Jimmy
Thanks for the link. This thread may take on a life of it's own if we can ban togeather and take another look
at the inner workings of random within a closed system. Clusters should not be associated with patterns
in the general sense but more like the DNA of randomness. Everyone that has ever looked at the skips for
any number should have noticed that the highest hitting skip is (0) meaning the number hits more often in
the next drawing then any other value. I have found very few exceptions to this rule as it seems to be a
product of randomness which ties in with the link provided by Jimmy. Anyone that wants to improve there
understanding of the lottery should be very very interested in this. Tracking skips has been used by many
to help predict a time frame for the next occurance for that number. I track skips and clusters much the same
way but how I use this information may be very different. Once one trains the brain to sort information in a
certain way it becomes a part of the natural process used by that person. When this process is different from
the common understanding then the ability to effectively convey an idea is hindered. First we must break this
data into strands of possible outcomes. I will use a (1) to indicate a hit and a (0) to indicate a skip.
(1) clusters first
010
0110
01110
011110
0111110
01111110
011111110
0111111110
01111111110
011111111110
now (0) clusters
101
1001
10001
100001
1000001
10000001
100000001
1000000001
10000000001
100000000001
The overall clusters can be much shorter or longer depending on the data that is being tracked and which value you
are looking for. One needs to analyze both hits and misses because each contains usefull data. I use this same
logic for tracking everything I use be it a filter, digit, group, ect....
Example tracking string
000110000101001111001011000001000001000000001000100101101101011101111000001010010010001
From this data one can see that the value has not shown for the last three draws then it hit twice and then
skipped 4 draws. The left most value in this string indicates the current draw so reading the data from left to right
would be the decending order from latest to earliest. This is not real data and is used for exmaple only.
This string has clusters of hits and misses and appears to show no consistant data that could be used to help
predict the next value. One can count the custers of (0's) that were greater than three of which there are six.
From this a person might say that it is logical that a (0) would be a good choice for the next drawing. One could
also say that in this string (1) has hit 33 and (0) showed 54 times and from this calculate an average hit rate
and percent.
average for (1) = 87/33 =2.63
average for (0) = 87/54 =1.61
percent for (1) = 33+54=87, 33/87 = 38%
percent for (1) = 33+54=87, 54/87 = 62%.
One could then make the caluclations using a smaller sample using say 50% of the data and then make a choice
based on the performance comparing the results. This however will lead to as many misses as it does hits. I
don't say not to use this sort of data but you need to add in some of what makes random, random. If clusters
repeated at a constant rate then random would not be random. If the value that makes up the cluster remained
constant then random would not be random. Knowing what the random element is can explain why and how it
turns what looked like a pattern into a liar at the exact time you choose to play it. Counts, average hits, percents
are needed but adding a bit of randomness is also important. To do this takes a little rewiring of the brain while
still maintaining a common sense approach. When you study the data from some random event and find what
looks like a pattern first you must consider the time between the events that make up the pattern. If random
repeated it's self on a regular bases then it would not be random. In my many test I have found that random
can produce two or more or the same values in succession and this gives the appearence of a pattern but it is more
a random process going through a sequence of possible outcomes so most of the time you see what looks like a pattern
is nothing more then the wheels on a one arm bandit lining up. If you know how many spaces are on the wheel
then this can help you calculate the odds for it happening in the next draw. We need to know this before we can
choose what to play, it's like putting random against random. Breaking the data down to a 50/50 choice as I have done
above by converting it to a binary style string works very well with this logic. There will always be mistakes in the choices
we make and I think that everyone expects this. A very simple test that will help you understand this can be done by
looking at a hit/skip for your lotteries numbers. You will find that the best time to play a number using hit rates is right
after it hit. The problem is that you have 5 or 6 numbers that hit in the last draw and while all of them could repeat, my
guess is that they won't. The second best time is one draw between and the third best is two days out and so on and
on until you reach the threshold value where this no longer applies. The problem that most people face is which one
of the skip values should I play. By looking at each of the possible values one at a time which gives you a yes or no choice
you can then analyze them based on the random element. I have found that random works on a random cycle that can
be reduced to a series of smaller cycles, Kind of like the pointers on a clock certain events are like the second hand
while others are like the minuet hand and some like the hour hand. Second hand events happen most then minuets
then hours, days, months, years, ect.....
I will try to add more to this as time permits
RL