|Posted: November 1, 2009, 4:23 pm - IP Logged|
Running some numbers through the database on GA F5 gives some interesting stats.
Got to play the percentages, just like baseball. .333 batting average means that one out of
3 at bats gets a hit. So if he is 0 for 2 he is probably going to get a hit at the next at bat.
But if the doesn't, he will be at 0 for 3 then 2 for 6 is expected.
The same with right handed pitchers vs right and left handed batters.
In GA F5, there are 56 combinations of 0s, 10s, 20s, 30s.
Numbers from 1-9, 10-19,20-29,30-39
The top 4 combinations are from even distribution of the
0s, 10s, 20s, 30s and another. They total up to 26% of all draws.
And the other 52 combinations add up to the other 75% of draws.
So an even distribution would have a 25% chance and
something weird or a result missing one or 2 sets of tens would yield the other 75%.
Here is a sampling.
There has not been a draw of all 30s.
This is a filter that I am using to figure the possibilites.
I have another query that lists the draw numbers individually, the draw dates and the number of
days prior to that date that it hit.
Then another query that tells me where each number is currently at.
For example, 28 has been going cold to warm.
I try to predict at along those lines.
If something is getting cold, then it does not get really cold
as it hasn't before so its time is coming up.
Last night, it gave me 16 that I designated to play yesterday.
I only had the money to play one line.
Of the numbers that were drawn, 4 were on my list of 16.
The other was the repeat of the prior day which I didn't choose.
It's a start to try to narrow things down.