
position the other horses that are more like each other ..
_also look at the par figure that the racing form's figures give. is that
near to the par that you have calculated, as the "expected" value. It is
time to consider how this group of horses is compared to the par group (all
horses in the type of races we have here).
next paragraph

_assume there is some inference for you, good. now its time to tweak, so
look at the original speed figure line, and the "expected" that was
calculated for the problem. look only at the figures that are below the
"expected" for the moment. look for the average among that set of figures
(the below average ones). is there a typical figure among this set we are
considering. otherwise consider the average of the figures on the other side
(the set of figures above average). average the above average figures. is
there one or more fpeedfigures on wither side that does not approach the
boundary limit of the average (above or below). now do these figures belong
to different horses, the same horse, or a combination.
_we are trying to see if there is some way that the problem can not be
solved, using our approach so far. . if there is only random things going
on, we have to rely on random chance. if there is one horse whose figures do
not belong to the set for the problem, we have to consider that one horse is
a possible random factor for this race.
_also we need to consider the average above the "expected", vs the average
below the "expected". Should we create a new measure of variance for only
those horses that are above the average, since a current variance we
calculated doesn't show the "spread" among those on the "positive" side of
the line we had for the "expected".
_we will want to draw the "expected" line on a chart and add our next set of
horses to the chart .. so we can have a better idea of these horses in
general ..
next paragraph

_at this point, the calculations are about over. we have only to look at
which side of the "expected" does each horse have more speed figures. the
most bettable horses will have alot of speed figures that are on the higher
side. the horses that are longshots will have their speed figures to the
lower side of the "expected".
_the statistics people calculate the variation associated with each speed
figure:

speed figure - expected / standard deviation

that is known as the standard Z score,.. the standard Z score is thought of
as dimensionless .. (it is the regular guy score). Z is the score of the
conformists in society (each score has its own Z score) .. each has the Z
written on the birth certificate, and they stay like that .. dimensionless
next paragraph


The Prof. considers Central Tendency from the reverse standpoint in his
subsequent presentation. The Prof. offered a definition of "closeness". The
"mean" is the closest of the main statistical measures to all of the scores,
using one definition of "close". He suggests that this analysis is
speculative, and that it would be an unusual definition of "closeness", at
that. The topic is still focused on the Central Tendency but the detail
involves the data now, rather than the issue of the "mean" which was and
seems now, to be an introduction to more statistics. He departs here with
the statement, "the means are nearly always abstract concepts." While the
Prof. lecture moves into a discussion of statistics which help us to study
the numerical data set which was proposed, I will place his summary point
about the Markov inequality for the "mean" in context here. The Markov
inequality is a way that Markov adopted to determine the limits of the
boundaries for his data. It is sometimes necessary to approximate the data,
especially rare data at the limits of the boundary. So Markov proved that
the Maximum portion that is greater than K "means" will equal 1/K.

For an Example of Markov's inequality, suppose the data about male tallness
indicates that the stature of a man averages 6 feet tall. We want to test
the limits of the boundary which theoretically exists at 3 "means", or put
in other words, "What proportion of the sample of men were measured to be 3
X 6 feet tall (18 feet tall).. Since we are using 3 "means", K =-3. Markov
says that the Maximum portion that is greater than K "means" will equal 1/K.
Or in our example, the Maximum portion that is greater than 3 "means" will
equal 1/3.., which is 33.33%. So as diverse as the numerical data may be,
once the Central Tendency is known, the outer limits of the boundary can be
inferred, to the extent that the insignificance of such a data sample could
be explained (Markov's inequality).

The Prof. considered the dimensions of the numerical data set. A "median" he
explained, is that score (data score) which is exactly higher than half of
the sample data, and lower than half of the sample data. Whatever the data,
if it can be arranged in increasing order, the "median" is the particular
score in the middle. The "median" would often be the most indicative
(descriptive) statistical measure. We see that quite a few biased scores
would have to be added to the sample, and the scores themselves would need
to have wide variation at the center, in order to have a significant change
of the value of the "median". In fact, the Prof. says that the "median"
meaures the outcomes of the process in terms of sensitivity, and in term of
typicality. How much is the process fluctuating, or what is the typical data
result, are suggested by the statistical "median".

In terms of "closeness" the "median" is closest to all of the data scores
using a definition which is, "the difference between any score, and the
central score, regardless of which direction that score is from the central
score". Of course we are considering scores which are the outcome of an
orderly process, otherwise the "median" itself is rather, statistically
insignificant.

The Prof. explains, opposite to what I just stated about the significance of
the "median", that the "median" can be used to divide the data into two
groups (in fact many groups). The "median" can be used for marking off
various fractions of the data set. The "median" represents the 50%
percentile (hundredths of the data). A quantile is a fraction of the data,
other than a hundredths portion. Usually, about 50% of the scores are above
the "median" and 50% of the scores are below the "median". The "median" is
not rich in theoretical usage, but its general usage is popular in the
statistics field, so the Prof. incites.

Actually, this is another Central Tendency measure which is affected by the
nature of the statistical significance of the data, within the data sample
itself. And as I proposed in the first part of the introduction, the nature
of data collection is as important as the data itself. We will continue
shortly: there is a significant remainder to the Prof.s lecture ..
next paragraph

The lecture by the Prof. offers the idea that, "large sets of data are
expected to be uni-modal". Then he introduces an idea of, "a mixture of two
groups of data". He presents a suggestion of data dimensionality, with the
introduction of the modal score. The modal score is used as an independent
variable, to which we can add an attribute relevant for all values in the
data set. Here is the best example of the mixture, or a complex mixture of
two groups of data, since the common attribute is related to two (2)
distinct foreign sets.

What is the modal or typical score: the score that occurs most often. A data
set can in fact have two (2) or many modes. And if the scores happen to be
unique, without repeats, it is said that the data set has no mode. And while
the mode is a main Central Tendency measure, it is a unique score which
corresponds to the most dependent variable values. The dimensionality of the
data set is defined as the correspondence between the dependent data, and
the data location (the independent variable).

The idea of modal data would be the same as the idea of data location. That
would be the static value to which the dependent data would be assigned. The
idea of Central Tendency in this case, is not horizontal, but in the
vertical direction. The most popular points on the data grid would cluster
about the Modal score(s). The Prof. professes that the mode is the least
sensitive of the three main measures of Central Tendency. The mode's
location, determines "symmetry" of the set of data scores when graphing the
scores.

All three types of Central tendency measures can be misleading. Nearly all
questions of "typical-ness" are affected by the nature of the group
(population) being studied. And when a few values included in a data set are
found to be much greater, or smaller, then no measure of Central Tendency
will (can) reflect that fact by itself.

When two or more distinct groups are mixed together into one "population"
any measure of Central Tendency may need to be augmented with information on
the "variation" found in the group. That is the subject of Prof.s next
lecture. Most of the time it helps to gather some information about the
variance around the mean, rather than to use such a crude mathematical limit
boundary, as Markov's inequality finds mathematically.

The Effect Size, is defined to be the difference metween the grand mean and
the group mean. At this point the Prof.s lecture wanders into subgroups
(quantiles) dispersed about the main median (grand median). It is easy to
visualize the mixing of two data sets which have a common location
(independent variable) and are clustered in three dimensions on the graph,
which is centered at the grand mean common to both data sets which we choose
to study in unison. The measurement of each Effect (Size) is the sum of the
lengths of the legs through the grid, between the points. What would the
average of those Effects be but a vertical median, which makes sense to me.
The Prof. says, the Effect Size "summarizes" complicated experiments. It is
a very important concept in statistics.

The "mean", in any case, is also called the "expectation". The Effect Size,
is what we "expect" to be the result of being in a particular group (of
measurements).

On the availability of the three main measures of Central Tendency: All
three (3) are possible with some data. The median, and the mode are
available with orderable (quantified-numeric) data. The Mode, is used with
purely nominal (labeled by name) or type data (names - mixed type).

The three main measures of Central Tendency: can be identical. The median is
the mid-point of the data, and the mean is most effected by extremes among
the data set scores.

Prof. is summarizing, notes that other measures of Central Tendency, would
be:
Geometric Mean - mass created by multiplying scores together instead of
adding them. the mass is cut apart by taking the Nth root, instead of
dividing by the number of scores (N).
Harmonic Mean - reciprocals of scores are averaged, and take the reciprocal
of the average.

There is a book on Central Tendency which describes over seventy (70)
measures of Central Tendency. His lecture explains how and why they are
used, but does not explain all seventy calculation formulas. After the first
lecure (this lecture) Prof. will ignore the median and the mode for the rest
of the statistics course. (its a good thing for youse)//

The homework assignment is - (1) find a variable of interest. Collect some
data on that variable. Calculate the mean, median, and the mode. Note
differences in the three measures. (2) take a look at the book, "How To Lie
With Statistics" by D. Hoff. (3) Look through the Almanac and other books of
numeric information. Calculate a few averages of interest. Average over time
as well, as in constant time.

The last point Prof. makes is that, all three measures of Central Tendency
can be helpful but much of the instruction focuses on the mean. And that
very different configurations of scores can produce the same mean.. The mean
is the figure that would have produced the same total from Equal
Contributions, if all had contributed Equally.

next paragraph

All of the typical lessons of the television statistics course present the
topic, then offer the main presentation, then consider the applications and
associated types of problems (how successful are those applications), and
then the homework excercises. Today's lecture is about the topic of
statistics, and in particular the Central Tendency. The lecturer presents a
sample of numerical data to the class. The sample data is a set of eight (8)
numbers. The eight numbers are {13,11,5,11,9,15,7,9}. He explains that there
are many types of data. It was suggested in the course introduction above
that data should be stamped within a time, but it would be just as relevant
to stamp the data with a location, or with a way of translation (a language
or code), or with a mechanical process, or something else. Our course will
consider combinations of data processing methods, involving abstract data
types along with real and abstract processes. The professor indicated that
his sample numerical data set was proposed, because such data can be ordered
from lesser to higher values. And because several values could be easily
identified as identical to other values within the set, for purpose of
explanation.

The main presentation begins, as Prof. Bill Kirby and he directs the class
to consider that "the time order in data is often lost or ignored by social
science reseachers". The time or temporal order of the data is often
indicative, he broadly suggests. Cental Tendency would be a first step in
the usual data analysis. He considers how the Central Tendency can
"describe" the data. In a large set of numerical data, each number has only
limited worth, but the collection of these numeric values can be descriptive
of some effect from some process (%problems with a process). The Central
Tendency is both "descriptive" and "inferential". But he lecturers, as we
work with the data we should first describe, and only then should we infer.

He returns to the numerical data set, and propositions the student about
"capturing" the data. I consider how he is using "capture" in the operative
sense. The professor mans to "use" the data, he has to capture it. This is a
question common to analysis of information. The considerations are too broad
to analyse anything at some levels (the unknown unknowns etc). In database
information systems we call this the scheme. The scheme can be all that we
propose to have in the system, there is no limit to the scheme .. (how about
that..). The information system can contain abstract formulas, guesses, maps
and charts, or anything else the human mind has been able to conjure. In
fact perhaps the words "capture" and "conjure" would be synonomous in their
usages.

The professor continues describing the problem of capturing the data. He
proposes several processes including, (1) memorization, (2) sampling, (3)
calculate the "center" of the numerical data, (4) then measure "variation"
from the center. The Prof. presents the formula for a "mean" (the arithmetic
mean). He explains that a "mean" of any sort, is a target point of the
process which involves the data. The place where the data would be "normal",
in some respect. He describes the "mean" as the sum of observations divided
by the number of observations. What he says does not apply solely to a type
of data, but to all information. The "mean" is the Fair Share, or the
abstract result of Equal contributions to the total (which begs the question
of how equal were the actual contributions). The "mean" is that Mass that is
the Result of Equally divided contributions.

As a measurement, the "mean" is known as the most sensitive main measure
used in statistics. It is also one of the primary methods used for
mathematical theory. That is to say, that we can look at the limits of the
boundaries by using the "means".

next paragraph (half a lecture onward, toward utopia)

The Public Television Statistics Course has begun this Sunday (Jan 25th). I
caught the first lesson, and I plan to take some notes and store them within
this file. The Professor Bill Kirby has notes on the Web Site for Univ. Of
Wisc. Stevens Point (Education Dept.) We'll use his notes and the television
lecture presentation to develop our notes. We'll add some object-oriented
concepts.

Lecture 1 : The Course, Central Tendency
The statistics course is a common requirement in graduate schools for
education. The statistics are also commonly used in disciplines such as
psychology, nursing, chemistry, handicapping, etc. The improvement of the
high school curriculum offerings will now include the statistics course.

Statistics is usually taught as an alytical tool, but many researchers in
social and population related studies are using qualitative methods more and
more. Depending upon the scope of the study being considered, the use of
interviews and observations may be just as good or better than statistics.
Statistics itself was created by many non-mathematicians, and it can be used
to describe all the phenomenon in the world. It is useful in understanding
life.

The term "Statistics", meaning numerical facts about the state seems to have
first appeared about 1770. That would place the subject in a neck and neck
race with the "Communist Manifesto", and now who ever hard of a "Statistical
Dictatorship". Some influential people at the time of the French Revolution,
Barron Tourgot, St. Simon, Condourcet, Victor Hugo (The Hunchback Of Notre
Dame), Napolean, and many Chemists, and Physiological Researchers (doctors),
began movements in their fields, to adopt the inclusion of statistics in
their work. Today we are fairly advanced in many of these fields, but at the
time there were few scientific instruments, and there were not many
calculators either. The knowledge on Chemistry and Biology at the time,
would be analyzed completely in the first several pages of a modern text.
The unfortunate result of the statistical science has been new forms of
government which revolve around elections, where the outcomes demonstrate no
statistical reason to have an election, but to exchange the method for the
cause. The dictators hide among the pollsters, now. The margin of error is
not measured by studying the opinion of the population, it is measured by
studying the dictator, or isn't it ?? Yes statistics can be useful for
understanding life.

The term "Statistics" has a common usage now, and is not as often related to
a particular theory. With a broad study of physical science, or population
variables, and use of algebra, the variables while abstract have
applications in "Statistics". Sometimes in algebra, the variables are used
to approximate or to determine the values of other variables. In statistical
studies they politely describe or determine whatever has gone wrong already,
or conclude that someone or other has not been statistically involved. Ha Ha
Ha ..

The topic of the first lesson in "Statistics" is the topic of Central
Tendency. Each week there is another lesson, including these other topics:
Measures of Variation, Charts and Graphs, Basic Concepts of Probability,
Concepts of Process Control, Studying a Single Variable, Using a Computer in
Statistics, Contingency Tables I, Contingency Tables II (what is wrong?),
Analysis of Variation I, Analysis of Variation II, Correlation and
Regression I, Correlation and Regression II, Other topics: non-parametric,
Bayesian, multivariate statistics.

The statistics course runs for 13 weeks on television, but the applications
which are related extend back to earlier message board topics. Firstly, If
you want to retire on horse racing income, keep money records .. i'll try to
design a system, looking at common practices and strategy; looking at common
financial statements, we will modify the variables to suit multi-ways of
abstraction. From the financial basic statements we may develop the ratio
analysis .., and that analysis is used to set a course, make predictions,
time, index, account, chart, ect.
To find where things are known, is is far from finding that a knowledge
exists somewhere. t.s. eliot was known to say that a certain political
philosopys was based on a release of political power, not accumulation, and
that destiny is as much a question as it ever has been. but he was talking
about a totalitarian world where up meant down and the good was the evil.; a
theatre of law and order .. power and righteousness meets televangelist
snake charmer .. so a piece of information for one thing, must be compared
with the reality, and it has a formulae of intent behind it, it comes from
somewhere ..

Natural philosophy pursues the relationship between things that do happen
and how evident it was, that those things would in fact happen. but to
reinforce the idea of recordkeeping, and our use of acounts to gauge the
things that have happened, or that will happen, in terms of the gambling
business, we need to have ways to work with a system where we collect
information that will be useful, according to our predictions. information
that is based on predictions
With our system design we can keep track of any details (i.e. expected
retained earnings - operating costs, for first meet at Tampa Bay Downs). The
admistrators office can do this for every day, at every track, for every
player, and pay out in predicted returns per share.

Predictions can be negative. a major winner is running in the race,
decidedly lowers expections for other horses in the race. its not too
difficullt to make a list of feared opponents, eliminating alternatives to a
degree. And then information coming in may show a profitable trend (30%
winners), yet other factor may also help, and the odds can be added 30%
(with a full field) and this is a short field, so the average even better
for this event. A better payout than expected .. it is a prediction
callculated backward from the odds payout. If the expected retained earnings
is 2-1 (a 3-1 odds minus a bet of one unit), and posted odds for the horse
are 5-1, then expected retained earning have doubled. We are moving toward
looking at charting details which effect costs, expectations, analysis
ratios, ect

To determine the cost/benefit ratio planned projects we add the effects of
planned projects, but the principle of constancy requires a start and end
for each planned project. When science is done the measurements usually
include some time related element, i.e, gallons per hour, or clocking a
workout. the use of time can be applied in different stages of the projects.
in this case, we want to use prediction per unit of time. one winner
per/hour based on the top trainers horses. One winner a week, based on a
breeding shed, or a sloppy track. if we have knowledge, then it should have
a duration, otherwise it has no immediate measurable effect. time is not
necessarily costly to work with however. all of our analysis ratios,
expectations, and liabilities require time stamp data, without which we
can't chart anything of consequence.. But by combining different types of
causes using relational calculus (set theory) lets us eliminate and
approximate from one subject to the next. A complex matrix the involves a
production flow would use an index which measures the accurcy of the work.
In industry they adopted quality control management. our system uses
time-sensitive information therefore we can index a ratio of progress over
time. With queuing operations graph, the probability of an the queue moving
forward during a period of time. Whereas, we have many inter-related
subjects to study and their progress, either in time or relative to one
another.
Some readings i've come across suggest that time can be operative throughout
a process in a multiple of ways. Firstly, time can establish a status. This
would be a variation on the statement, "The time is right". This issue of a
status is probably worth looking into, the mix of cause and effect is
notable though. The next aspect of, time to consider is as, it is a divide.
I would relate, that as some history is not repeatable, so that shows an
example of two different times. Thirdly, there is a deterministic view,
where one event actually determines the next, in time (one has to be done
before the other can). They say a unit-time advance scenario is simple to
work with, but the simplcity results from the lack of an events list (task
list) to work on. The event-list adds a dimension, and adds a need for
calculations beyond the intervals of constant minutes or seconds, but the
time-process is managed by events just so as not to advance to periods of
inactivity. Which perhaps does not propose results by causes. But for events
that happen on a fairly orderly and regular basis, unit-time intervals can
be used, so long as periods of inactivity do not become a major issue. For
random events, the only solution is an event-list, and the relative time
system. The system requires an update between events however, which is why
systems which have to handle many variables are inclined to use unit-time
where advances to inactivity will not be an issue. In real-time systems the
time advance- should generally be less than half of the shortest delay
(inactivity) unit. Simulator systems programmers conclude that time-quality
is determined by the data structure in use, and the state changes as
designed into the system which can mean another event, in our considered
design.. Combining process increases memory use, but improves
time-efficiency ..
What a betting system should do, is handle many many small details that
influence predictions which become a strategy of betting. This handling of
information is improved by a process called object-oriented systems. using
the lambda calculus, object programs are called by each other for the
answers to detail analysis. Applicable to probability, frequency,
expectation, and hypothesis

As of the last cup o coffee, the "use" of information is based on the
conjecture that such information could be found, sooner or later. a
subsection of all this information is that information which we already
have, so the two possible information categories can be considered together,
or apart from each other. Perhaps, at some levels of abstraction, there is
only one group information that encompasses all objective or subjective
thoughts, and whatever symbols we can design to represent and manipulate
that set of all thoughts. however, in terms of "use" the set is limited to a
certain problem set, where the values apply. In looking toward a conclusion
based on information (infomatics?), it is necessary to look for the
overlaps, where incompleteness becomes a problem. its a problem if the
information is altered during the evaluation process. so it is safer to set
up groups of information which will be unaffected relative to each other,
and so the corruption throughout the process of evaluation will be limited ,
or at least detectable. but we see already, the evaluation has stages of
completeness and a moving goal line, where all the answers melt into another
set of unknowns .. The topic here is the level of reality where prediction,
affirms hypothetical issues .. A possible method to control the consistency
in a process of information analysis, would be to copy every common piece of
information from one source. the idea here is to limit the variations. To
use the best abstract concepts over and over, noting how one causes an
effect on another. and take measures to increase the effects for the sake of
observation or boundary testing. Where possible combine information into a
single point, prior to distributing it throughout the abstract model.

A couple of ideas to test here are variance (fixed or moving center), and
recurrence. the issues are pretty well defined to the point where those
calculators come in contact with our variance matrix which is effected by
some uncontrolled conditions which may point to a need to guess at
information..

In this introduction, the topic revolved aroud a level of reality where
prediction, affirms hypothetical issues, and a "use" of information which is
perceived as "being in the details", even though we do not happen to have
all of the details. But, having a definition of use, we looked at constancy
in the time of analysis, for both the formulae, and the data. even guesses
can be used as information, under controlled conditions. Defining a
calculus, the author D. Berlinski's book states that the world of numbers is
severely cold, yet analytic geometry presents a program in which arithmetic
comes vibrantly to life within geometry, and goes a process where the world
is made to bloom. We may consider highly abstract possibilities, lo and
behold the work of each day is unique, even beautiful. A well made system
defines a way to see a way beyond our own potential, to utopia (just
kidding).

TODAY'S LECTURE

Probability

Z Probability (handicap with statistics)
Explaining complex recurrence with statistics. Lecture IV Statistics Course
_Welcome to the the mesage board for horse racing professionals. This is
about the use of speed figures, as well as the types of standard reasons for
elimination of possibilities, which would aid with obtaining more advantage
in the win and place pools. You could say that it would make things work
more efficiently, should these approaches be used as the professionals
would.
_the connections between probability and statistics is somehow according to
the person doing the figures.. When statistics are used, we work with the
data set, to explain its quality, to some degree. Discussed ways to corrupt
data, even to move it through time changes. We managed the capture of data,
and having data as opposed to not having some data "use", as well as the
stages of processing including inferrence (reasoning by logic).
_today, we are going to give more form to these issues, by assigning numbers
to the ways that we've been discussing data. Firstly, we will add more
abstract concepts, then we'll define them as probability numbers.
_it turns out by mathematical investigation, that throwing darts involves
probability. The best possibility of hitting a winner by throwing a dart
will occur if you throw that dart toward the dartboard with the winners
names on it. And if you hit the dart board with the dart (don't throw like a
girl), that would also raise the probability. the message board's dr. cutz
advises his fellow boarders, not to play this game like an old hag.. we can
look in the general direction of those posting on this message board, to
consider the long run probability for those strategies.
_the simplest probability involves a single logical dimension. The p or q. q
means that something will not happen. q is the complement of p. so p means
that something will happen.

let P(e) mean the probability for the occurrence of an event "E".

1 - P(e) = q which says that there is a number, defined to be less than 1,
and the smaller that number is (lesser P(e)), the greater the probability
that P(e) will not happen.

_so we have established a ration for the occurrence of an event which is
proportional to the nonoccurrence of an event. and if we consider two
questions that would relate to each other, such as leaving the house with
both shoes on rather than just one shoe, then we have to consider the
chances of both events, which is less than either (unless one is a certain).
_the connection between the events is important, in that when one influences
the other the probability is greater than if these events are distinct from
each other.
_a porbability is certain if P(e) is known to be equal to 1. If P(e) is
equal to zero (0) where the event is impossible. The probability of someone
walking out of the house with one shoe and being ignited by the sun, is
virtually zero. Not because of the shoe Event but because of the sun Event
being impossible P(e) is 0 for the sun event. The combination is associative
which means the same non chance exists for him to ignite outside before
putting on that one shoe.
_next paragraph>

_once we have formed an idea about simple probability, we might also
consider statistics which gives us a way to consider the questionable data,
the central tendency, as well as variation. now, we would like to evaluate
the data as probability, according to availability, range, influence from
one event on the next, other fundamental consistency, and chance.
_on the basic level, probability is measured one Event at a time. And the
measurement of probability is a measure taken from a series of occurrences.
If the coin is flipped 20 times it could end up as ten heads, and ten tails,
the point is that we are looking at the probability after a series of
occurrences, the twenty flips of the coin. If heads was considered a H and
tails considered a T, the probability of H would be H / H + T

20 flips with 4 heads and 16 tails would evaluate to the possibility for
heads as:

4 (heads)/ 20 (flips of coin) = 4/20

the idea of knowing that the coin will land 50% heads and 50% tails, is not
necessarily true with the information that was just collected. many
judgement call are going to be made such as, it that heads or tails ? The
thing about probability is that the samples of the data, and the standards
we use to judge the values (center of opinion) and the concepts we use to
identify or collect our data, would be related to what probability we migh
arrive at for some series of occurrences. In theory though, if we have 10
heads 6 tails and 4 unknown, from 20 flips of the coin the probability for
50% heads and 50% tails is less the .50 (50%). it is impossibile and
unlikely unless we throw out the unknowns..
_next paragraph>

_so if we have more unknowns, then we also know that there is less
probability for two occurrences of the event, in the series of events we are
looking at. if we had three certain known events (together) that happen
every time (blood pressure, breathing, consciousness), then we can look at
the probability that the person is alive as a certainty. Without knowing
some of the information, we cant be certain.
_sometimes prople use a system for playing a lottery, they use the same
numbers (their birthday), or they use a machine to generate random
combinations since that gives them an even chance (by definition). Their
stategy would involve "expecting" the unknown. And as we all work with
statistics, its evident that playing the strategy that is a known loser, is
the same as throwing darts like a girl
next paragraph

_at this point, we can arrange the problem we want to look at, and affirm to
our supporters that we have no support from the old hags in the room. . the
problem with probability involves the evaluation of likely Events or P(e).
If an event is Certain to occur P(e) =1, Likely P(e) > .50, Uncertain P(e) =
..50, Unlikely P(e) < .50, Impossible P(e) = 0

_the events can be independent of each other, and we can add up their
probability of occurrence, then divide by the number of Events. So the
probability for two Uncertain Events occurring is .50. For two Impossible
Events is 0. For two Likely Events > .50. And for two Unlikely Events < .50

_there is no scale to measure P(e). The numbers can be manipulated without
consideration for the statistical problems of data collection, so all that
has to be out of the way before probability of an Event is considered.
Events can be either simple or complicated, but each qualification has to be
met in order for the Event take place. Its the "use" of the Event. The Event
with respect to its P(e) probability.
next paragraph

_one aspect of estimating probability would be an estimation of the effects
of chance occurrences. Unless we can know what the chances are, we can not
predict P(e) with accuracy. this issue grows harder to cope with as the
number of Events of the series grows. The proportion would seem to be
related to the question of statistical variation in time, the uncertainty
can operate as a malfunction in the process, similar to a machine that
works, but unreliably. The entire day's production run can not be estimated,
between zero % and 100%, there's no estimate there.. Whenever two events are
not independent, then if the chance of one is unknown, the other's chance is
unknown.
_the dimensions of probability actually include the P(e) and the complement,
NOT P(e) or the NOT.
_for the element chance to be negligible, we have to consider what it does
to the odds. for 5-1 odds the element of NOT is 1(chance is also unknown).
The element of the P(e) probability of an Event is 5 (with chance as an
unknown here). That is where 5-1 came from. Look at
5(+/- chance)/1 (+/- chance).

Then look at

17(+/- chance) / 130 (+/- chance)

with each chance the variation could be swayed either way. . i think i got
one right .
next paragraph

the management of P(e) can be acheved by weighing the probabilities, along
with the possibilities. when you have a choice of 8 horses to bet on,,
considering them equally leads to a weighed probability of 1/8. Odds are 8-1
against your bet. The weight is not a problem here, because thinking people
can make models, and use counting strategies i.e, play two , or five,
numbers out of 8. The probabilities of winning seem likely better when the
grandson of Grindstone is running in a four horse race against two maiden
fillies and a three legged pony. But there is an example for a probability
problem that points out there is a 37% for a hat check girl to give you the
right hat without knowing who you are even while she is wearing a
blindfold..
_there is the idea of continuous but rare possibility of rare events, which
will make things look easier than they really are..


so chance turns out to oppose probability .. the typical person thinks the
two are similarly defined, but actually the idea of probability denies the
random chance will appear. Probability is the avoidance of Chance ..

============

Welcome to the message board for high flying big-time hot shot superstar
handicappers. This is a preview of another of the weekly study topics, this
one: handicapping with data and statistics-out of control variations.
_and i figured out what to do in the training of Funny Cide .. he can go a
distance carrying weight, and would be better in handicaps where he would
have no problems with weight of 124 lbs.

_the upcoming discourse is about how to address the times when the target of
the handicapping process is badly missed .

_how to judge the possibilites of wagering
.. Out of control variation
.. The process is "tightened"
.. "Unwanted looseness or noise" is removed
.. If the process is not reliable, make it so
.. "Reliable" means statistically reliable
.. Variation is expected but we want variation within acceptable limits

_Central tendency can be out of control
.. Drift or trend : seek stability
When process is in control, it is ripe for improvement- Reliability first,
then experimentation
.. Try modifications, redefinition, simplification
.. Alternative approaches, including elimination

-------

_a manager for a bank thought he was managing his staff well. every week he
chose the big champion employee, and the employee with one foot out the
door. but the next time he did this, the superstar was someone else, and the
office dunce was someone else. the bank manager, thinking about personal
initiative, intelligence, loyalty, did not have any idea why none of the
standards he considered, was relevant to the management job he was doing,
poorly. he created a hostile work environment where no one showed any
initiative to improve anything for the company. there investment systems
that are managed as badly. investment takes information, strategy, trial and
error. the exploitative system would capitalize on human error, poor
judgement, bad reasoning, and helplessness
------------
a group with a purpose is a type of association. a group that lacks a
purpose is also a type of association. there is no similarity between the
two groups, because one is real and the other is imaginary. but a violence,
stupidification, rape, bastardization and terrification, don't suffice as
purpose. so the misidentification problem is broad. a group that is
solidified by stupidity and violence, with the purpose of terrification and
bald deception is hardly suitable for a scientific investigation.
_a preliminary evaluation of a process should show it to be useful, and
controllable, otherwise it is producing mistakes. out of control variation -
bingo was his nameo
----------
_the map illustrates associations of imaginary issues of ethnicity, history,
nations, power, and glory.. ole glory ?
_accounting for change easier for some than for others. the value of the
system needs to be identified, before anyone with a legitimacy will be ready
to work with it. human error is not the issue, just the truth that there is
no strategy in worthlessness
=======
_if the problem is not human error due to fatigue, impatience, psychological
trauma, how do we invent modification. give the computers a good breakfast?
the technology cant be cited for absence from work, or lateness..
_only when we know how the system is supposed to work, can we suggest
improvements. if we can effect large changes we have in a way, invented a
new system. most times this would not be the focus.
_to design a modification, we need to consider each variable for
modification of its effect. to achieve the modification all members of the
process have to judge their mistakes, as they are part of the process being
modified.
_when the environment for improvement has been created, each modifiable
variable is checked against possible (probable) changes.
_probability is related to modification and improvement. but chance is
related to the regularity we depend upon, and its variation. we can rethink
and reinvent the system if we want, when the change is dispersed it might
effect the whole system, even when it is a slight modification.
_reducing the imbalances, keep repeating the improvement cycle, restabilize
and recalculate. ..then reinvest and repeat.
_control the sample being tested. have the explanation of the sample in
order to reduce the mistake of judgement of the test. and note that many
small chances for improvement will be lost, when we fail to detect their
effects.
_to identify the wagering issues, the identifiers have to be studied.
statistics are useful to assume (forecast) future results
_business research is not easy or shy of failure

====
end


.