This system parallels the system we described in the thread, Generating A Set Of Key Numbers For The P3
We'll look for 4 distinct numbers, which we'll call key numbers. We are going to shoot for at least one of these key numbers being present in at least 3 out of the next 5 draws.
In the P4 case if you take any 4 distinct numbers between 0 and 9, in the long run,at least one of these numbers will appear approximately 76% of the timein the past draw sample set. Since it's common for hot spots to occurover a short run, we should also be able to find a set of 4 numbers, fromwhich, at least one will occur in 100% of the past N draws, for somenumber N.
When we experimented with this hypothesis, we often got many sets of 4 key numbers. To thin out the key number sets, we added an additional requirement. Instead of demanding that at least one number in each of the past draws be present in a 4 number set, we bumped the "at least one number" requirement to "at least two numbers". In many tests, we never got more than two key number sets, although it's not uncommon to get more than 2 every now and then. In such a case pick one set and stick to it.
In a seperate blog, the C code for generating the execuatable code will be provided.
In the meantime, here are the key number sets keyNumberFinderP4 gave for a few states, good from 3/24/09 - 3/29/09:
NY Day and Eve: Set1 -> 2-4-7-9
Ga Day and Eve: Set1 -> 1-3-7-8
Texas: Day Set1 -> 1-7-8-9
Eve Set1 -> 1-2-3-9
NJ Day and Eve: Set1 -> 0-1-2-4** Set2 -> 0-1-4-5**Set3 -> 1-2-4-7** Set4 -> 1-4-5-7**
Set5 -> 2-4-5-7
Ca: Set1 -> 2-6-7-8
Keep us posted.
jayemmar